Investors are getting nervous about AI infrastructure spending just as the Big Tech companies double down. Over the past two weeks, Alphabet, Amazon, Meta, and Microsoft have outlined plans to spend more than $740 billion on capital expenditures this year, which Wall Street is reading as AI infrastructure spending, including data centers and chips. Earlier in the AI boom, investors largely rewarded such big outlays, viewing them as a sign of confidence in the technology’s growth potential. But as we enter the fourth quarter, that enthusiasm seems to be giving way to concerns about negative free cash flow (spending more cash than a company brings in), mounting debt, and whether these enormous investments will ever generate enough returns. The anxiety spilled across the market on Friday. The Nasdaq fell 2.9%, and chip stocks took a drubbing, too: Nvidia dropped 4%, Broadcom more than 5%, and Qualcomm 11%. Alphabet, which plummeted 12% on Thursday, its worst day on record, wiping out $500 billion in market value, fell again. A big part of the concern is that companies that used to have overflowing free cash flow from their core businesses are getting hit as AI threatens the economics of those businesses. Alphabet provided the clearest example of that this week. Google’s parent company reported negative free cash flow of $9.1 billion in the third quarter, its second consecutive quarter in the red, even as quarterly capex grew to $51.4 billion. Alphabet expects to spend up to $205 billion this year, and expects that spending to keep rising in 2027. Meanwhile, the wild success of AI Mode — Google’s AI-powered chatbot feature built directly into Search — is eating into the company’s money-maker. Google’s Search advertising business, which makes up roughly half of its revenue, saw growth slow to 12%, its weakest pace in five quarters. That leaves Google walking a tightrope. It’s spending billions on the AI technology that could undermine the search business that pays it. Amazon is in a similar boat after reporting negative free cash flow of $42.4 billion over the past 12 months, compared with $14.8 billion in positive free cash flow the year before. The company says it expects to spend $220 billion in capex this year. Even though revenue at its cloud business, AWS, grew 39%, Amazon’s stock tumbled 7% after it reported earnings Thursday. Meta is also burning cash, reporting negative free cash flow of $2.1 billion in the third quarter as capex outpaced its operating cash. It’s guiding to as much as $145 billion in capex this year. The company noted that Muse, its new free personal AI agent, is a bit of a money pit (it’s “very compute-intensive to run,” CEO Mark Zuckerberg has said). But ad revenue grew 26%. Still, no amount of good news saved Meta from taking a 10% hit on its stock last week. Microsoft was the only exception when it came to free cash flow — it recorded $18.6 billion in positive free cash flow in its latest quarter, with spending on capex at $45.2 billion, less than the $50 billion it projected. Microsoft said that wasn’t an intentional pullback from AI spending, blaming delays in capacity from third-party providers. As a result, shares of smaller companies that rent out AI computing power sank. It’s worth noting that the money to pay for all of this AI infrastructure isn’t just coming from the Big Tech companies’ profits. It’s also showing up in stock and bond sales. According to S&P Global Market Intelligence, Microsoft, Alphabet, Amazon, Meta, and Oracle raised a combined $326 billion in debt in 2026, triple the amount raised last year. Demand for AI compute remains strong. Google, Amazon, and Microsoft all saw gains in their cloud businesses. But investors are getting spooked by the costs of meeting that demand. These muddied waters could complicate Anthropic’s highly anticipated IPO. The AI company filed to go public earlier this week, and its future growth depends on securing enormous amounts of compute. Financial Times reported Friday that Blackstone has warned investors the company might not qualify for an investment-grade rating after its IPO, meaning it might end up being more reliant on the hyperscalers and third-parties to fund its infrastructure ambitions.
Intel’s stock suffered one of its worst days ever on Friday, crashing around 28% after the chip giant reported a weaker-than-expected outlook for the fourth quarter. This rout reduced the value of Intel’s holdings by the government by $12 billion, CNBC reports. The Trump administration had the government buy 10% stake over a year ago. Nvidia, which also took a big stake in Intel in 2025, saw its stock drop a bit on Friday as well. Much of the tech industry — especially chipmakers — were caught in the sell-off. Intel’s stock is still a profitable investment for the government so far (and for Nvidia), as it’s up more than 70% for the year, even if it’s dropped significantly from the sky-high prices it commanded over the summer. The scare-off was caused when Intel warned to expect revenue of $15 billion to $16 billion in the fourth quarter when analysts were looking for around $17 billion. It also warned that its gross profit margin would decline to 37% from 43%, and it was investing so heavily, it expected no positive cash flow. The main problem for Intel is that PC sales are being hurt by the AI industry gobbling memory and other components, and causing PC prices to rise. “Our customers began signaling a PC slowdown based on the memory prices and availability. Year-to-date system endpoint prices have increased significantly, placing additional pressure on already limited IT budgets,” Intel CEO Lip-Bu Tan warned on the call. The bright side is that revenue from Intel’s data center business grew 78% over the year-ago quarter. But it is hamstrung on that front until it can bring more capacity to its fabs and book more external customers for them, competing with powerhouse Taiwan Semiconductor Manufacturing Co. One of Intel’s new fabs in Arizona is operational but the next one isn’t expected to be online until 2028.
Intel Panther Lake 18A At Computex 2025 2 Intel reported third-quarter 2026 earnings with revenue of $16.5 billion, up 21% year-over-year. Non-GAAP gross margin reached 44.1%, and non-GAAP EPS came in at $0.52. Data center demand is surging while the client PC business is struggling (mightily) under memory market conditions. We do not always cover these, but Patrick asked me to do this one because it shows a trend we have seen recently at STH. Intel Q3 2026 Results Show Contrasting Segments First off, at an overall level, Intel’s headline numbers are relatively positive. Up almost 21% Y/Y is good, but last year was not great for Intel. Intel Q3 2026 Earnings Financial Highlights The Data Center and AI (DCAI) group delivered $7.3 billion in revenue, up 78% year-over-year. David Zinsner, Intel’s CFO, called it Intel’s strongest data center revenue growth in 15 years. Segment operating income reached $3.0 billion with a 41.5% margin. One of the interesting bits on why we were covering this announcement is that the demand in the market has shifted towards AMD’s 192-core Turin parts, which Intel does not have a direct competitor for (maybe Clearwater?), but partners tell us that Xeon 6 is selling because basically everything is selling. Intel Q3 2026 Earnings Data Center And AI Demand is strong across hyperscale, enterprise, and neo cloud customers, with “Purpose Built Silicon” revenue more than doubling year-over-year. Lip-Bu Tan, Intel’s CEO, said Intel’s immediate priority is unlocking additional supply to meet extraordinary demand for the Xeon family. On the call, he said that substrate availability is limiting output and it would take some time to ease. The client side looks very different. Client Computing and Physical AI Group (CCPG) revenue fell 5% year over year to $8.1 billion. Overall inflation is well north of zero depending on the index you use, so revenue falling is not good. On the call Intel cited industry-wide memory shortages impacting the total addressable market. Higher DRAM and NAND prices are pushing system costs up and hurting demand. On the call, it sounded like Intel was looking for 18A supply to be used on the data center side pending substrate availability. Intel Q3 2026 Earnings Client Computing And Physical AI Intel expects the PC slowdown to continue at least through the first half of next year. Memory and NAND pricing pressure is expected to persist into 2027, which means client system costs are likely to stay elevated for a while longer. For those who see today’s memory prices and think they are shocking, the earnings call commentary did nothing to make me think Q1 would see a decline in memory pricing. Intel Foundry posted $6.7 billion in revenue, up 58% Y/Y, though that figure includes intersegment transactions rather than external customer sales alone. The segment still recorded an operating loss of $1.9 billion with a negative 27.7% margin, but that is narrower than the $2.3 billion loss a year ago. Intel Q3 2026 Earnings Intel Foundry Improved yields and cycle times are reducing wafer costs. Intel released the Intel 14A 0.9 process design kit in October as committed, and expansion continues across sites including Ohio. Final Words We do not cover all earnings calls, but Patrick’s thought was that this was an important one since it also tells a story of STH. Folks wonder why we have been doing a lot of AI desktop and server/ networking coverage, and fewer mini PCs. Intel’s Panther Lake is a cool product, but it is not running larger local models. Right now, DRAM and NAND are being priced for the value they bring attached to AI nodes. Likewise, the NVIDIA-Intel investment and the many NVIDIA GPU-based systems with Intel Xeon CPUs are helping drive Xeon 6 sales and push pricing up. As a result, we are seeing the market demand for non-local AI PCs drop due to higher prices, while the AI-related markets are cooking. We know folks want more mini PC reviews, but this is a fundamental industry challenge. A smaller client-side market also means fewer small companies trying to innovate on form factors, while the data center market is the exact opposite. This came out very clearly in Intel’s earnings.
