An anonymous reader quotes a report from Ars Technica: Qualcomm has a new chip for flagship phones, and the best part is that it uses an improved version of the Oryon CPU architecture that the Snapdragon X Elite chips brought to Windows PCs earlier this year. The Snapdragon 8 Elite is the follow-up to last year's Snapdragon 8 Gen 3 -- yet another change to the naming convention that Qualcomm uses for its high-end phone chips, though, as usual, the number 8 is still involved. The 8 Elite uses a "brand-new, 2nd-generation Qualcomm Oryon CPU" with clock speeds up to 4.32 GHz, which Qualcomm says will improve performance by about 45 percent compared to the Snapdragon 8 Gen 3.

Rather than a mix of large, medium, and small CPU cores as it has used in the past, the 8 Elite has two "Prime" cores for hitting that high peak clock speed, while the other six are all "Performance" cores that peak at a lower 3.53 GHz. But it doesn't look like Qualcomm is using a mix of different CPU architectures anymore, choosing to distinguish the higher-performing core from the lower-performing ones by clock speed alone. Qualcomm promises a similar 40 percent performance boost from the new Adreno 830 GPU. The chip also includes a marginally improved Snapdragon X80 5G modem, up from an X75 modem in the Snapdragon 8 Gen 3 -- its main improvement appears to be support for additional antennas, for a total of six, but the download speed still tops out at a theoretical 10Gbps. Wi-Fi 7 support appears to be the same as in the 8 Gen 3, but the 8 Elite does support the Bluetooth 6.0 standard, up from Bluetooth 5.4 in the 8 Gen 3.

Qualcomm says the new chip's CPU features "44% improved power efficiency" and "40% greater power efficiency" for the GPU, which ought to keep power usage in line despite the performance improvements -- these gains are probably attributable to the new 3 nm TSMC manufacturing process, compared to the 4 nm process used for the Snapdragon 8 Gen 3. And no 2024 chip announcement would be complete without some kind of AI mention: Qualcomm's image signal processor is now an "AI ISP," which Qualcomm says "recognizes and enhances virtually anything in the frame, including faces, hair, clothing, objects, backgrounds, and beyond." These capabilities can allow it to remove objects from the background of photos, among other things, using the on-device processing power of the chip's Hexagon neural processing unit (NPU). The NPU is 45 percent faster than the one in the Snapdragon 8 Gen 3. Phones using the Snapdragon 8 Elite should begin appearing in "the coming weeks."

jd writes: Linus Torvalds is not a happy camper and is condemning hardware vendors for poor security and the plethora of actual and theoretical attacks, especially as some of the new features being added impact the workarounds. These workarounds are now getting very expensive, CPU-wise.

TFA quotes Linus Torvalds:

"Honestly, I'm pretty damn fed up with buggy hardware and completely theoretical attacks that have never actually shown themselves to be used in practice.

"So I think this time we push back on the hardware people and tell them it's *THEIR* damn problem, and if they can't even be bothered to say yay-or-nay, we just sit tight.

Because dammit, let's put the onus on where the blame lies, and not just take any random shit from bad hardware and say 'oh, but it *might* be a problem.'"

Libreboot is a distribution of coreboot "aimed at replacing the proprietary BIOS firmware contained by most computers."

So then what exactly is GNU Boot? Its home page explains... In November 2022, Libreboot began to include non-libre code. We have made repeated efforts to continue collaboration with those developers to help their version of Libreboot remain libre, but that was not successful. Now we've stepped forward to stand up for freedom, ours and that of the wider community, by maintaining our own version — a genuinely libre Libreboot, that after some hurdles gave birth to this project: GNU Boot.
But today, Phoronix writes: While priding itself on being "100% free", last December [GNU Boot] had to drop some motherboard support and CPU code after discovering they were shipping some files that are non-free by their free software standards. Today they announced another mistake in having inadvertently been shipping additional non-free code.

