The DeepComputing RISC-V Mainboard that Framework announced last year for its 13-inch laptops is now available for $199. Liliputing reports: If you already have a Framework Laptop 13 with an Intel or AMD motherboard, the new board is a drop-in replacement. But if you don't have a Framework Laptop you can also use the mainboard as a standalone computer: Framework sells a $39 Cooler Master case that effectively turns its mainboards into mini desktop computers. The RISC-V Mainboard comes from a partnership between Framework and DeepComputing, the Chinese company behind the DC-ROMA laptops, which were some of the first notebook computers to ship with RISC-V processors.

The board features a StarFive JH7110 processor, which is a 1.5 GHz quad-core chip featuring SiFive U74 RISC-V CPU cores and Imagination BXE-4-32 graphics, 8GB of onboard RAM, and a a 64GB SD card for storage (there's also support for an optional eMMC module, but you'll need to bring your own). Since the board is designed to fit in existing laptop frames, it's the same size and shape as AMD or Intel models and has four USB ports in the same locations. But these ports are a little less versatile than the ones you might find on other Framework Laptop 13 Mainboards [...]. There's also a 3.5mm audio jack. You can check out the new board via the Framework Marketplace.

Further reading: Late last year, Framework CEO Nirav Patel delivered one of the best live demos we've ever seen at a tech conference -- modifying a Framework Laptop from x86 to RISC-V live on stage.

On January 24th Brooklyn blogger Jeffrey Emanuel made the case for shorting NVIDIA, remembers MarketWatch, "due to a number of shifting tides in the AI world, including the emergence of a China-based company called DeepSeek."

He published his 12,000-word post "on his personal blog and then shared it with the Value Investors Club website and across Reddit, X and other platforms." The next day he saw 35 people read his post. "But then the post started to go viral..." Well-known venture capitalist Chamath Palihapitiya shared Emanuel's post on Nvidia's short case with his 1.8 million X followers. Successful early stage investor Naval Ravikant shared the post with his 2.6 million followers... Morgan Brown, a vice president of product and growth at Dropbox, pointed to it in a thread that was viewed over 13 million times. Emanuel's own X post got nearly half a million views. He also quickly gained about 13,000 followers on the platform, going from about 2,000 to more than 15,000 followers...

[Emanuel] pointed to the fact that so many people in San Jose were reading his blog post. He theorized that many of them were Nvidia employees with thousands — or even millions — of dollars worth of Nvidia stock tied up in employee stock options. With that much money in a single asset, Emanuel speculated that many were already debating whether to hold the stock or sell it to lock in profits. He believes his blog post helped convince some of them to sell. "A lot of the sell pressure you saw on Monday morning wasn't necessarily what you might think. I believe a fair amount of that was from shares that had never been active because they had been sitting in workplace.schwab.com accounts..."
Emanuel stresses he's "the most bullish on AI," with MarketWatch emphasizing that "while the points Emanuel laid out in his blog post might be bearish for Nvidia, he still thinks they paint a positive future for AI." Nevertheless, Monday NVIDIA's market capitalization dropped $600 billion, which MarketWatch calls "the largest single-day market-cap drop to date for any company." What countless Wall Street firms and investment analysts had seemingly missed was being pointed out by some guy in his apartment.... Matt Levine, the prominent Bloomberg News financial columnist, noted the online chatter that claimed Emanuel's post "was an important catalyst" for the stock-market selloff and said it was a "candidate for the most impactful short research report ever." Emanuel spent the rest of the week booked solid as hedge funds paid him $1,000 per hour to speak on the phone and give his take on Nvidia and AI...