'Gamble on democracy'
8 h ago
Spencer Pratt appeared to be a shoo-in to win the June primary in the Los Angeles mayor's race, based on millions of dollars wagered on prediction markets aiming to call the race. Pratt was an unlikely candidate for mayor. The former reality TV star had never held elective office, but after the Palisades fire destroyed his home, he turned his criticism of the city's response into a campaign rallying cry. Using prediction markets including Kalshi, Polymarket U.S. and Polymarket International, speculators flocked to the race, which drew more than $47 million in trades, making it the most heavily traded mayoral contest of 2026. On the eve of the June primary, one market gave Pratt a 78% chance of advancing to the November runoff. The prediction markets were wildly off base, and Pratt finished third. Pratt supporters, though, used the failed speculation to cast doubt on the integrity of the election. That drew a response from the city: "Prediction markets can add to that confusion when people treat their projections as more authoritative than the actual votes being counted," Dean Logan, the registrar-recorder for the Los Angeles County Clerk's Office, said in a statement to CNBC. Prediction markets' failure to call the election correctly also raised the possibility that markets can be manipulated to make candidates appear to have more support than they really do. Exclusive reporting from CNBC shows that this risk is real. The dissonance between voter polling, speculation-fueled prediction markets and the actual outcome has grown in importance as election wagers become a prominent feature of elections. Odds and gambling platforms do not use methodologies used by traditional political polling, and therefore are not substitutes for political polls. Polymarket and Kalshi are the two largest prediction market platforms. They let people speculate on almost anything — from which team will be the 2027 NBA champions to whether the U.S. will confirm that aliens exist. Already, on these platforms the amount wagered on contracts related to the 2026 congressional elections has more than tripled wagers on the 2024 races, according to the Anti-Corruption Data Collective, or ACDC, a nonprofit based in Washington, D.C. And people are increasingly using the platforms as a source of information. For midterm election markets specifically, Kalshi reports that viewers outnumber actual traders by a 5-to-1 ratio. Spotting suspicious tradesDespite the Pratt race being a heavily traded, liquid contract, Kalshi detected and reported on price spikes that occurred during trading on that race in a case study it published in July. A price spike is a temporary but large price jump that can make an outcome look more likely to happen, according to experts. Price spikes can happen for a variety of reasons, such as new information changing trader expectations or a trader buying a large number of contracts, which can drive up contract prices as cheaper ones get bought up. On May 20, the company said, a trader on Kalshi spent $1.25 million on "Yes" contracts on Pratt becoming LA's next mayor. One of the trades caused the price of contracts on Pratt's victory to jump from 29 cents to 45 cents, suggesting that traders saw his chances of winning rise from 29% to 45%. However, according to Kalshi's case study, within nine seconds, the price reverted to earlier levels and the markets self-corrected. "You can absolutely guarantee there are rational traders in that market willing to take those trades and bring that price right back down to where it belongs," said Robert DeNault, the head of enforcement at Kalshi, who is tasked with tracking unusual activity on the platform. "And by the end of the day, his price was lower than when it began." Los Angeles mayoral candidate Spencer Pratt hosts a campaign block party event on 10th Avenue in Los Angeles, May 20, 2026.Robert Gauthier | Los Angeles Times | Getty ImagesBut during the LA mayoral race, there were numerous price spikes across Kalshi and Polymarket International that pushed the odds in Pratt's favor, according to an analysis for CNBC by ACDC. The group didn't consider trades from Polymarket U.S. in its analysis. Data that ACDC collected confirms Kalshi's findings. In the case study, Kalshi provided insight through about 30 minutes after the large order was placed. However, the group found that if you expand the time frame to three hours after the spike, the price lands three cents higher and that level is sustained for about two hours. In the long term, the price does correct. But according to ACDC's research, the longer a price increase lasts, the greater the risk of that information being interpreted as market sentiment. And as ACDC's research suggests, election contracts are vulnerable to unusual price spikes — similar to those in the Pratt race. The group found that of 167 congressional markets across Polymarket International and Kalshi worth more than $100,000, 38 markets — or about 23% — experienced price spikes moving the contract price 30 cents or more. One example of such a spike was the Kalshi contract on whether Juliana Stratton would be the Democratic nominee for the Senate in Illinois. On March 3, according to ACDC, two purchases stood out as large — the median order was $12 before election day. One buyer spent $5,000 on contracts that favored the longtime public official's candidacy, taking the price from 47 cents to 79 cents. The price normalized after three seconds, going back down to 47 cents, but about one minute later, it shot back up to 99 cents after $3,000 of additional contracts were purchased. The price eventually sank back to 52 cents — 5 cents more than where it started — meaning a person or persons spent $8,000, and, in the process, drove up the price that made a Stratton victory look more likely. Two weeks later, Stratton won the Democratic primary. Stratton declined to comment on this story, but a spokesperson from her office pointed CNBC to a poll from Public Policy Polling released on the same day as the price spike. It showed Stratton leading. In a statement to CNBC, a spokesperson for Kalshi wrote, "As is common on Kalshi, the market corrected quickly to a price near fair value." ACDC said it's impossible to know the motive behind these wagers. While profit could be a factor, it is an imperfect one. On prediction markets, participants can buy and sell contracts valued between $0 and $1, with the price reflecting how likely traders think an event is to happen. The more expensive a contract, the smaller the potential payout. Winning contracts settle at $1, while losing contracts become worthless. So traders buying in at 99 cents expect to earn only 1 cent if their predicted outcome happens, and transaction fees charged by prediction markets can also eat into that profit. "If the theory is that prediction markets are accurate because of a price motive, [a lack thereof] begs the question: What type of goal does someone making a trade like that really have for themselves?" said Michelle Kendler-Kretsch, a researcher at ACDC. Influencing election outcomes may be an answer. Voters were between 10 and 20 percentage points more likely to believe a candidate would win if prediction markets favored the candidate, no matter what the polls reported, according to a new study funded by the Brennan Center for Justice, a New York City-based left-leaning advocacy organization. That survey, which took place in September, included the perspective of 1,994 registered voters. Data ACDC shared with CNBC also shows shifting the odds is even easier and cheaper in markets where not many wagers are placed. Across 18,000 active congressional markets on Polymarket International, Polymarket U.S., and Kalshi, 88% could have their prices moved by 10 cents with a single wager under $1,000. Policing market manipulation Still, Kalshi market outcomes are often correct. "Across U.S. elections since 2024, when Kalshi's markets favored a candidate to win, that candidate went on to win roughly 9 times out of 10," the company wrote in a website post. Further, The Washington Post examined 268 candidates in this year's primaries who received a 70% to 80% chance of winning on Kalshi or Polymarket on any day in the two months before the primary election. Three-quarters of them won. Kalshi says it's working to assuage election distortion concerns. It added new transparency-centered features to its midterm election markets earlier this month to help users understand how much confidence to place in a contract's odds, including labels for markets with little trading activity and live feeds showing individual trades. Polymarket also rolled out its own election-specific surveillance system "to detect potential attempts to manipulate election markets, distort probabilities, or manufacture apparent market consensus."Kalshi told CNBC it uses algorithms to generate alerts based on parameters for unusual trading activity, including fishy trading times and volumes as well as price jumps. It's unclear if Kalshi's internal protocols alerted over the Pratt trades; the company didn't respond to CNBC's request for clarity. Both Kalshi and Polymarket U.S. are governed by the Commodity Futures Trading Commission, or CFTC, an independent federal agency whose rules are designed to prevent market manipulation and insider trading. Trading prices and activity are publicly visible, and at least in the case of Kalshi, its order book is public on derivatives exchanges and people can escalate and flag unusual trades to the exchanges. In addition, Kalshi and Polymarket U.S. collect "know your customer" information including full legal names and Social Security numbers. The outside world, however, has no insight into who is behind each trade. "I formerly worked at the FBI and in centralized finance off the blockchain," said Shana Bautista, the head of global intelligence and investigations at Polymarket. "If you want to get information on a ca