GNU Boot discovered an issue with non-free code affecting not only them but also some of the Linux distributions that pride themselves on being fully free software / 100% open-source. This latest snafu they say is "more problematic" than their prior non-free code discover due to impacting the free software Linux distributions too. The issue at hand though comes down to test data contained within the archive and that containing non-free code in the form of microcode, BIOS bits, and Intel Management Engine firmware.
"We also contacted Replicant..." according to the announcement, "a free Android distro that also ships vboot source code." And in addition, "We had to re-release all the affected tarballs." (Which at this point is three release candidates...)

Tom's Hardware reports on some interesting hardware history being shared on X.com: AMD engineer Phil Park identified a curious nugget of PC architectural history from, of all places, a year-old Quora answer posted by former Intel engineer [and Pentium Pro architect] Robert Colwell. The nugget indicates that Intel could have beaten AMD to the x86-64 punch if the former wasn't dead-set on the x64-only Itanium line of CPUs.
Colwell had responded on Quora to the question "Shouldn't Intel with its vast resources have been able to develop both architectures?" This was a marketing decision by Intel — they believed, probably rightly, that bringing out a new 64-bit feature in the x86 would be perceived as betting against their own native-64-bit Itanium, and might well severely damage Itanium's chances. I was told, not once, but twice, that if I "didn't stop yammering about the need to go 64-bits in x86 I'd be fired on the spot" and was directly ordered to take out that 64-bit stuff. I decided to split the difference, by leaving in the gates but fusing off the functionality. That way, if I was right about Itanium and what AMD would do, Intel could very quickly get back in the game with x86. As far as I'm concerned, that's exactly what did happen.
Phil Park continued the discussion on X.com. "He didn't quite get what he wanted, but he got close since they had x86-64 support in subsequent products when Intel made their comeback." (So, Park posted later in the thread, "I think he won the long game.")

Park also shared a post from Nicholas Wilt (NVIDIA CUDA designer who earlier did GPU computing work at Microsoft and built the prototype for Windows Desktop Manager): I have an x86-64 story of my own. I pressed a friend at AMD to develop an alternative to Itanium. "For all the talk about Wintel," I told him, "these companies bear no love for one another. If you guys developed a 64-bit extension of x86, Microsoft would support it...."

Interesting coda: When it became clear that x86-64 was beating Itanium in the market, Intel reportedly petitioned Microsoft to change the architecture and Microsoft told Intel to pound sand.

In a press release this morning, Apple announced a new iPad Mini with a faster A17 Pro chip that supports Apple Intelligence. The Verge reports: The new Mini is mostly a spec bump: it runs a new A17 Pro chip, which Apple says has a 30 percent faster CPU, 25 percent faster GPU, and a Neural Engine twice as fast as the previous model. The device also supports the new Apple Pencil Pro, which is a nice touch for the Mini-toting artists out there, and comes with 128GB of storage in the base model rather than 64GB. (Those AI models need all the space they can get.) The Wi-Fi 6E chip is faster, the USB-C port is faster, everything about the iPad Mini is the same as before only faster this time.

The only real design change with the new Mini is the colors. Apple's gone more colorful with a lot of its products this year, and the Mini comes in new purple and blue models. In photos they look muted rather than vivid, though, so don't expect the eye-popping new colors on the iPhone 16.

Boston-based engineer and photographer Wenting Zhang built his own full-frame camera and open-sourced the project on GitLab for anyone else to build upon. The camera, named Sitina S1, features a 10MP CCD sensor, custom electronics, and a 3D-printed body. Digital Photography Review reports: Zhang says he started the project in 2017, and it's not finished yet. "Engineers are usually bad at estimating how long things will take. I am probably particularly bad at that. I expected this project to be challenging, so it would take a bit longer, like probably one year. Turned out my estimation was off," he says. He makes clear to point out that this is a hobby project, purely for fun, and that his camera isn't going to achieve the level of image quality found in commercially available products from established companies. Despite that, his project provides a fascinating look into what's involved in building a camera from the ground up.