Emanuel wrote that the industry may be running low on quality data to train that AI — that is, a potential "data wall" is looming that could slow down AI scaling and reduce some of that need for training resources... Some of these companies, like Alphabet, have also been investing in building out their own semiconductor chips. For a while, Nvidia's hardware has been the best for training AI, but that might not be the case forever as more companies, such as Cerebras, build better hardware. And other GPU makers like AMD are updating their drivers software to be more competitive with Nvidia... Add all these things together — unsustainable spending and data-center building, less training data to work with, better competing hardware and more efficient AI — and you get a future where it's harder to imagine Nvidia's customers spending as much as they currently are on Nvidia hardware... "If you know that a company will only earn supersized returns for a couple years, you don't apply a multiple. You certainly don't put a 30-times multiple," Emanuel told MarketWatch.
The article notes that DeepSeek "is open-source and has been publishing technical papers out in the open for the past few months... The $5.6 million training-cost statistic that many investors cited for sparking the DeepSeek market panic was actually revealed in the V3 technical paper published on Dec. 26."

"Nothing horrible or unexpected happened last week," Linux Torvalds posted tonight on the Linux kernel mailing list, "so I've tagged and pushed out the final 6.13 release."

Phoronix says the release has "plenty of fine features": Linux 6.13 comes with the introduction of the AMD 3D V-Cache Optimizer driver for benefiting multi-CCD Ryzen X3D processors. The new AMD EPYC 9005 "Turin" server processors will now default to AMD P-State rather than ACPI CPUFreq for better power efficiency....

Linux 6.13 also brings more Rust programming language infrastructure and more.
Phoronix notes that Linux 6.13 also brings "the start of Intel Xe3 graphics bring-up, support for many older (pre-M1) Apple devices like numerous iPads and iPhones, NVMe 2.1 specification support, and AutoFDO and Propeller optimization support when compiling the Linux kernel with the LLVM Clang compiler."

And some lucky Linux kernel developers will also be getting a guitar pedal soldered by Linus Torvalds himself, thanks to a generous offer he announced a week ago: For _me_ a traditional holiday activity tends to be a LEGO build or two, since that's often part of the presents... But in addition to the LEGO builds, this year I also ended up doing a number of guitar pedal kit builds ("LEGO for grown-ups with a soldering iron"). Not because I play guitar, but because I enjoy the tinkering, and the guitar pedals actually do something and are the right kind of "not very complex, but not some 5-minute 555 LED blinking thing"...

[S]ince I don't actually have any _use_ for the resulting pedals (I've already foisted off a few only unsuspecting victims^Hfriends), I decided that I'm going to see if some hapless kernel developer would want one.... as an admittedly pretty weak excuse to keep buying and building kits...
"It may be worth noting that while I've had good success so far, I'm a software person with a soldering iron. You have been warned... [Y]ou should set your expectations along the lines of 'quality kit built by a SW person who doesn't know one end of a guitar from the other.'"

"I know, I know: 'Year of the Linux desktop ... yadda, yadda'," writes Steven Vaughan-Nichols, a ZDNet senior contributing editor. "You've heard it all before. But now there's a Linux-powered PC that many people will want..."

He's talking about Nvidia's newly-announced Project Digits, describing it as "a desktop with AI supercomputer power that runs DGX OS, a customized Ubuntu Linux 22.04 distro." Powered by MediaTek and Nvidia's Grace Blackwell Superchip, Project DIGITS is a $3,000 personal AI that combines Nvidia's Blackwell GPU with a 20-core Grace CPU built on the Arm architecture... At CES, Nvidia CEO Jensen Huang confirmed plans to make this technology available to everyone, not just AI developers. "We're going to make this a mainstream product," Huang said. His statement suggests that Nvidia and MediaTek are positioning themselves to challenge established players — including Intel and AMD — in the desktop CPU market. This move to the desktop and perhaps even laptops has been coming for a while. As early as 2023, Nvidia was hinting that a consumer desktop chip would be in its future... [W]hy not use native Linux as the primary operating system on this new chip family?