Hackers working for the North Korean regime have been blamed for a record cryptocurrency heist following a cyberattack at crypto exchange Binance. The hackers stole at least $6.3 billion in crypto from the company’s wallets, though about half of that amount was frozen as the stolen crypto traveled across various blockchains, according to security researchers. The breach is by far the largest known theft of crypto in history, though the incident has not been confirmed yet to be resolved, and the amount of crypto stolen may change. In a post on X, Binance said its initial investigation showed that the hackers broke into the company’s network after targeting a “compromised” Binance engineer’s laptop, which “allowed an attacker to gain access to a service container and submit unauthorized withdrawal transactions.” Binance said it has seen no evidence of a leak of its private keys, which would grant the hacker near-unfettered control over the company’s stores of crypto. The crypto exchange — the world’s largest — said it has now suspended all deposits and withdrawals while its investigation into the breach continues. TechCrunch reported on Thursday that the exchange had paused withdrawals across some of its blockchains. Binance said “user funds are safe,” but did not immediately provide proof of this claim. The company said it will “publish an updated proof of reserves once the remediation is complete.” Spokesperson Andie Cooper said in a statement shared with TechCrunch on Friday that some $3 billion has been frozen by industry partners. The spokesperson said Binance has “begun recovering affected assets for users and continues to work with law enforcement, issuers, blockchain intelligence firms and industry partners to trace, freeze and recover additional funds.” It’s unclear how the hack went down or under what circumstances the employee’s laptop was compromised, but it’s not uncommon for security incidents to stem from the use of compromised software on employee devices. Software is generally built from many components, including open source projects, which hackers regularly target, compromise and modify in order to steal data from devices that the affected software is installed on. Other times, malware bundled with regular-looking apps and games can allow hackers to steal usernames, passwords, and session tokens from infected computers, which may let them log in as if they were a legitimate employee. Binance will face questions about why the company’s security allowed a single employee’s laptop to access systems capable of controlling billions of dollars in funds. Spokespeople for Binance did not respond to an email with questions about the breach. TechCrunch has also asked the company who, if anyone, performed a security audit of Binance’s network and systems prior to the hack. Companies regularly ask third-party companies to stress-test their security to identify weaknesses that could be exploited by attackers, but it’s unclear if Binance sought such a review. Security researchers at TRM Labs, a blockchain intelligence and forensic company, attributed the breach to North Korean hackers, and called it the “largest hack in history.” Researchers describe the network of hackers, known in the cybersecurity industry as TraderTraitor, as a North Korean cell focused almost exclusively on stealing crypto from large companies and exchanges. This group was also blamed for the now-second-largest heist on record, a $1.4 billion theft from crypto exchange Bybit in 2025. According to TRM, the theft of Binance’s billions in crypto took about 20 minutes. Blockchain security firm Elliptic also attributed the Binance theft to North Korean hackers. Elliptic said some of the stolen funds ended up in the same wallet that was used in a previous cyberattack on crypto exchange Bitget earlier this year, in which hackers stole $351 million. The U.S. government and its allies have long warned about the threat from North Korea, which remains cut off from the U.S. and the global financial system for its ongoing development of its nuclear weapons program. Researchers say the breakaway country’s sole motivations are to steal as much crypto as possible in order to continue developing its sanctioned weapons of mass destruction. Updated with statement from Binance.
There’s a big uproar about AI agents taking over commerce and orders, but that might not be the case just yet. Delivery company DoorDash revealed during its third-quarter earnings call that it has only seen “several hundred” orders from third-party agents like Muse, Instinct, Fo, and others. For comparison, the company said in its 2025 report it completes an average of 3 million deliveries daily. “I would say that this [using agents to order] is something that is still a very dynamic area. And my take on all things technology is you never want to fight gravity. But at the same time, you don’t want to be delusional,” DoorDash CEO Tony Xu said in an answer to an analyst’s questions about the topic. “What I mean by that is if you look at where AI agents are today, we’re seeing maybe several hundred orders in total from some of these third-party agents,” he said. Xu mentioned that users are still predominantly placing orders through DoorDash’s app. However, the company wants to learn more about agentic use cases and develop a healthy relationship with third-party agent makers. He added that, in the future, there could be an opportunity for it to advertise within these agents. DoorDash has its own agents, including a text-to-order agent it launched in September, a CLI agent introduced in July, and a chatbot called Ask DoorDash it launched in June, which allows users to place orders through prompts and photos. The company also released an MCP to enable ordering from other chatbots for corporate ordering. The space of agent-based ordering is evolving, as websites are still deciding whether to allow them to access their content or not. Amazon has been blocking such personal agents, but Walmart has laid out a red carpet for them in order to snag some business from its competitors.
Nvidia has begun securing advanced substrate capacity in Taiwan, Japan and South Korea, using capacity reservation agreements and upfront payments to support supplier expansion as next-generation AI accelerators ramp up...
Simplifies complex testing while reducing reliance on manual scripting and specialized expertise Traditional lab automation requires engineers to manually configure script and maintain connections across a growing ecosystem of test and measurement tools. As network and security architectures evolve at an unprecedented rate, manual test configuration and automation have become increasingly difficult to scale. As AI agents and agentic workflows become part of network and security operations, test infrastructure must be able to participate in those same AI-driven workflows. The integration of VIAVI’s award-winning MCP Server Framework simplifies the interaction between infrastructure testing tools and AI clients, and reduces reliance on GUIs, scripts or APIs. The framework exposes testing functions as standardized tools that AI agents can discover and execute through conversational commands. By translating natural-language requests into test actions, the MCP Server Framework shifts the focus from needing deep product, testing and scripting expertise to intent- and outcome-driven validation workflows for complex test environments. The MCP Server Framework includes an autonomous error correction loop to detect and self-correct test tool misconfigurations, with human review checkpoints maintaining oversight and safe operations at critical points in the testing process. Structured AI workflows provide the context needed to improve reliability, while real-time data allows tests to adapt as results arrive. “Network infrastructure is evolving faster than the supply of specialists who can validate it, so validation has to become AI-native,” said Sashi Jeyaretnam, Senior Director, Security Product Management, VIAVI. “The MCP Server Framework helps close that gap by reducing time-to-test and operational cost while making complex validation accessible to a broader range of engineers.” About VIAVI VIAVI (NASDAQ: VIAV) is a global leader in test and measurement and optical technologies. Our test and measurement, and resilient Position, Navigation and Timing (PNT) solutions enable and secure critical infrastructure ranging from data center ecosystems and communication networks to military, aerospace, railway and first responder communications. In addition, we develop and advance technologies used in high-volume optical applications across anti-counterfeiting, consumer electronics, aerospace, industrial and automotive end markets.
