Intel is grappling with an unexpected market shift as customers eschew its new AI-focused processors for cheaper previous-generation chips. The company revealed during its recent earnings call that demand for older Raptor Lake processors has surged while its newer, more expensive Lunar Lake and Meteor Lake AI PC chips struggle to gain traction.

This surprising trend, first reported by Tom's Hardware, has created a production capacity shortage for Intel's 'Intel 7' process node that will "persist for the foreseeable future," despite the fact that current-generation chips utilize TSMC's newer nodes. "Customers are demanding system price points that consumers really want," explained Intel executive Michelle Johnston Holthaus, noting that economic concerns and tariffs have affected inventory decisions.

Intel plans to cut more than 20% of its workforce this week, marking the first major restructuring under new CEO Lip-Bu Tan, according to Bloomberg. The cuts aim to eliminate bureaucracy and restore an engineering-centric culture at the struggling chipmaker. This follows last year's reduction of approximately 15,000 positions, with Intel's headcount already down to 108,900 employees from 124,800 a year earlier.

The Santa Clara-based company has suffered three consecutive years of declining sales while losing technological ground to competitors, particularly Nvidia in the AI computing sector. Tan, who took over last month, has already begun divesting non-core assets, recently selling a 51% stake in Intel's programmable chips unit Altera to Silver Lake.

The Trump administration is "taking measures to restrict the sale of AI chips by Nvidia, Advanced Micro Devices and Intel," especially in China, reports the New York Times. But that's triggered a series of dominoes. "In the two days after the limits became public, shares of Nvidia, the world's leading AI chipmaker, fell 8.4%. AMD's shares dropped 7.4%, and Intel's were down 6.8%." (AMD expects up to $800 million in charges after the move, according to CNBC, while NVIDIA said it would take a quarterly charge of about $5.5 billion.)

The Times notes hopeful remarks Thursday from Jensen Huang, CEO of Nvidia, during a meeting with the China Council for the Promotion of International Trade. "We're going to continue to make significant effort to optimize our products that are compliant within the regulations and continue to serve China's market." But America's chipmakers also have a greater fear, according to the article: "that their retreat could turn the Chinese tech giant Huawei into a global chip-making powerhouse." "For the U.S. semiconductor industry, China is gone," said Handel Jones, a semiconductor consultant at International Business Strategies, which advises electronics companies. He projects that Chinese companies will have a majority share of chips in every major category in China by 2030... Huang's message spoke to one of his biggest fears. For years, he has worried that Huawei, China's telecommunications giant, will become a major competitor in AI. He has warned U.S. officials that blocking U.S. companies from competing in China would accelerate Huawei's rise, said three people familiar with those meetings who spoke on the condition of anonymity.

If Huawei gains ground, Huang and others at Nvidia have painted a dark picture of a future in which China will use the company's chips to build AI data centers across the world for the Belt and Road Initiative, a strategic effort to increase Beijing's influence by paying for infrastructure projects around the world, a person familiar with the company's thinking said...

Nvidia's previous generation of chips perform about 40% better than Huawei's best product, said Gregory C. Allen, who has written about Huawei in his role as director of the Wadhwani AI Center at the Center for Strategic and International Studies. But that gap could dwindle if Huawei scoops up the business of its American rivals, Allen said. Nvidia was expected to make more than $16 billion in sales this year from the H20 in China before the restriction. Huawei could use that money to hire more experienced engineers and make higher-quality chips. Allen said the U.S. government's restrictions also could help Huawei bring on customers like DeepSeek, a leading Chinese AI startup. Working with those companies could help Huawei improve the software it develops to control its chips. Those kinds of tools have been one of Nvidia's strengths over the years.
TechRepublic identifies this key quote from an earlier article: "This kills NVIDIA's access to a key market, and they will lose traction in the country," Patrick Moorhead, a tech analyst with Moor Insights & Strategy, told The New York Times. He added that Chinese companies will buy from local rival Huawei instead.