Linux, after all, already runs on the Grace Blackwell Superchip. Windows doesn't. It's that simple. Nowadays, Linux runs well with Nvidia chips. Recent benchmarks show that open-source Linux graphic drivers work with Nvidia GPUs as well as its proprietary drivers. Even Linus Torvalds thinks Nvidia has gotten its open-source and Linux act together. In August 2023, Torvalds said, "Nvidia got much more involved in the kernel. Nvidia went from being on my list of companies who are not good to my list of companies who are doing really good work." Canonical, Ubuntu Linux's parent company, has long worked closely with Nvidia. Ubuntu already provides Blackwell drivers.
The article strays into speculation, when it adds "maybe you wouldn't pay three grand for a Project DIGITS PC. But what about a $1,000 Blackwell PC from Acer, Asus, or Lenovo? All three of these companies are already selling MediaTek-powered Chromebooks...."

"The first consumer products featuring this technology are expected to hit the market later this year. I'm looking forward to running Linux on it. Come on in! The operating system's fine."

The world's most powerful supercomputer, capable of 2.79 quintillion calculations per second, has been unveiled at Lawrence Livermore National Laboratory in California, designed primarily to maintain the U.S. nuclear weapons stockpile and run other classified simulations. The $600 million system, named El Capitan, consists of 87 computer racks weighing 1.3 million pounds and draws 30 megawatts of power.

Built by Hewlett-Packard Enterprise using AMD chips, it operates alongside a smaller system called Tuolumne, which ranks tenth globally in computing power. "While we're still exploring the full role AI will play, there's no doubt that it is going to improve our ability to do research and development that we need," said Bradley Wallin, a deputy director at the laboratory.

At CES 2025 today, Lenovo introduced the Legion Go S handheld gaming console. It marks the first officially licensed handheld that comes pre-loaded with Valve's Arch Linux based SteamOS operating system. Phoronix reports: This first officially licensed SteamOS handheld is making use of the AMD Ryzen Z1 Extreme SoC with Radeon 700M graphics, an 8-inch 1200p LCD touchscreen with VRR support, up to 32GB of LPDDR5x-6400 memory, up to 1TB of PCIe Gen4 SSD storage, and a 55 Whr battery. Pricing starts at $500 USD with availability beginning in May. Sadly this Lenovo Legion Go handheld running SteamOS is making use of the Ryzen Z1 Extreme and not the Ryzen Z2 announced by AMD yesterday with the Zen 5 cores. But at CES Lenovo is showing off the Lenovo Legion Go (8.8", 2) prototype that uses the AMD Ryzen Z2 Go SoC along with an OLED display albeit a Windows gaming device. Additional details are available in Lenovo's press release.

At CES 2025, AMD unveiled the Z2, Z2 Go, and Z2 Extreme chipsets -- all powered by Zen 5 CPU cores and designed for handheld gaming PCs. IGN reports: The AMD Zen 2 Extreme, along with lower-specced cousins the Z2 and Z2 Go, are powered by Zen 5 CPU cores. While the Zen 2 Extreme is using a RDNA 3.5-based GPU, the Z2 and Z2 Go are still using RDNA 3 and RDNA 2, respectively. This creates an entire family of APUs (Advanced Processing Units) for handheld gaming PCs that should hopefully cause the price of handhelds to go down a bit.

With the Z2 Extreme, AMD is hoping to dramatically improve battery life, while also delivering console-like gaming performance to devices like the Lenovo Legion Go. By and large, the biggest limiting factor of these handhelds, especially at the high end, is how quickly their batteries drain when playing demanding games away from a wall outlet. The company also introduced the "Fire Range" HX3D processors for gaming laptops, leveraging 3D V-cache technology for enhanced gaming performance and efficiency. "All of these mobile chipsets, from 'Fire Range' HX3D to the AMD Z2 Extreme, will end up in gaming laptops and handhelds over the next few months," adds IGN.

AMD published a press release with additional details and specifications.