Summary: Reports revenues of 104.6155 trillion won, operating profit of 83.3338 trillion won, net profit of 39.7756 trillion won; Quarterly sales surpass 100 trillion won for the first time CEO Kwak Noh-Jung explains the background to the decision to discontinue the review of Solidigm IPO, as well as the company’s shareholder return policy Sales of HBM4 and high-value-added server products expanded; HBM4 expected to be the mainstream product from the fourth quarter Long-term agreements secured for more than 50% of 2027~2028 production capacity; A structure that reflects market conditions and ensures stable profitability Most of the gains from the Kioxia investment to be shared with shareholders, with approximately 100 trillion won in shareholder returns to be executed this year SK hynix Inc. announced that it has recorded 104.6155 trillion won in revenues, 83.3338 trillion won in operating profit (with an operating margin of 80%), and 39.7756 trillion won in net profit (with a net margin of 38%) in the third quarter.During the earnings presentation, CEO Kwak Noh-Jung explained the background of the company’s decisions regarding the Solidigm IPO and its principle to share the returns from strategic investment with shareholders. “The IPO was intended to raise external capital for the expansion of Solidigm’s production facilities, and thereby to lighten our own investment burden and broaden our capacity for shareholder returns,” said Kwak Noh-Jung, CEO, SK hynix. “However, after carefully considering recent financial market conditions and other factors, we concluded that proceeding at this time would not be the best way to achieve the originally anticipated outcome. We have therefore discontinued our review.” “Our principle has been to share the majority of the gains from successful strategic investments with our shareholders. These gains help replenish our capacity for shareholder returns, which may temporarily decline as we prepare for the future, allowing us to continue investing boldly in sustainable growth while protecting shareholder interests,” he continued. “In line with this principle, we have decided to return the gains realized from our investment in Kioxia this year to shareholders through a special dividend and we will remain committed to upholding this principle so that our shareholders can directly benefit from the returns from our strategic investments.” Click to enlarge Record quarterly results in both revenue and operating profit and revenue surpasses 100 trillion won for the first time In a market environment in which prices of key products have risen on the back of growing demand for AI memory, the company once again posted all-time-high quarterly results in both revenue and operating profit, underpinned by the competitiveness of its high-value-added products. In particular, revenue exceeded 100 trillion won for the first time in the company’s history. Revenue and operating profit increased by 32% and 38%, from the previous quarter, respectively. By product, DRAM results were driven by a significant increase in sales of server DDR5 while HBM4 shipments began in earnest. In NAND, sales of 321-layer TLC eSSD expanded, and high-value-added products for data centers continued to grow across the board. The company’s financial structure has also become significantly concrete. At the end of the third quarter, cash and cash equivalents stood at 114.5 trillion won, up 26.5 trillion won from the end of the previous quarter, while borrowings decreased by 2.8 trillion won to 15.8 trillion won. As a result, net cash increased to 98.7 trillion won. Memory supply shortage to persist through 2030 SK hynix analyzed that a virtuous cycle has taken hold in which AI-driven productivity gains and new revenue generation in areas such as software development, customer service, and search lead to greater AI adoption, which in turn drives increased investment in infrastructure. It also projected that, as the drivers of AI investment move beyond the early stage, when a small number of companies accounted for most of it, and broaden to include general enterprises and nation-level initiatives, the global AI data center infrastructure market will continue to grow steadily. The company likewise expects the memory market to sustain its growth, led in the short term by enterprise and personal AI and, over the mid-to-long term, by physical AI such as autonomous driving and robotics. The company assessed that supply, on the other hand, is not keeping pace with the growth of the market. Given technical difficulty of leading-edge processes and the long lead time required to build new production facilities, it expects the structure in which demand exceeds supply to persist through 2030. Expanding production capacity and strengthening product competitiveness to achieve a further step up in growth next year Building on these favorable conditions, the company plans to accelerate the expansion of production capacity and the strengthening of product competitiveness in order to sustain the current growth momentum, and to scale up revenue and profit one step further next year. First, the company plans to rapidly increase the portion of its 10nm-class 6th generation (1c) process technology and expand sales of DDR5 and LPDDR5 SOCAMM for AI servers, thereby further strengthening its DRAM portfolio centered on high-value-added products. To respond to surging HBM demand, the company will ramp up M15X to maximum capacity by the end of the year to increase HBM4 production. Accordingly, the company expects HBM4 to surpass HBM3E in both shipment volume and revenue in the fourth quarter and to establish itself as a mainstream product. In NAND, the company will raise the ratio of its 321-layer process to more than half of its domestic production capacity in the fourth quarter, and thereby increase the supply of high-performance, high-capacity eSSD. Over the mid-to-long term, the company intends to develop NAND into a core solution that enhances the throughput and economic efficiency of AI services, based on HBF and system optimization technologies. Long-term agreements secured for more than 50% of 2027~2028 production capacity and the company to drastically expand investment next year To improve mid-to-long term demand visibility and solidify its growth foundation, SK hynix is expanding long-term agreements (LTAs) with customers. The company has currently signed LTAs with 14 major customers, including leading global cloud service providers (CSPs), and is also proactively discussing with a number of customers the possibility of extending the contract terms, which are typically five years. The LTA volume secured to date corresponds to more than 50% of total production capacity for 2027 and 2028. In particular, by introducing a mechanism that reflects market conditions within a price range agreed upon with customers, the company has established a structure that secures stable profitability while allowing flexible responses to market changes. Backed by the demand visibility, the company is also moving to expand its production capabilities. Next year, with the Yongin Phase 1 entering full-scale operation and domestic and overseas infrastructure investment gaining momentum, investment is expected to increase significantly. The company nevertheless plans to adjust the scale and timing of its investment flexibly, after carefully weighing customer demand, product competitiveness, profitability and investment efficiency. Most of the gains from the Kioxia investment to be shared with shareholders, with approximately 100 trillion won in shareholder returns to be executed this year As previously indicated in August, SK hynix announced its additional shareholder return plan. For fiscal year 2026, the company will pursue shareholder returns totaling approximately 100 trillion won, comprising cash dividends of 30 trillion won or more and share repurchases and cancellations of 60 trillion won or more in total, including the 40 trillion won announced in August. Through this, the company intends to fulfill its commitment to return 50% or more of free cash flow (FCF). Of this, the 40 trillion won share repurchase announced in August was completed last week, and the shares will be cancelled shortly. The company also set out its principles: the dividend amount will be determined on the basis of operating results; if significant investment gains arise, they will be returned additionally through special dividends; and share repurchases will continue every quarter. At today’s meeting, the Board of Directors resolved on a third-quarter cash dividend of 14 trillion won and a share repurchase of 20 trillion won. Of the shares to be repurchased, 10 trillion won, earmarked for shareholder returns, will be cancelled in full, while the remaining approximately 10 trillion won will be used for employee compensation and is not included in the shareholder return total. The third-quarter dividend is 20,000 won per share, consisting of a 10,000-won dividend reflecting improved earnings and a 10,000 won special dividend that shares the gains from the Kioxia investment. The company has decided to return to shareholders most of the gains realized from its Kioxia investment this year. “The customer trust built through our technological competitiveness is leading to long-term contracts, and our business structure is transforming into one with earnings stability and mid-to-long term demand visibility,” emphasized said Kwak Noh-Jung, CEO. “We will carry our HBM competitiveness over to next-generation products, and we will also increase the rewards we share with our shareholders as our business outcome and investment result grow.”
MaxLinear: Fiscal 3Q26 Financial Results Generating $218.1 million, up 29% QoQ and up 72% YoY This is a Press Release edited by StorageNewsletter.com on October 30, 2026 at 2:01 pm Summary: Q3 net revenue of $218.1 million, up 29% sequentially and up 72% YoY Infrastructure business up 210% YoY reflecting accelerated adoption of optical AI data center products MaxLinear, Inc., a provider of radio frequency (RF), analog, digital and mixed-signal integrated circuits, announced financial results for the third quarter ended September 30, 2026.Third Quarter Financial Summary Click to enlarge Click to enlarge Management Commentary “Our Q3 results underscore the significant growth inflection underway across MaxLinear’s overall business. We delivered 72% year-over-year revenue growth, including 210% growth YoY in AI datacenter-centric infrastructure revenue,” said Kishore Seendripu, chairman and CEO, MaxLinear. “Strong momentum in our optical AI data center business reflects the volume ramp of our Keystone PAM4 DSP platform, as well as the strength of our expanding infrastructure portfolio and customer activity centered around our Rushmore and Annapurna product families of 1.6T-capable optical and electrical products. These solutions enable a broad range of emerging AI data center architectures. With the convergence of multiple growth drivers fueling our data center-focused business, new product introductions expanding our footprint and content, our relentless focus on innovation, operational excellence, and disciplined execution, we believe MaxLinear is well-positioned to deliver sustained growth, expanding profitability, and long-term value for our shareholders.” Credit Agreement MaxLinear announced that it has amended its existing credit agreement with Wells Fargo and other lenders to increase the amounts available under the revolving credit facility by $70 million to $200 million. Additional details regarding the amended credit agreement are described in the company’s quarterly report on Form 10-Q for the quarter ended September 30, 2026, filed with the SEC.