Longtime Slashdot reader hackingbear shares a report from Interesting Engineering: A research team at Fudan University in Shanghai, China has built the fastest semiconductor storage device ever reported, a nonvolatile flash memory dubbed "PoX" that programs a single bit in 400 picoseconds (0.0000000004 s) -- roughly 25 billion operations per second. Conventional static and dynamic RAM (SRAM, DRAM) write data in 1-10 nanoseconds but lose everything when power is cut while current flash chips typically need micro to milliseconds per write -- far too slow for modern AI accelerators that shunt terabytes of parameters in real time.

The Fudan group, led by Prof. Zhou Peng at the State Key Laboratory of Integrated Chips and Systems, re-engineered flash physics by replacing silicon channels with two dimensional Dirac graphene and exploiting its ballistic charge transport. Combining ultralow energy with picosecond write speeds could eliminate separate highspeed SRAM caches and remove the longstanding memory bottleneck in AI inference and training hardware, where data shuttling, not arithmetic, now dominates power budgets. The team [which is now scaling the cell architecture and pursuing arraylevel demonstrations] did not disclose endurance figures or fabrication yield, but the graphene channel suggests compatibility with existing 2Dmaterial processes that global fabs are already exploring. The result is published in the journal Nature.

Microsoft has introduced BitNet b1.58 2B4T, the largest-scale 1-bit AI model to date with 2 billion parameters and the ability to run efficiently on CPUs. It's openly available under an MIT license. TechCrunch reports: The Microsoft researchers say that BitNet b1.58 2B4T is the first bitnet with 2 billion parameters, "parameters" being largely synonymous with "weights." Trained on a dataset of 4 trillion tokens -- equivalent to about 33 million books, by one estimate -- BitNet b1.58 2B4T outperforms traditional models of similar sizes, the researchers claim.

BitNet b1.58 2B4T doesn't sweep the floor with rival 2 billion-parameter models, to be clear, but it seemingly holds its own. According to the researchers' testing, the model surpasses Meta's Llama 3.2 1B, Google's Gemma 3 1B, and Alibaba's Qwen 2.5 1.5B on benchmarks including GSM8K (a collection of grade-school-level math problems) and PIQA (which tests physical commonsense reasoning skills). Perhaps more impressively, BitNet b1.58 2B4T is speedier than other models of its size -- in some cases, twice the speed -- while using a fraction of the memory.

There is a catch, however. Achieving that performance requires using Microsoft's custom framework, bitnet.cpp, which only works with certain hardware at the moment. Absent from the list of supported chips are GPUs, which dominate the AI infrastructure landscape.

Intel will sell a 51% stake in its Altera programmable chip unit to private equity firm Silver Lake for $4.46 billion, aiming to cut costs, raise cash, and streamline the company's focus as it shifts toward becoming a contract chip manufacturer. CNBC reports: The deal, announced on Monday, values Altera at $8.75 billion, a sharp decline from the $17 billion Intel paid in 2015. [...] Since last year, Intel has taken steps to spin Altera out as a separate unit and said it planned to sell a portion of its stake. "Today's announcement reflects our commitment to sharpening our focus, lowering our expense structure and strengthening our balance sheet," [CEO Lip-Bu Tan], who took the helm after former top boss Pat Gelsinger's ouster, said.

Altera makes programmable chips that can be used for various purposes from telecom equipment to military. Reuters had first reported in November that Silver Lake was among potential suitors competing for a minority stake in Altera. The deal is expected to close in the second half of 2025, after which Intel expects to deconsolidate Altera's financial results from Intel's financial statements, the company said.

Nvidia has announced plans to manufacture AI supercomputers entirely within the United States, commissioning over 1 million square feet of manufacturing space across Arizona and Texas. Production of Blackwell chips has begun at TSMC's Phoenix facilities, while supercomputer assembly will occur at new Foxconn and Wistron plants in Houston and Dallas respectively.