Phoronix reports on Valve's "Steam Survey" results for December 2024, saying the new numbers "reflect a nice upward trend for the Linux gaming statistics and a high point in recent times." In November the Steam Survey reflected a 2.03% marketshare for Linux... Roughly inline with what we have been seeing for Linux right at around the 2% threshold. With the just-published December survey numbers, there is a 0.29% increase to 2.29%...! When looking at the Linux numbers, SteamOS Holo accounts for around 36% of all Linux gamers... SteamOS Holo being the operating system of the Steam Deck and beginning to appear on other devices as well... Driven in large part by the Steam Deck relying on a custom AMD SoC/APU and AMD being popular with Linux gamers/enthusiasts for their open-source driver support, AMD CPU use on Linux commands a 73.6% marketshare.
In fact, December "saw AMD reach another record-high share among participants of Valve's survey," according to TechSpot — "up 3.02% last month, taking its total to 38.7% as Intel fell slightly to 63.4%..." Elsewhere, Windows 11 is now comfortably the most popular OS in the survey. It pulled ahead another 2% to an almost 55% share in December as Windows 10 dropped to 42.3%... However, it's a different story when looking at global users: Windows 10's share has increased two months in a row to 62.7% while Windows 11 has declined to 34.1%. Rounding up the rest of the survey, 16GB of RAM remains the most popular amount of system RAM but it's lead is declining as second-place 32GB grows; a trend that is mirrored in the VRAM category...
Phoronix adds that the Windows percent "pulled back by 0.51% to 96.1% while Apple macOS also gained 0.22% going up to a 1.61% marketshare."

Earlier this year, Russia President Vladimir Putin called on the government to develop its own domestically produced gaming consoles with proprietary operating systems and cloud-based platforms. "With Russia heavily sanctioned and looking to promote its own products, one of its in-development consoles is powered by the Elbrus processor," notes TechSpot. However, the processor is "designed primarily for domestic applications in critical infrastructure, defense, and other sensitive areas" and "can't match high-end CPUs from Intel, AMD, and Arm." From the report: The Russian government admits that this device isn't going to be on the same level as current-gen machines. "I hope my colleagues will approach this task with full responsibility and come up with something truly groundbreaking," said Anton Gorelkin, Deputy Chairman of the State Duma Committee on Information Policy. "It is obvious to everyone: Elbrus processors are not yet at the level required to compete equally with the PS5 and Xbox, which means the solution must be unconventional." Gorelkin said that Russian consoles aren't being designed only to play ports of hundreds of old, less-demanding games. He added that they should primarily serve the purpose of promoting and popularizing domestic video game products.

Another organization following Putin's instructions is Russian telecommunications firm MTS. Its console (above) will use the company's cloud-based gaming platform, called Fog Play. It allows owners of high-end PCs to rent out their computing power to those with less-powerful equipment, charging an hourly price. Those with more powerful PCs can access games on the service and use their own hardware to play them. MTS' device is expected to cost no more than $45 and come with an Xbox-like controller, suggesting it's unlikely to appeal to those who enjoy current-gen console games.

"I still can't quite believe that I'm using an Arm-powered Windows laptop every day," writes a senior editor at the Verge: After more than a decade of trying to make Windows on Arm a reality, Microsoft and Qualcomm finally nailed it this year with Copilot Plus PCs. These new laptops have excellent battery life and great performance — and the app compatibility issues that have plagued Windows on Arm are mostly a thing of the past (as long as you're not a gamer). Microsoft wanted 2024 to be "the year of the AI PC," but I think it was very much the year of Windows on Arm...

The key to Windows on Arm's revival this year was Qualcomm's Snapdragon X Elite processors, which were announced in April. They've provided the type of performance and power efficiency only previously available with Apple's MacBooks and challenged Intel and AMD to do better in the x86 space. After much debate over Microsoft's MacBook Air-beating benchmarks, the reviews rolled in and showed that Windows on Arm was indeed capable of matching and beating Apple's MacBook Air. Qualcomm even hired the "I'm a Mac" guy to promote Windows on Arm PCs, showing how confident it was in challenging Apple's laptop dominance.