Circle signage during the company's initial public offering in front of the New York Stock Exchange, June 5, 2025.Michael Nagle | Bloomberg | Getty ImagesBinance said Friday that a security incident had an estimated impact of $6.3 billion, with around $3 billion in cryptocurrency assets frozen so far, and pledged to cover all losses.The company's co-CEO Richard Teng said the exchange would cover all losses from the $6.3 billion hack using its SAFU fund and company funds, with about $3 billion in stolen assets frozen on-chain so far.The world's largest crypto exchange said an initial investigation suggests "the security incident may involve a compromised DevOps engineer's laptop, which allowed an attacker to gain access to a service container and submit unauthorized withdrawal transactions." Later in the day, Teng said the attack involved a complex process rather than the misuse of a single engineer's permissions, underscoring that the investigation had not attributed responsibility to any individual. He added that there was no evidence of any wallet private key leak.Binance has temporarily suspended withdrawals while it overhauls its withdrawal security systems and has revoked the affected address authorizations, Teng said. The exchange is also working with law enforcement and blockchain intelligence firms to trace and recover the stolen funds, while deposits remain unaffected.Binance co-founder Yi He said deposits remained unaffected and continued to operate normally, while the exchange was working on enhancements to its withdrawal service and related address controls. It expects to begin resuming withdrawal services within the next 48 hours.The exchange said it would cover all losses using its Secure Asset Fund for Users and company funds, adding that customer funds remained safe.Binance also temporarily suspended deposits and withdrawals, revoked the affected address authorization and enlisted law enforcement and blockchain intelligence firms to trace and recover funds.The company said it would publish a technical report, and an updated Proof of Reserves after remediation efforts."We have found no evidence of any wallet private key leak. We will share a detailed technical post-mortem once the investigation is complete and the facts are fully verified," Binance said.The announcement followed moves by several crypto firms, including Circle Internet Group, Tether and Ripple, to freeze funds tied to the hack.Bitcoin fell more than 4% to $77,994.96 on Friday as the hack weighed on crypto markets. Shares of Circle tumbled about 5% in after-hours trading on Thursday. "Our teams are aware of reports of a potential security issue and are actively investigating," a Binance spokesperson said. "Rest assured that user funds are completely safe." Binance also said on X that Binance Alpha trading had been temporarily suspended, and that "the current verified total frozen is $2.54 billion."Outside effortsCircle CEO Jeremy Allaire said Thursday evening in a social media post that the company had frozen more than 800 million in USD Coin associated with the incident. USDT stablecoin issuer Tether CEO Paolo Ardoino also said in on X that his firm had frozen more than 1.5 billion of its stablecoin that may have been related to a potential breach. Jack McDonald, Ripple's senior vice president of stablecoins, said it was actively freezing affected funds held in RLUSD, its U.S. dollar-backed stablecoin. Later, Yi He said in a statement that the XRP and RLUSD-tied firm Ripple helped halt the movement of $200 million of funds linked to the possible attack. But those frozen funds could represent only part of those impacted. Crypto data provider Arkham Intelligence linked nearly $6 billion in digital assets — including bitcoin, ether, solana and various stablecoins — to 55 wallet addresses it has linked to the potential exploit.Arkham identifies, tracks and flags crypto wallets by analyzing publicly available blockchain data. It also solicits tips from its community members.Exploits are not uncommon in the crypto space. In 2025, Bybit suffered roughly $1.5 billion in losses due to a hack. More recently, Bitget lost more than $350 million worth of digital assets in a cyber theft incident. Over the past several years, bad actors have also drained billions of dollars in crypto from decentralized exchanges.
Binance has confirmed to TechCrunch that it is dealing with a security incident as users reported that funds were being drained from the crypto exchange’s wallets. The world’s largest cryptocurrency exchange said in an emailed statement to TechCrunch that it is “aware of reports of a potential security issue and are actively investigating.” “Rest assured that user funds are completely safe,” Binance said in a statement. “Our security teams are reviewing all relevant information with urgency, and we will provide updates through official Binance channels as soon as we have verified information,” the statement added. “Users should remain cautious of scams and should not trust unofficial links, messages, or requests for account information.” Binance spokesperson Ross O’Leary said that Binance has “temporarily suspended withdrawals on multiple networks following a security incident as a precautionary measure.” News of the breach comes hours after Binance founder Changpeng Zhao said in a post on X that the company likely had been hacked. Zhao, who no longer runs the company but is its largest shareholder, said that he expects users’ funds to be “fully covered” but did not provide an explanation for the claim. Financial tech giant Circle froze about $800 million in Tether as it passed through its networks, while crypto giant Tether froze about $1.56 billion related to Binance’s hack. Crypto watchers say at least $2 billion has been drained from Binance’s hot wallets, which are internet-connected cryptocurrency wallets that store crypto that are connected to the internet. By some estimates, Binance may have lost upward of $5.7 billion in crypto, though the numbers have not been confirmed and it’s possible that the theft is ongoing. It’s unclear who is behind the hack, and Binance did not say if it had received any correspondence from the hackers. If confirmed into the billions, this would likely rank as one of the largest crypto breaches in history, according to leaderboards tracking crypto thefts. In 2025, hackers stole a record $1.4 billion in crypto in a single breach from exchange Bybit, blamed on North Korean hackers. In what appears to be a separate incident earlier on Thursday, prediction market giant Polymarket briefly halted trading after it confirmed that a third-party company was compromised, which led to the company pausing withdrawals. According to security researchers, hackers stole around $2 million in user funds after compromising a piece of software developed by a company called DebugBear.
Traders work on the floor of the New York Stock Exchange during morning trading on Oct. 14, 2026 in New York City. Michael M. Santiago | Getty ImagesFutures tied to the Nasdaq-100 dropped Thursday evening after Amazon and Intel shares slid following the companies' latest quarterly results.Nasdaq-100 futures lost 1.7%, while futures tied to the S&P 500 slid 0.7%. Dow futures fell 13 points.In Asia, Japan's Nikkei 225 slipped 0.68% while the Topix rose 0.26%. The Kospi dropped 3.11% at open, while the small-cap Kosdaq declined 2.35%. Australia's benchmark S&P/ASX 200 was 0.65% higher.Amazon shares lost 6% in extended trading after the company reported earnings and revenue for the third quarter that fell short of Wall Street's expectations. The company also posted negative free cash flow of $42.4 billion and maintained its 2026 capital expenditures forecast for $220 billion.Intel slid 21% after issuing a weak outlook for the current quarter, pointing to a slowdown in PCs. Revenue in the company's client computing and physical artificial intelligence group declined 5% from the year-ago quarter to $8.1 billion.Sandisk also fell in extended trading, down 3% even as its fiscal first quarter results beat estimates. The data storage company announced an eight-for-one forward stock split.The declines come as investors scrutinize the AI trade.In regular trading Thursday, Alphabet shares lost 12% after the company missed on earnings and revenue in the third quarter and posted negative free cash flow of $9.07 billion. The Google-parent's slide dragged the Nasdaq Composite to a 1.4% loss in the regular session, while the S&P 500 fell 0.3%.Friday marks the final trading day in October. The month has been rocky for the AI trade, with investors paying close attention to companies' spending around the build-out. Hyperscalers Alphabet and Microsoft haven't lifted their capital expenditures forecasts.At the same time, neoclouds — the cloud-computing companies that are behind the infrastructure needed for AI datacenters — are facing investor scrutiny over their debt levels. CoreWeave, which has seen a surge in its bond yields, is the poster child of that worry; shares have cratered 30% in October. Nebius is heading for a 27% decline on the month.The Roundhill Memory ETF (DRAM), which includes AI memory plays Micron and Seagate Technology, is off 14% in October, while the iShares A.I. Innovation and Tech Active ETF (BAI) is down 10%.As traders prepare to wrap up the month, the Nasdaq Composite is heading for a 2.5% fall. The S&P 500 is on track for a 0.9% advance, while the Dow is on pace to rise 2.8%.
Amazon shares plunged more than 6% in extended trading Thursday after the company reported disappointing third-quarter results and ballooning negative free cash flow, reflecting its swelling AI investments. Here's how the company did, compared with estimates from analysts polled by LSEG: Earnings per share: $1.89 vs. $1.96 estimatedRevenue: $201.94 billion vs. $202.31 billion estimatedWall Street was also looking at other key revenue numbers:Amazon Web Services: $45.98 billion vs. $45.29 billion, according to StreetAccountAdvertising: $19.64 billion vs. $20.82 billion, according to StreetAccountThe company held steady on its capital expenditure forecast for 2026, saying it continues to expect spending will reach $220 billion this year. Tech's hyperscalers have been on a blistering spending run to meet the surge in demand for artificial intelligence infrastructure, but signs have begun to emerge that the capex splurge is moderating. Earlier this week, Microsoft lowered its capex forecast for the year to $170 billion from $175 billion, while Alphabet maintained its guidance at $195 billion to $205 billion. CEO Andy Jassy said on a conference call with investors that the company expects to "spend meaningfully more in 2027," and that Amazon has a "clear line of sight to attractive financial returns." The aggressive AI spending has eaten into Amazon's cash flow. The company posted negative free cash flow for the second consecutive quarter, recording an outflow of $42.4 billion for the trailing twelve month period, a swing of 386% year over year. In the prior quarter, the company reported negative free cash flow of $7.6 billion.The company said the shortfall was primarily due to an increase in purchases of property and equipment as a result of AI investments. Capital expenditures reached $77.3 billion during the quarter, ticking up from $35.1 billion one year earlier, and coming in well above analysts' estimated $58.7 billion, according to FactSet.Amazon said it expects fourth-quarter revenue to be in the range of $240 billion and $247 billion. The midpoint of its guidance was below Wall Street's expectations of $244.5 billion, according to LSEG. The fourth quarter is typically one of Amazon's biggest by revenue, due to the peak holiday shopping season. Sales in Amazon's cloud segment soared 39% year over year to $45.98 billion. That topped the consensus forecast for $45.29 billion and growth of 38%, according to StreetAccount. Jassy highlighted AWS acceleration in the earnings release, saying the company expects revenue growth rates of "over 40%" next year and in 2028. He added, "Our AWS investment in functionality and infrastructure is working."Jassy told investors on the call that he expects AWS will be a $1 trillion revenue business in the next five to 10 years. Amazon is facing pressure from Wall Street to show that its AI investments are lifting its revenues. The company pointed to growing traction for Bedrock, its AI model marketplace, along with its Trainium and Graviton AI chips. Jassy said Amazon's chips business "continues to grow triple-digit percentages year over year." He added that the company plans to make its Trainium chips available externally "late in 2027," assigning a timeframe to comments Jassy made earlier this year suggesting such a move.Third-quarter net income came in at $20.69 billion, or $1.89 per share, vs. $21.19 billion, or $1.95 per share, a year ago. The company said the decrease was driven primarily by a $9.5 billion pre-tax gain from its investment in the AI startup Anthropic, which was recorded in the 2025 third quarter. While the AWS segment continues to hum along, results weren't as bright in Amazon's retail business, which remains its largest unit by revenue, or its advertising segment. Online sales gained 2% year over year to $68.47 billion, which fell short of analysts' projected $69.68 billion, according to StreetAccount. It marked a slowdown from one year ago, when sales increased nearly 10%.Advertising revenue rose 11% year over year to $19.64 billion. That missed StreetAccount consensus estimates of $20.82 billion.Both units were impacted in part by Amazon's decision to pull its annual Prime Day sale forward in several countries from its typical third quarter timeframe into the second quarter so as to not conflict with the FIFA World Cup in July.Stock Chart IconStock chart iconAmazon stock chart.