"The engines of the world's AI infrastructure are being built in the United States for the first time," said Jensen Huang, Nvidia's founder and CEO. "Adding American manufacturing helps us better meet the incredible and growing demand for AI chips and supercomputers, strengthens our supply chain and boosts our resiliency."

The company will deploy its own AI, robotics, and digital twin technologies in these facilities, using Nvidia Omniverse to create digital twins of factories and Isaac GR00T to build manufacturing automation robots. Nvidia projects an ambitious $500 billion in domestic AI infrastructure production over the next four years, with manufacturing expected to create hundreds of thousands of jobs.

Late Friday news broke that U.S. President Trump's new tariffs included exemptions for smartphones, computer monitors, semiconductors, and other electronics. But Sunday morning America's commerce secretary insisted "a special-focus type of tariff" was coming for those products, reports ABC News. President Trump "is saying they're exempt from the reciprocal tariffs," the commerce secretary told an interviewer, "but they're included in the semiconductor tariffs, which are coming in probably a month or two.... This is not like a permanent sort of exemption."

The Wall Street Journal notes that Sunday the president himself posted on social media that "NOBODY is getting 'off the hook' for the unfair Trade Balances, and Non Monetary Tariff Barriers... There was no Tariff 'exception' announced on Friday. These products are subject to the existing 20% Fentanyl Tariffs, and they are just moving to a different Tariff 'bucket.'"

"The administration is expected to take the first step toward enacting the new tariffs as soon as next week," reports the New York Times, "opening an investigation to determine the effects of semiconductor imports on national security."

More from ABC News: Commerce Secretary Howard Lutnick said Sunday that the administration's decision Friday night to exempt a range of electronic devices from tariffs implemented earlier this month was only a temporary reprieve.. Lutnick said on "This Week" that the White House will implement "a tariff model in order to encourage" the semiconductor industry, as well as the pharmaceutical industry, to move its business to the United States. "We can't be beholden and rely upon foreign countries for fundamental things that we need," he said.... "These are things that are national security that we need to be made in America."

Razer's upcoming Blade 16 and other laptops are no longer available for preorder or purchase on its US site. From a report: The configurator for preordering its new Blade 16 laptop was available as recently as April 1st, according to the Internet Archive -- one day before the Trump administration announced sweeping US tariffs on China, Taiwan, and others that make laptop components.

When asked recently if tariffs might affect Razer's prices or availability, its Public Relations Manager, Andy Johnston, told The Verge, "We do not have a comment at this stage regarding tariffs." Razer may not be openly talking about the impact of tariffs, but Framework halted sales of its entry-level Laptop 13 in the US on April 7th, and Micron reportedly confirmed surcharges for its memory chips will apply once the tariffs take effect after midnight tonight.

Software engineer and longtime Slashdot reader, Dmitry Grinberg (dmitrygr), shares a recent project they've been working on: "an interactive-speed Linux on a tiny board you can easily build with only 3 8-pin chips": There was a time when one could order a kit and assemble a computer at home. It would do just about what a contemporary store-bought computer could do. That time is long gone. Modern computers are made of hundreds of huge complex chips with no public datasheets and many hundreds of watts of power supplied to them over complex power delivery topologies. It does not help that modern operating systems require gigabytes of RAM, terabytes of storage, and always-on internet connectivity to properly spy on you. But what if one tried to fit a modern computer into a kit that could be easily assembled at home? What if the kit only had three chips, each with only 8 pins? Can it be done? Yes. The system runs a custom MIPS emulator written in ARMv6 assembly and includes a custom bootloader that supports firmware updates via FAT16-formatted SD cards. Clever pin-sharing hacks allow all components (RAM, SD, serial I/O) to work despite the 6 usable I/O pins. Overclocked to up to 150MHz, the board boots into a full Linux shell in about a minute and performs at ~1.65MHz MIPS-equivalent speed.

It's not fast, writes Dmitry, but it's fully functional -- you can edit files, compile code, and even install Debian packages. A kit may be made available if a partner is found.