Microsoft and Qualcomm also worked closely with developers to make key apps compatible, and it's now very rare to run into an app compatibility issue that can't be solved by a native Arm64 version or Microsoft's improved emulator. Even Google, which previously shunned Windows Phone, has created Arm64 versions of Chrome and Google Drive to support Microsoft's efforts. With developers continually providing native versions of their apps, it makes it a lot easier to switch to a Windows on Arm laptop. The only big exception is gaming, where x86 still reigns supreme for compatibility and performance...

It's hard not to see 2025 as the year that Windows on Arm continues to eat into the laptop space. A Dell leak revealed Qualcomm is preparing new chips for 2025, and the chip maker has also been rolling out cheaper Arm-based chips to bring laptop prices down.
The article acknowledges that both AMD and Intel "have the key advantage of game compatibility that Windows on Arm is definitely not ready for..." But "Given the Windows on Arm gaming situation, a new generation of Nvidia's GPUs could help generate fresh excitement around x86 laptops throughout 2025." And "Nvidia might also be planning to help the Windows on Arm effort. The chip maker has long been rumored to be planning to launch Arm PC chips as soon as 2025... Whatever happens to laptops in 2025, you can guarantee that there's going to be fierce competition between Intel, AMD, and Qualcomm."

But the author still complains about the dedicated Copilot key on his new WIndows-on-Arm laptop. "While the Copilot experience on Windows has gone through several confusing revisions, it's still a key I accidentally press and then get frustrated when a Copilot window appears."

They provide training data to top AI companies including OpenAI and Meta, according to its web site. Founded in 2016, San Francisco-based Scale AI now has over 900 employees, eventually growing beyond "unicorn" status with over $1.35 billion in ivnestments. In May the company's valuation was over $14 billion, with investors including Amazon, Meta, Nvidia, Cisco, Intel, and AMD (as well as earlier investments from Y Combinator and $100 million from Peter Thiel's Founders Fund). SFGate calls them "a buzzy San Francisco startup with high-dollar ties across the tech industry".

But SFGate also report Scale AI "was sued Tuesday by a former worker with allegations that the company is committing wage theft and misclassifying workers." Steve McKinney filed the suit against Scale and several top executives, including 27-year-old billionaire CEO Alexandr Wang, in San Francisco Superior Court. With the filing, the former contractor aims to be a lead plaintiff for a class-action lawsuit against Scale; a judge will need to certify his proposed class of current and former contractors within California...

McKinney, whose complaint says he was paid on an hourly basis and worked on a project eventually sold to Meta, is accusing Scale of amassing its clout and cash by exploiting workers. "Scale AI is the sordid underbelly propping up the generative AI industry," the complaint says, before rattling off a list of allegations about its treatment of contractors. The document accuses Scale of bait-and-switch hiring promises; demanding off-the-clock, unpaid work; denying overtime pay; and unfairly booting contractors from projects...

The Tuesday complaint calls Scale's control over its contractors "Orwellian." The company makes contractors download a tool to track much of their computer use, including by taking periodic screenshots, the suit alleges. The lawsuit also alleges that Scale reassigns the workers to varyingly paid projects and docks pay if a task takes more than it was supposed to, plus posits that Scale is in violation of California's "ABC" test, which monitors use of the designation "independent contractor." It argues that contracted "Taskers" like McKinney should be classified as employees instead...

The complaint, along with arguing for class-action certification, seeks restitution, punitive damages and changes to Scale's worker classification model.
The article adds that "Per Fortune, Scale's armies of contractors marked up images for Cruise and Waymo to help autonomous cars understand their surroundings..."

HDMI Forum will announce new specifications with increased bandwidth capabilities at CES 2025, ahead of anticipated graphics card launches from AMD and NVIDIA. The announcement, scheduled for January 6, is expected to introduce HDMI 2.2 standard alongside a new cable supporting higher resolutions and refresh rates.

Current HDMI 2.1 specification maxes out at 48 Gbps bandwidth, allowing 10K resolution at 120 Hz with compression. The upgrade aims to compete with DisplayPort 2.1, which offers up to 80 Gbps bandwidth and is already supported by recent AMD and Intel GPUs.