Amazon reported its third-quarter earnings on Thursday and missed analysts’ expectations for both its total revenue and earnings per share. There were other murky numbers in the report that scared investors, too, including how much cash the company is burning through. But the report did have a big, bright spot: Its cloud unit, Amazon Web Services, beat revenue expectations and grew a whopping 39% compared to the year-ago quarter. Here are the numbers: Net sales increased 12% to $201.9 billion. But analysts were expecting $202.2 billion, according to FactSet. Fully diluted earnings per share were $1.89. Analysts were expecting $1.97. AWS sales reached $45.98 billion. Analysts were expecting $45.29 billion. In addition, Amazon said that AWS operating income grew to $17.8 billion, compared with $11.4 billion in the third quarter of 2025. But here’s the biggest ouch of the report: Free cash flow fell to a negative of $42.4 billion for the trailing 12 months — meaning the company was spending far more cash than it generated. This compares to positive free cash flow of $14.8 billion for the trailing 12 months ended September 30, 2025. The big spending was driven largely by Amazon’s investments in AI, the company said. With one quarter left in its fiscal year, Amazon reiterated that it still expects to end its 2026 with cash capital expenditures of about $220 billion. To put that in perspective, the company forecast fourth-quarter net sales of between $240 billion and $247 billion. The fourth quarter is typically a retailer’s biggest, thanks to the holiday shopping season. That means Amazon expects to spend almost as much on capital expenditures, largely for AI, over the full year as it expects to bring in from sales in the fourth quarter.
Rivian’s new R2 SUV was its best-selling vehicle in the third quarter, the company said Thursday, marking an auspicious start for the all-important EV. While Rivian didn’t say exactly how many R2s it sold over the three-month period, the company is confident enough in the pace of sales that it raised its delivery guidance for Wall Street. Rivian had previously expected to deliver between 65,000 and 70,000 EVs this year. On Thursday, it bumped up that floor to 67,000. It’s the second time Rivian has raised its delivery guidance this year, and the company is now on track to break a three-year stalemate during which it struggled to ship more than 50,000 vehicles annually. The company launched the R2 just in June, but it has made a major bet on scaling production of the lower-cost SUV at a historically quick rate. The early success of the R2 is also a sign that Rivian is finding new customers despite last year’s loss of the $7,500 federal EV tax credit. The death of that tax credit, along with changes made by the Trump administration to U.S. regulations that supported the growth of EVs, has caused many major automakers to shift their attention back to internal combustion and hybrid vehicles — leaving room in the market for pure-EV companies like Tesla and Rivian. “In a U.S. automotive marketplace starved for high-quality EV choice, R2 is an attractively priced option for everyday adventures that is resonating with a broad set of consumers and we expect will be a game changer for Rivian’s long-term growth and profitability,” the company said in a letter to shareholders released Thursday. Rivian revealed earlier this month that it built 19,751 vehicles in the third quarter and delivered 19,248. That was roughly 45% more than it delivered in the same period last year and a big jump from the 12,613 vehicles built and 12,194 delivered in the second quarter of 2026. The increase was driven in large part by the company launching a second production shift at its factory in Normal, Illinois. Rivian said Thursday that the R2 was responsible for “much of the gain quarter-over-quarter,” but didn’t go into specifics. The company generated $2.2 billion in revenue in the third quarter and suffered a net loss of $913 million — down from $1.2 billion in the same period last year. Rivian launched the R2 at a starting price of just under $60,000. A more affordable “premium” trim is supposed to be available later this year that starts at $53,990. Even cheaper versions are slated for 2027, which will be available for as low as $44,990.
DoorDash announced on Thursday that it has agreed to acquire rival Instacart for $17 billion. The acquisition would bring two major players in on-demand delivery under one roof, combining DoorDash’s dominance in restaurant takeout with Instacart’s reach in grocery delivery. DoorDash will buy Instacart for a total value of $65 for each Instacart share, the company said. The deal has been approved by the boards of both companies. “Over the last 14 years, Instacart has transformed online grocery by helping retailers of all sizes compete and grow, and joining forces with DoorDash allows us to do even more,” said Instacart CEO Chris Rogers in a press release. “When our purpose-built grocery technology and trusted retailer and brand relationships are combined with DoorDash’s global consumer reach and logistics capabilities, we can create new opportunities for our retail and brand partners, offer an even better experience for customers and shoppers, and deliver compelling value for our shareholders.” DoorDash plans to keep Instacart’s brand, enterprise technology products, and app. The company also notes that the two brands will continue to operate as separate and independent companies until the transaction closes. DoorDash says the deal could help more grocers build out their own online ordering platforms and advertising businesses using Instacart’s technology, while grocers could also use DoorDash’s fulfillment services to handle more online orders and offer lower-cost delivery. “Grocery is one of the hardest things to get right online,” DoorDash CEO Tony Xu said in the press release. “Someone has to pick a ripe avocado, find a substitute you’d choose yourself, and get the ice cream to you before it melts. Instacart has spent years learning to do that well for grocers, while DoorDash has built a delivery network and technology that local stores rely on every day. Together, we could give stores of every size more of those tools and unlock the potential of grocery delivery. It’s still early innings, but we know that when more people shop local, more dollars stay local,” said Xu. Instacart currently has a market cap of more than $13 billion, and its stock price is up 25% for the year to date. The company likely viewed this as an opportune time to sell at a desirable price, ahead of intensifying competition in the grocery delivery market. Bloomberg previously reported that Instacart was in talks with Uber, but that the company had walked away from discussions. The deal is expected to close toward the end of 2027 or in the first half of 2028.