Intel and Taiwan Semiconductor Manufacturing Co. have reached a preliminary agreement to form a joint venture operating Intel's chipmaking facilities, with TSMC taking a 20% stake, The Information reports [non-paywalled source]. Intel and other U.S. semiconductor companies would hold the majority of shares in the proposed venture. Instead of capital investment, TSMC has discussed sharing chipmaking methods and training Intel personnel.

The talks face internal opposition from some Intel executives concerned about widespread layoffs and the abandonment of Intel's own technology, according to the report. The deal could help TSMC neutralize a struggling competitor while potentially giving Taiwan more leverage with the U.S. administration, which recently imposed tariffs on Taiwanese goods excluding chips.

Tech manufacturers continue misleading consumers with impressive-sounding but less useful specs like milliamp-hours and megahertz, while hiding the one measurement that matters most: watts. The Verge argues that the watt provides the clearest picture of a device's true capabilities by showing how much power courses through chips and how quickly batteries drain. With elementary math, consumers could easily calculate battery life by dividing watt-hours by power consumption. The Verge: The Steam Deck gaming handheld is my go-to example of how handy watts can be. With a 15-watt maximum processor wattage and up to 9 watts of overhead for other components, a strenuous game drains its 49Wh battery in roughly two hours flat. My eight-year-old can do that math: 15 plus 9 is 24, and 24 times 2 is 48. You can fit two hour-long 24-watt sessions into 48Wh, and because you have 49Wh, you're almost sure to get it.

With the least strenuous games, I'll sometimes see my Steam Deck draining the battery at a speed of just 6 watts -- which means I can get eight hours of gameplay because 6 watts times 8 hours is 48Wh, with 1Wh remaining in the 49Wh battery. Unlike megahertz, wattage also indicates sustained performance capability, revealing whether a processor can maintain high speeds or will throttle due to thermal constraints. Watts is also already familiar to consumers through light bulbs and power bills, but manufacturers persist with less transparent metrics that make direct comparisons difficult.

Intel Chief Executive Officer Lip-Bu Tan said the chipmaker will spin off assets that aren't central to its mission and create new products including custom semiconductors to try to better align itself with customers. From a report: Intel needs to replace the engineering talent it has lost, improve its balance sheet and better attune manufacturing processes to meet the needs of potential customers, Tan said. Speaking at his first public appearance as CEO, at the Intel Vision conference Monday in Las Vegas, Tan didn't specify what parts of Intel he believes are no longer central to its future.

"We have a lot of hard work ahead," Tan said, addressing the company's customers in the audience. "There are areas where we've fallen short of your expectations." The veteran semiconductor executive is trying to restore the fortunes of a company that dominated an industry for decades, but now finds itself chasing rivals in most of the areas that define success in the field. A key question confronting its leadership is whether a turnaround is best served by the company remaining whole or splitting up its key product and manufacturing operations. Tan gave no indication that he will seek to divest either part of Intel. Instead, he highlighted the problems he needs to fix to get both units performing more successfully. Intel's chips for data center and AI-related work in particular are not good enough, he said. "We fell behind on innovation," the CEO said. "We have been too slow to adapt and meet your needs."

Taiwanese authorities have accused 11 Chinese companies, including SMIC, of secretly setting up disguised entities in Taiwan to illegally recruit tech talent from firms like Intel and Microsoft. The Register reports: One of those companies is apparently called Yunhe Zhiwang (Shanghai) Technology Co., Ltd and develops high-end network chips. The Bureau claims its chips are used in China's "Data East, Compute West" strategy that, as we reported when it was announced in 2022, calls for five million racks full of kit to be moved from China's big cities in the east to new datacenters located near renewable energy sources in country's west. Datacenters in China's east will be used for latency-sensitive applications, while heavy lifting takes place in the west. Staff from Intel and Microsoft were apparently lured to work for Yunhe Zhiwang, which disguised its true ownership by working through a Singaporean company.