The Verge thinks Valve "could make a play to dethrone the Sony PlayStation and Microsoft." And it's not just because there's lots of new SteamOS hardware on the way (including a wireless VR headset and a pair of trackable wands, a Steam Controller 2 gamepad, and a living room console.

"Valve has also now seemingly revealed plans for partners to create third-party SteamOS hardware too." It won't be easy to take on Sony, Microsoft, or Meta. Those companies have a lot to lose, and they're deeply entrenched. But the Steam Deck has revealed a massive weakness in each of their businesses that may take them years to correct — the desire to play a huge library of games anytime, anywhere. And while they figure that out, Valve may be building an entire new ecosystem of SteamOS hardware, one that could finally let PC and peripheral makers tap into the huge and growing library of Windows games on all sorts of different hardware without relying on Microsoft or subjecting their customers to the many annoyances of Windows...

Valve has long said it will open up SteamOS to other manufacturers, even recently committing to some direct support for rival handhelds like the Asus ROG Ally — and the other week, Valve quietly updated a document that may reveal its larger overarching strategy. It won't just leave SteamOS sitting around and hope manufacturers build something — it'll hold their hand. Valve now has an explicit label for third parties to create "Powered by SteamOS" devices, which it explicitly defines as "hardware running the SteamOS operating system, implemented in close collaboration with Valve." It additionally lets companies create "Steam Compatible" hardware that ships with "Valve approved controller inputs," as well as SteamVR hardware and Steam Link hardware that lets you stream games from one device to another...

When Valve asked PC manufacturers to sign onto its Steam Machines initiative over a decade ago, with the idea of building living room PC consoles, it asked for a leap of faith with very little to show and a tiny chance of success. It took years for Valve to even build the oddball living room controller for its Steam Machines, and it didn't get far in convincing Windows game developers to port their games to Linux. But by the time it announced the Steam Deck, Valve had hammered out a Proton software compatibility layer so good that many Windows games now run better on Linux, and created the most customizable yet familiar set of controls ever made. If manufacturers could build their own Steam Machines rather than equivalent Windows machines, they could offer better gaming products than they do today. Maybe they'd even want to release a VR headset that isn't tied to Microsoft or Meta if it doubled as a Steam Deck, portably playing decades of flatscreen games.

It's not clear any of this will pan out; Valve is an exceedingly small company that tries not to chase too many things at a time. When I speak to PC industry executives about why they pick Windows over SteamOS, some say they're concerned about whether Valve would truly be able to support them. But it's just as intriguing an idea as it was 12 years ago when Gabe Newell explained the initial vision to us, and this time, there's a far better chance it'll work.
"Today, every major PC company is building one or more Steam Deck rivals," the article points out. "But without Valve's blessing and support, they're saddled with a Windows OS that doesn't start, pause, and resume games quickly and seamlessly enough to feel portable and easy..."

An anonymous reader shared this report from Phoronix: New Linux patches from Huawei engineers are preparing new driver support for controlling High Bandwidth Memory (HBM) with the ARM-based Kunpeng high performance SoC...

[I]t would appear there is a new Kunpeng SoC coming that will feature integrated High Bandwidth Memory (HBM).Unless I missed something, this Kunpeng SoC with HBM memory hasn't been formally announced yet and I haven't been able to find any other references short of pointing to prior kernel patches working on this HBM integration... It will be interesting to see what comes of Huawei Kunpeng SoCs with HBM memory and ultimately how well they perform against other AArch64 server processors as well as the Intel Xeon and AMD EPYC competition.

Lawrence Livermore National Laboratory's "El Capitan" supercomputer is now ranked as the world's most powerful, exceeding a High-Performance Linpack (HPL) score of 1.742 exaflops on the latest Top500 list. Engadget reports: El Capitan is only the third "exascale" computer, meaning it can perform more than a quintillion calculations in a second. The other two, called Frontier and Aurora, claim the second and third place slots on the TOP500 now. Unsurprisingly, all of these massive machines live within government research facilities: El Capitan is housed at Lawrence Livermore National Laboratory; Frontier is at Oak Ridge National Laboratory; Argonne National Laboratory claims Aurora. [Cray Computing] had a hand in all three systems.