On October 15–16, at the tail end of SEMICON West 2026, PDF Solutions held its 2026 CONNECT Conference at the Westin St. Francis in San Francisco. PDF Solutions Vitals Santa Clara–based PDF Solutions is a semiconductor data, analytics, and connectivity company, founded in 1991. Today, its business rests on three pillars, all feeding a single data layer: Exensio — the core analytics platform. It ingests 50+ data formats from FDC, test, assembly and packaging, harmonizes them into a semantic model (cloud or on-prem), and runs interactive analytics and ML on top. This is the recurring, scalable engine. Cimetrix + SecureWise — factory equipment connectivity and, since the March 2025 SecureWise acquisition, secure remote equipment connectivity for data acquisition and control. The connectivity layer is what makes PDFS a manufacturing data aggregator, not just an analytics vendor. DirectScan / Design-for-Inspection — on-chip electrical measurement hardware to catch defects during manufacturing. eBeam inspection to uncover buried systematic defects. This is the differentiated-data moat: PDFS generates proprietary characterization data at advanced nodes that optical inspection can’t see, then monetizes the analytics on it. The company reports about 350 customers spanning fabs, fabless, OSATs and equipment OEMs, i.e., it sits in the middle and broker data across the chain. It is publicly traded on the NASDAQ stock market under the symbol PDFS. The PDF Solutions CONNECT Conference CONNECT was built around the recent announcement of Exensio Aurora and the requirements for applying AI across manufacturing, inspection, test and the supply chain, supported by customer deployment discussions. Its intended audience included manufacturing engineers, data scientists, equipment companies and semiconductor executives. The agenda for Thursday morning, Oct 15, proposed two keynotes: Executive Keynote by John Kibarian, CEO, titled “Next Generation PDF Solutions Platform for the Global Semiconductor Industry.” Technical Keynote by Kimon Michaels, CTO, titled “PDF Solutions Platform – Architecture Overview and New Solutions Announcements.” And an Exensio Aurora panel at noon. Afternoon customer and solution talks: ST on advanced test solutions; Intel on reimagining manufacturing analytics for advanced packaging; Samsung on vector-scan eBeam inspection in advanced logic; Unimicron on advancing substrate manufacturing with FDC; PDF on its AI strategy and on transforming analytics with Exensio Aurora; POET on wafer-scale optics for the AI era; next-generation equipment integration and control; secureWISE updates. The evening dinner panel included the NYTimes, imec, Siemens EDA, Teradyne and PDF Solutions. The Agenda for Friday, Oct 16, included an industry keynote by Dr. Jon Koomey, “Plenty of room at the top (and also the bottom): What’s next for computing,” with a fireside chat with Kibarian. Then deep dives, clinics and demos on advanced test, scalable analytics, LLM-based solutions, equipment solutions, and Aurora development strategy and design. Customer speakers and what they were billed to cover Samsung: Dr. Hyung Keun Yoo, Sr. Director Metrology and Inspection, on adopting PDF’s DirectScan technology to ramp leading-edge chips. Intel: Syed Baquar, Principal Engineer and Director, Data & Analytics, on how Intel is transforming its manufacturing data infrastructure and moving to AI-driven analytics with PDF. STMicroelectronics: Massimo Longaretti, Software Automation Manager, on how to automate test execution, moving towards “dark factories.” Unimicron: Dr. Justin Chang, Senior Deputy Director of Smart Manufacturing, on FDC for substrates. POET: Raju Kankipati, Chief Revenue Officer, speaking on managing operational challenges in a fast-growing company. Dinner panel: Teradyne CEO Greg Smith, imec CEO Patrick Vandenameele and Siemens EDA EVP Ankur Gupta, joined by Kibarian. The panel moderator originally announced was Bloomberg’s Emily Chang who in the end was travelling in Europe that week with Jensen Huang. Tripp Mickle, NYTimes tech reporter, very successfully moderated the panel in her place. Main announcement: Exensio Aurora Unveiled September 9 as the new, highly scalable Exensio analytics architecture, with first public demos at CONNECT. PDF’s stated rationale is that 3D architectures, advanced packaging and heterogeneous integration blur front-end/back-end boundaries and cause a data explosion, and that in its experience under 5% of semiconductor manufacturing data gets analyzed with traditional tools. It was built because conventional BI tools hit in-memory processing limits and general-purpose cloud data platforms lack semiconductor relevance, at the urging of several customers. It is cloud-first but also supports on-premises deployment, for both online and offline applications. Said Akar, VP Enterprise Development, wrote a blog post on the four questions that led to Aurora, including how Exensio supports a range of AI approaches, LLMs among them. One trade outlet noted the caveat: the September announcement defined the architecture and direction rather than offering an independently benchmarked production result. Executive and Technical Keynotes CEO John Kibarian argued that AI could now transform semiconductor manufacturing itself. His executive keynote and co-founder Kimon Michaels’s technical presentation outlined how Exensio Aurora aims to connect manufacturing knowledge, scalable analytics and coordinated action across the supply chain. Kibarian framed the opportunity around efficiency. Demand for AI hardware is growing, but equipment availability, energy and manufacturing capacity constrain production. New fabs require expertise and successful process ramps as well as capital. Customers must therefore extract more useful output from existing resources, even as they invest in additional capacity. More products and more complex data Specialized processors are changing manufacturing requirements. Foundation-model companies and data-center operators are developing bespoke chips, creating more products and potentially smaller production runs. Some customers, Kibarian observed, effectively never leave new product introduction: the extensive measurements needed during development remain important throughout production. Assembly and test are also becoming more complicated. Fabless companies must manage information across multiple manufacturing partners and test stages. Earlier results can influence downstream screening, grading and routing. Advanced packaging, photonics and co-packaged optics add further complexity, making traceability and data integration increasingly important. The challenge extends beyond collecting measurements. An experienced engineer understands relationships among products, tools, chambers and process histories that a general-purpose AI model may miss. During yield excursions, relevant warnings often become obvious only afterward. PDF Solutions’ strategy is to combine manufacturing data with semiconductor-specific context so engineers can identify meaningful signals sooner. Reusable workflows connect AI with engineering practice Kibarian described a practical role for large language models: constructing analytical workflows that engineers can inspect, execute and reuse. A request can generate blocks for querying data, performing calculations, producing plots and annotating results. Python provides the underlying language, while visual workflows expose the analytical sequence. Once validated, a workflow can be stored and reused, reducing repeated model interactions and associated computing costs. It also preserves engineering knowledge. Manufacturers need consistent, auditable decisions, particularly when screening products with long service lives. Retaining the procedure makes it possible to investigate how a result was reached rather than relying solely on an AI-generated answer. Michaels explained the architecture supporting this approach. Aurora shifts heavy analysis from desktop clients to distributed server-side computation. Its FUSE computing stack addresses both parallel processing and parallel data access, accommodating very wide development datasets and extensive high-volume manufacturing histories. Caching and reusable intermediate results further improve efficiency. Michaels reported approximately 25-fold performance improvements in internal testing at comparable hardware spending. These are company-reported figures whose applicability depends on the workload. The broader objective is to make analyses practical that previously required excessive time or computing capacity. Taking models into production Exensio StudioAI provides the framework for model development, deployment, monitoring and governance. Tracking model versions and operating locations matters when an automated screening decision must be reconstructed months or years later. Deployment depends on the application. Time-critical algorithms may run directly on testers or process tools. Factory-level analysis can combine information from multiple machines, while less urgent work can run centrally or in the cloud. Michaels emphasized an open architecture that supports customers’ own models, libraries, data stores and agents. Exensio Aurora also expands the representation of manufacturing history beyond flat tables. Repeated measurements, retests, derived values and relationships among wafers, dies and assembly stages must remain connected. Integrating design and product-lifecycle information supplies additional context for analysis. A detailed demonstration was provided of how LLMs can be used to “dialog with the data” and create successive analytical workflows that can be saved and reused later for similar purposes ensuring repeatability, predictability and efficiency while encoding the specific engineering knowledge of the users. Kibarian linked these capabilities with secureWISE, which supports controlled, encrypted collaborat
Broadcom Announces Continued Infrastructure Agreement with FedEx to Support Global Operations
1 d ago