The Investigation Bureau also alleged that China's largest chipmaker, Semiconductor Manufacturing International Corporation (SMIC), used a Samoan company to establish a presence in Taiwan and then hired local talent. That's a concerning scenario as SMIC is on the USA's "entity list" of organizations felt to represent a national security risk. The US gets tetchy when its friends and allies work with companies on the entity list.

A third Chinese entity, Shenzhen Tongrui Microelectronics Technology, disguised itself so well Taiwan's Ministry of Industry and Information Technology lauded it as an important innovator and growth company. As a result of the Bureau's work, prosecutors' offices in seven Taiwanese cities are now looking into 11 Chinese companies thought to have hidden their ties to Beijing.

Micron will raise prices for DRAM and NAND flash memory chips through 2026 as AI and data center demand strains supply chains, the U.S. chipmaker confirmed Monday. The move follows a market rebound from previous oversupply, with memory prices steadily climbing as producers cut output while AI and high-performance computing workloads grow.

Rivals Samsung Electronics and SK Hynix are expected to implement similar increases. Micron cited "un-forecasted demand across various business segments" in communications to channel partners. The price hikes will impact sectors ranging from consumer electronics to enterprise data centers.

The U.S. has added 80 entities to its export blacklist to prevent China from acquiring advanced American chips for military development, including AI, quantum tech, and hypersonic weapons. The Verge reports: More than 50 of the new entities added to the list are based in China, with others located in Iran, Taiwan, Pakistan, South Africa, and the United Arab Emirates. BIS says the restrictions have been applied to entities that acted "contrary to US national security and foreign policy," and are intended to hinder China's ability to develop high-performance computing capabilities, quantum technologies, advanced artificial intelligence, and hypersonic weapons.

Six of the newly blacklisted entities are subsidiaries of Inspur Group -- China's leading cloud computing service provider and a major customer for US chip makers such as Nvidia, AMD, and Intel -- which BIS alleges had contributed to projects developing supercomputers for the Chinese military. The Beijing Academy of Artificial Intelligence is another addition to the list, which has criticized its inclusion. "American technology should never be used against the American people," said Jeffrey Kessler, Under Secretary of Commerce for Industry and Security. "BIS is sending a clear, resounding message that the Trump administration will work tirelessly to safeguard our national security by preventing U.S. technologies and goods from being misused for high performance computing, hypersonic missiles, military aircraft training, and UAVs that threaten our national security."

An anonymous reader quotes a report from Bloomberg: Jack Ma-backed Ant Group used Chinese-made semiconductors to develop techniques for training AI models that would cut costs by 20%, according to people familiar with the matter. Ant used domestic chips, including from affiliate Alibaba and Huawei, to train models using the so-called Mixture of Experts machine learning approach, the people said. It got results similar to those from Nvidia chips like the H800, they said, asking not to be named as the information isn't public. Hangzhou-based Ant is still using Nvidia for AI development but is now relying mostly on alternatives including from Advanced Micro Devices and Chinese chips for its latest models, one of the people said.

The models mark Ant's entry into a race between Chinese and US companies that's accelerated since DeepSeek demonstrated how capable models can be trained for far less than the billions invested by OpenAI and Alphabet Inc.'s Google. It underscores how Chinese companies are trying to use local alternatives to the most advanced Nvidia semiconductors. While not the most advanced, the H800 is a relatively powerful processor and currently barred by the US from China. The company published a research paper this month that claimed its models at times outperformed Meta Platforms Inc. in certain benchmarks, which Bloomberg News hasn't independently verified. But if they work as advertised, Ant's platforms could mark another step forward for Chinese artificial intelligence development by slashing the cost of inferencing or supporting AI services.

A United States District Court judge has approved a settlement between HP and customers who sued the company for firmware updates that prevented printers from working with non-HP ink cartridges.