El Capitan has more than 11 million combined CPU and GPU cores based on AMD 4th-gen EPYC processors. These 24-core processors are rated at 1.8GHz each and have AMD Instinct M1300A APUs. It's also relatively efficient, as such systems go, squeezing out an estimated 58.89 Gigaflops per watt. If you're wondering what El Capitan is built for, the answer is addressing nuclear stockpile safety, but it can also be used for nuclear counterterrorism.

Slashdot unixbhaskar writes: Linus has released a fresh Linux kernel for public consumption. Please give it a try and report any glitches to the maintainers for improvement. Also, please do not forget to express your appreciation to those tireless folks who did all the hard work for you.
The blog OMG Ubuntu calls it "one of the most biggest kernel releases for a while," joking that it's a "really real-time kernel." The headline feature in Linux 6.12 is mainline support for PREEMPT_RT. This patch set dramatically improves the performance of real-time applications by making kernel processes pre-emptible — effectively enabled proper real-time computing... Meanwhile, Linus Torvalds himself contributes a new method for user-space address masking designed to claw back some of the performance lost due to Spectre-v1 mitigations.

You might have heard that kernel devs have been working to add QR error codes to Linux's kernel panic BSOD screen (as a waterfall of error text is often cut off and not easily copied for ad-hoc debugging). Well, Linux 6.12 adds support for those during Direct Rendering Manager panics...

A slew of new RISC-V CPU ISA extensions are supported in Linux 6.12; hybrid CPU scaling in the Intel P-State driver lands ahead of upcoming Intel Core Ultra 2000 chips; and AMD P-State driver improves AMD Boost and AMD Preferred Core features.
More coverage from the blog 9to5Linux highlights a new scheduler called sched_ext, Clang support (including LTO) for nolibc, support for NVIDIA's virtual command queue implementation for SMMUv3, and "an updated cpuidle tool that now displays the residency value of cpuidle states for a clearer and more detailed view of idle state information when using cpuidle-info." Linux kernel 6.12 also introduces SWIG bindings for libcpupower to make it easier for developers to write scripts that use and extend the functionality of libcpupower, support for translating normalized error addresses reported by an AMD memory controller into system physical addresses using a UEFI mechanism called platform runtime mechanism (PRM), as well as simplified loading of microcode patches on AMD Zen and newer CPUs by using the family, model, and stepping encoded in the patch revision number...

Moreover, Linux 6.12 adds support for running as a protected guest on Android as well as perf and support for a bunch of new interconnect PMUs. It also adds the final conversions to the new Intel VFM CPU model matching macros, rewrites the PCM buffer allocation handling and locking optimizations, and improves the USB audio driver...

French private AI lab PleIAs "is committed to training LLMs in the open," they write in a blog post at Mozilla.org. "This means not only releasing our models but also being open about every aspect, from the training data to the training code. We define 'open' strictly: all data must be both accessible and under permissive licenses."

Wednesday PleIAs announced they were releasing the largest open multilingual pretraining dataset, according to their blog post at HuggingFace: Many have claimed that training large language models requires copyrighted data, making truly open AI development impossible. Today, Pleias is proving otherwise with the release of Common Corpus (part of the AI Alliance Open Trusted Data Initiative) — the largest fully open multilingual dataset for training LLMs, containing over 2 trillion tokens of permissibly licensed content with provenance information (2,003,039,184,047 tokens).

As developers are responding to pressures from new regulations like the EU AI Act, Common Corpus goes beyond compliance by making our entire permissibly licensed dataset freely available on HuggingFace, with detailed documentation of every data source. We have taken extensive steps to ensure that the dataset is high-quality and is curated to train powerful models. Through this release, we are demonstrating that there doesn't have to be such a [heavy] trade-off between openness and performance.