Oded Nagel is CEO of CTERA Networks, since 2023. He joined the company in 2008 and occupied since then several positions: chief strategy officer, VP solutions and director, customer care. Previously he collaborated with Checkpoint, SofaWare Technologies and 013 Netvision. He holds a BA in Business and Economics. StorageNewsletter: Oded, following our recent meeting during the 70th edition of The IT Press Tour in California, I would like to take this opportunity to ask you a few questions about CTERA, the business strategy, and solutions, particularly in today’s challenging market environment, which, despite its complexities, also presents significant opportunities for growth and innovation. CTERA is a private company with strong enterprise recognition and adoption. Could you share a few key figures that show the company’s growth? Oded Nagel: We don’t publish financials, but a few numbers tell the story. We’ve grown at a CAGR of more than 25%, and CTERA now manages hundreds of petabytes of customer data. Internally, we’re at about 200 people. Over the past two years, marketing has more than doubled and sales is on track to do the same, while R&D has stayed roughly the same size because AI has made our engineers far more productive. That ratio reflects where we are. The platform is built, and we’re scaling how we take it to market. The AI wave is reshaping the market quickly, and CTERA has had to adapt its strategy. How would you describe your new mission? For 18 years, our job was to unify and protect file data across distributed environments. That foundation hasn’t changed, but the destination has. Our mission now is to make enterprise file data usable by AI, safely, and with a measurable return. Most organizations are buying GPUs, models, and copilots, yet few can prove value. We see three reasons: the data isn’t AI-ready, security hasn’t kept up, and not enough people know how to build agents. Our principle is simple: move the AI to the data, not the data to the AI. Prices for memory and storage components are under strong pressure. Is that affecting your business, and if so, how are you responding? It affects everyone in the market. We’re software-defined and price mainly on capacity, so we don’t carry hardware inventory risk, but our customers feel it directly. Our own research across 16PB of production data found that only 4.4% of capacity is actively used and 88% of files haven’t been touched in over a year. Paying flash prices for that data makes even less sense when component costs rise. We help customers respond in three ways: edge caching keeps only active data on local flash, archiving moves cold data to lower-cost object storage, and customers choose their own object storage, in the cloud or on premises, so they aren’t tied to one vendor’s hardware pricing. What does that mean for your partners? This is a major opportunity for partners to bring more value to their customers. With the CTERA platform, partners can go beyond storage and help customers get more out of their own data, so teams can find, use, and activate it more effectively and become more productive. Partners can also help customers cut costs, which matters even more as component prices rise. CTERA analyzes data activity to separate the data people use from the data they don’t. Active data stays fast and accessible, and unused data is archived to lower-cost storage. Customers spend less on storage, and partners become trusted advisors, which creates more services and recurring revenue. Your data platform is already very rich. What are the most important features and services you’ve added recently? Five additions stand out. First, CTERA Fusion Direct unifies file and object in one namespace, so existing S3 buckets attach to the Global File System without migration, and AI clusters can read the same data over S3 and S3 over RDMA. Second, CTERA InsightAI gives administrators plain-language visibility into what they store, what’s growing, and what’s risky. Third, MCP connectivity links file data to Microsoft 365 Copilot, Claude, Gemini, and Cursor, while our n8n integration brings it into agentic workflows. Fourth, AI-assisted data archiving identifies cold data and moves it under policy. And most recently, Forward Deployed Engineering embeds our engineers in customer teams until an AI use case is in production and the customer can run it on their own. Underpinning all of this is cyber protection, which matters even more as AI gains access to file data. CTERA Ransom Protect detects malicious behavior and blocks attacks at the edge, and we’re extending that protection from encryption to data exfiltration. AI is changing what enterprises expect from file services. What does that mean for CTERA and for your customers’ needs? In other words, modern challenges call for modern solutions: where does CTERA fit? File services used to be judged on access, performance, and protection. Those still matter, but there’s now a new consumer of file data: the AI agent. The common approach is to copy data into a separate AI pipeline, which adds cost and risk, creates another copy to govern, and strips away the context and permissions that make the data trustworthy. We take the opposite approach. We help customers bring AI to their data, not their data to AI. The same Global File System that serves users over SMB and NFS serves AI over S3 and MCP, with permissions enforced in place. The data stays where it is, governed as it always was, and customers use the models of their choice. Could you give examples of the AI services you deliver on the CTERA platform, along with some customer references? CTERA InsightAI analyzes metadata only, such as names, paths, sizes, and audit logs, so there are no file reads and no egress costs. A manufacturer with more than a petabyte across four regions and 33 CTERA Edge Filers uses it to understand what is growing, stale, valuable, or risky. CTERA Content Services goes deeper. It converts documents, scans, and audio into AI-readable text, then runs permission-aware search, RAG, classification, and extraction. A U.S. mortgage company uses it through MCP in Microsoft Copilot Studio to search loan files, tag personal information, and prepare training data for default-prediction models. A medical-legal firm uses it to turn scanned and handwritten records into case narratives for legal review. You’ve also introduced a data archiving service. What does it mean for enterprises, and what impact does it have on their storage TCO? Archiving has traditionally meant removing data from the working estate. Ours keeps it in the same namespace, under the same access model, and still reachable by AI. The cost reduction comes from two levers: the data moves to a lower-cost license tier and to a lower-cost storage class, in the cloud or on premises. AI identifies the candidates, and every move is logged with a reason, with retention controls for compliance. The TCO impact depends on how much cold data a customer has. Given that most estates are overwhelmingly inactive, the opportunity is significant, and we’d rather show customers their own numbers than quote a generic percentage. Competition in this space is active. How do you differentiate from the usual players addressing the same needs? You’ll find point products for each piece of this: file services, analytics, security, AI pipelines. What customers struggle with is stitching them together. We deliver the file system, the edge, protection, insights, and the RAG and embedding layer on one governed platform, plus the engineers to connect it to their workflows. Two principles matter. First, data stays in standard formats, with no proprietary encapsulation, so customers keep control and sovereignty. Second, on security, we play offense: CTERA Ransom Protect detects malicious behavior and blocks it at the edge rather than relying only on restore after the fact. How do you assess the growing threats of ransomware and data theft, and what measures should organizations prioritize to strengthen their resilience Backup is defense, and it’s still necessary. But restoring a billion encrypted files takes days or weeks, so we put the emphasis on stopping the attack early or altogether. CTERA Ransom Protect watches for malicious behavior and blocks it at the edge. In third-party lab testing by Synergy7, it detected all eight ransomware families tested, with a median time to block of 24.5 seconds, and kept every attack under 10% file encryption. The bigger shift now is exfiltration. If you can’t show what was stolen, you have to assume everything was, notify every customer, and absorb the fines. Knowing exactly what was touched is becoming as important as recovery. That’s why we brought Tal Sarfaty on as SVP of Cybersecurity, and it’s a major focus going into 2027. And before we wrap up, what is your perspective on data sovereignty, and how do you see its importance evolving? Sovereignty starts with control over where data lives and how it’s stored. CTERA runs fully on premises, in a private cloud, or in the customer’s own cloud tenancy, and data stays in standard S3 buckets with no proprietary encapsulation, so customers are never locked into us or a single provider. The same applies to AI. Customers bring their own model, including private LLMs running on their own infrastructure, so sensitive files never have to leave their environment to be useful. For many of our customers in government, healthcare, and financial services, that’s the only way AI gets approved at all. As 2026 draws to a close, what have been the key milestones for CTERA this year, and what are your priorities for 2027? 2026 was our busiest year. We launched CTERA Fusion Direct in March, CTERA InsightAI in May, n8n integration in June, Microsoft 365 Copilot support in July, and AI-assisted archiving in September. We also published research on how little enterprise storage is actually used, introduced Forward Deployed Engineering in Oct
Screens display the Fed rate announcement as traders work on the floor of the NYSE in New York City, US, Oct. 28, 2026. Jeenah Moon | ReutersU.S. equity futures dipped late Wednesday as investors digest the latest round of hyperscaler earnings. Nasdaq-100 futures lost 1.2%. S&P 500 futures declined 0.5%.In Asia, Japan's Nikkei 225 added over 0.58% while the Topix was flat. The Kospi rose 0.11% at open, while the small-cap Kosdaq fell 0.97%. Australia's benchmark S&P/ASX 200 was 0.51% higher.In after-hours trading, Microsoft advanced 3% after growth in the software giant's Azure cloud infrastructure business helped it post beats on the top and bottom lines in the fiscal first quarter, per LSEG. That contrasts sharply with Alphabet's 9% slide after adjusted earnings per share and revenue missed expectations in the third quarter, and the company posted negative free cash flow of $9.07 billion. The company kept its full-year capital expenditures forecast at $195 billion to $205 billion. The results come at a time when the artificial intelligence trade is in focus. AI stocks, such as the iShares Semiconductor ETF (SOXX) and the Roundhill Memory ETF (DRAM), have been through a rough patch for most of October, only to perk up in recent days. Chips helped lift the broader market to a winning regular session on Wednesday, with the S&P 500 rising for a third straight day as Micron Technology jumped almost 6% and Broadcom added more than 2%. The tech-heavy Nasdaq Composite also eked out a gain, advancing 0.3%.Earlier Wednesday, the Federal Reserve decided to hold steady on rates, leaving the central bank's benchmark interest rate in a range between 3.75% and 4%. The decision to hold was widely expected as the Federal Open Market Committee issued a quarter-percentage-point hike in September. This time, central bank policymakers pointed out that inflation "remains elevated," but also said that it was awaiting the "accumulation of more information" before it takes further steps.The Fed's decision "can be interpreted as mildly dovish at the margin, but we think it is better viewed as an expression of patience rather than a shift away from a tightening bias," said Daniel Siluk, head of global short duration and liquidity at Janus Henderson Investors. "The door remains open to further tightening, but the Committee appears willing to wait for incoming data to validate the need for additional action."Traders will get further insight into the economy on Thursday, when the Personal Consumption Expenditures price index report for September comes out. The PCE is a preferred inflation gauge for the Federal Reserve. The first reading of third-quarter GDP and weekly jobless claims are also due.On the earnings front, all eyes will be on quarterly reports from Caterpillar and Eli Lilly in the morning, followed by Amazon and Intel after the market close.