The class-action lawsuit, filed in December 2020, alleged HP "wrongfully compels users" to buy only HP ink by issuing updates that block competitors' cartridges. Under the settlement, HP admits no wrongdoing and won't pay monetary damages to affected customers, though it will pay $5,000 each to the three plaintiffs and $725,000 in attorneys' fees.

HP has agreed to allow users of specific printer models impacted by the November 2020 update to decline firmware updates containing "Dynamic Security" features -- HP's term for technology that blocks cartridges using non-HP chips. The settlement applies only to 21 specific printer models, leaving numerous other HP printers subject to Dynamic Security restrictions. HP has previously paid millions in similar cases in Europe, Australia, and California related to printer bricking.

General Motors is partnering with Nvidia to enhance its self-driving and manufacturing capabilities by leveraging Nvidia's AI chips, software, and simulation tools. "GM says it will apply several of Nvidia's products to its business, such as the Omniverse 3D graphics platform which will run simulations on virtual assembly lines with an eye on reducing downtime and improving efficiency," reports The Verge. "The automaker also plans to equip its next-generation vehicles with Nvidia's 'AI brain' for advanced driver assistance and autonomous driving. And it will employ the chipmaker's AI training software to make its vehicle assembly line robots better at certain tasks, like precision welding and material handling." From the report: GM already uses Nvidia's GPUs to train its AI software for simulation and validation. Today's announcement was about expanding those use cases into improving its manufacturing operations and autonomous vehicles, GM CEO Mary Barra said in a statement. (Dave Richardson, GM's senior VP of Software and Services Engineering will be joining NVIDIA's Norm Marks for a fireside chat at the conference.) "AI not only optimizes manufacturing processes and accelerates virtual testing but also helps us build smarter vehicles while empowering our workforce to focus on craftsmanship," Barra said. "By merging technology with human ingenuity, we unlock new levels of innovation in vehicle manufacturing and beyond."

GM will adopt Nvidia's in-car software products to build next-gen vehicles with autonomous driving capabilities. That includes the company's Drive AGX system-on-a-chip (SoC), similar to Tesla's Full Self-Driving chip or Intel's Mobileye EyeQ. The SoC runs the "safety-certified" DriveOS operating system, built on the Blackwell GPU architecture, which is capable of delivering 1,000 trillion operations per second (TOPS) of high-performance compute, the company says. [...] In a briefing with reporters, Ali Kani, Nvidia's vice president and general manager of automotive, described the chipmaking company's automotive business as still in its "infancy," with the expectation that it will only bring in $5 billion this year. (Nvidia reported over $130 billion in revenue in 2024 for all its divisions.)

Nvidia's chips are in less than 1 percent of the billions of cars on the road today, he added. But the future looks promising. The company is also announcing deals with Tier 1 auto supplier Magna, which helped build Sony's Afeela concept, to use Drive AGX in the company's next-generation advanced driver assist software. "We believe automotive is a trillion dollar opportunity for Nvidia," Kani said.

Nvidia unveiled its next wave of AI processors at GTC on Tuesday, announcing Blackwell Ultra chips that will ship in the second half of 2025, followed by the Vera Rubin architecture in 2026. CEO Jensen Huang also revealed that its 2028 chips will be named after physicist Richard Feynman.

The Blackwell Ultra maintains the same 20 petaflops of AI performance as standard Blackwell chips but increases memory from 192GB to 288GB of HBM3e. Nvidia claims these chips can process 1,000 tokens per second -- ten times faster than its 2022 hardware -- enabling AI reasoning tasks like running DeepSeek-R1 models with 10-second response times versus 1.5 minutes on H100 chips.

Vera Rubin will deliver a substantial leap to 50 petaflops in 2026, featuring Nvidia's first custom Arm-based CPU design called Olympus. Nvidia is also changing how it counts GPUs -- Rubin itself contains two dies working as one chip. The annual release cadence represents a strategic shift for Nvidia, which previously introduced new architectures every two years before the AI boom transformed its business.