Common Corpus is:

— Truly Open: contains only data that is permissively licensed and provenance is documented

— Multilingual: mostly representing English and French data, but contains at least 1B tokens for over 30 languages

— Diverse: consisting of scientific articles, government and legal documents, code, and cultural heritage data, including books and newspapers

— Extensively Curated: spelling and formatting has been corrected from digitized texts, harmful and toxic content has been removed, and content with low educational content has also been removed.


Common corpus builds on a growing ecosystem of large, open datasets, such as Dolma, FineWeb, RefinedWeb. The Common Pile currently in preparation under the coordination of Eleuther is built around the same principle of using permissible content in English language and, unsurprisingly, there were many opportunities for collaborations and shared efforts. But even together, these datasets do not provide enough training data for models much larger than a few billion parameters. So in order to expand the options for open model training, we still need more open data...

Based on an analysis of 1 million user interactions with ChatGPT, the plurality of user requests are for creative compositions... The kind of content we actually need — like creative writing — is usually tied up in copyright restrictions. Common Corpus tackles these challenges through five carefully curated collections...
Last week AMD also released its first series of fully open 1 billion parameter language models, AMD OLMo.

And last month VentureBeat reported that the non-profit Allen Institute for AI had unveiled Molmo, "an open-source family of state-of-the-art multimodal AI models which outpeform top proprietary rivals including OpenAI's GPT-4o, Anthropic's Claude 3.5 Sonnet, and Google's Gemini 1.5 on several third-party benchmarks."

AMD has announced plans to cut 4% of its global workforce as it repositions to compete in the AI chip market dominated by Nvidia. The layoffs will affect approximately 1,040 employees of its 26,000-strong workforce reported at the end of 2023. CNBC adds: AMD produces powerful AI accelerators for data centers, including the MI300X, which companies such as Meta and Microsoft purchase as an alternative to Nvidia-based systems. But Nvidia dominates the market for powerful AI chips, with over 80% market share, partially because it developed the core software that AI engineers use to develop programs such as OpenAI's ChatGPT.

Red Hat, the IBM-owned open source software firm, is acquiring Neural Magic, a startup that optimizes AI models to run faster on commodity processors and GPUs. From a report: The terms of the deal weren't disclosed. MIT research scientist Alex Matveev and professor Nir Shavit founded Somerville, Massachusetts-based Neural Magic in 2018, inspired by their work in high-performance execution engines for AI. Neural Magic's software aims to process AI workloads on processors and GPUs at speeds equivalent to specialized AI chips (e.g. TPUs). By running models on off-the-shelf processors, which usually have more available memory, the company's software can realize these performance gains.

Big tech companies like AMD and a host of other startups, including NeuReality, Deci, CoCoPie, OctoML and DeepCube, offer some sort of AI optimization software. But Neural Magic is one of the few with a free platform and a collection of open source tools to complement it. Neural Magic had so far managed to raise $50 million in venture capital from backers like Andreessen Horowitz, New Enterprise Associations, Amdocs, Comcast Ventures, Pillar VC and Ridgeline Ventures.

An anonymous reader shares a report: AMD has gained a substantial 5.7 percentage points of share of the desktop x86 CPU market in the third quarter compared to Q2, the largest quarterly share gain since we began tracking the market share reports in 2016. It also represents an incredible ten percentage point improvement over the prior year. AMD also raked in a strong increase in revenue share, jumping 8.5 percentage points over the prior quarter, indicating that it is selling a strong mix of higher-end CPU models.

During the quarter, AMD launched its new Ryzen 9000-series family of processors amid a scandal related to stability issues with Intel's Raptor Lake chips, which generated a flood of negative press for the company over the course of several months, and inventory adjustments for one of Intel's customers. AMD now commands 28.7% of the desktop processor market. AMD also continued to gain share in the laptop and server markets, though its gains on the desktop side of the business were the most impressive, according to Mercury Research.