Although CMOS has become the dominant sensor technology in consumer cameras, owing to factors like speed, lower power consumption and cost, Zhang's camera is built around a 10MP Kodak KAI-11000CM CCD sensor with a global electronic shutter, which he selected for a rather pragmatic reason: it was easy to source. "Most manufacturers (like Sony) aren't going to just sell a sensor to a random hobbyist, so I have to buy whatever is available on eBay. This 10MP CCD turned out to be available," he explains. The choice of sensor has a useful benefit. As he explains in one of his videos, designing and building a mechanical shutter is complicated and beyond his area of expertise, so his DIY design is based on using an electronic shutter. For similar reasons, he chose to use an LCD screen as a viewfinder rather than a prism-based optical design, resulting in a mirrorless camera.

Zhang wanted his design to be compatible with existing lenses. His mirrorless design, with a short flange distance, provided a great deal of flexibility to adapt different lenses to the camera, and he's currently using E-mount with active electrical contacts. And that's just the start. Zhang also needed to integrate a CCD signal processor with an ADC (analog to digital converter), a CPU, battery, an LCD screen and buttons. He also designed and built his own circuit board with a power-only USB port, flash sync terminal, power button and SD card slot, and create the software and user interface to tie it all together. Finally, everything fits inside a 3D-printed enclosure that, to my eye, looks rather attractive.

Phoronix's Michael Larabel reports: OpenBSD 7.6 is out this evening as another major step forward for this BSD operating system with enhanced hardware support, security improvements, updating various user-space software, and enabling other kernel enhancements. There are a ton of changes to find with the just-released OpenBSD 7.6.

Some of the new OpenBSD 7.6 features include:

- OpenBSD 7.6 provides initial support for Qualcomm Snapdragon X1 Elite (X1E80100) SoCs. The 7.6 release also has initial Samsung Galaxy Book4 Edge boot support in ACPI mode with OpenBSD 7.6.
- ARM64 has additional CPU security mitigations with Spectre-V4 now in place on ARM64 and adding Spectre-BHB for Cortex-A57 cores.
- OpenBSD 7.6 on RISC-V now supports the Milk-V Pioneer board.
- OpenBSD 7.6 on AMD64 has finally implemented support for AVX-512.
- Various SMP kernel improvements. You can view the full list of features and download the OpenBSD 7.6 release via OpenBSD.org.

"Alleged photos and videos of an unannounced 14-inch MacBook Pro with an M4 chip continue to surface on social media, in what could be the worst product leak for Apple since an employee accidentally left an iPhone 4 prototype at a bar in California in 2010," writes MacRumors' Joe Rossignol. From the report: The latest video of what could be a next-generation MacBook Pro was shared on YouTube Shorts today by Russian channel Romancev768, just one day after another Russian channel shared a similar video. The clip shows a box for a 14-inch MacBook Pro that is apparently configured with an M4 chip with a 10-core CPU and a 10-core GPU, 16GB of RAM, 512GB of storage, three Thunderbolt 4 ports, and a Space Black finish. According to the "About This Mac" software menu shown in the video, the MacBook Pro in the video is allegedly an unreleased November 2024 model. [...]

Apple is well known for having a culture of secrecy, so this magnitude of leak is rarely seen for its products. As previously mentioned, this could be the most significant leak for Apple since Gizmodo obtained and shared photos of an iPhone 4 prototype that a then-employee of the company accidentally left behind at a bar in California. In that case, Apple got law enforcement involved, but how it acts this time around remains to be seen.

He's the long-time Slashdot reader who installed Linux on a 1993 PC — and then installed a 1994 version of MS-DOS on a modern Thinkpad X13. (And somewhere along the way, he even built a ChatGPT client for DOS...)

But in a new blog post, yeokm1 reveals "I recently built myself a PC," salvaging parts from a previous desktop system to bootstrap an upgrade. And "I decided to build one with the ability to still reach back into the past to run MS-DOS..."

The result? A Ryzen 5 7600 and GeForce 4060 Ti system, but with a floppy drive, optical drive, Sound Blaster card, serial, parallel and PS/2 ports — that runs MS-DOS. The fact that a 30-year-old MS-DOS 6.22 can still work well enough on such a modern hardware is testament to the efforts made by the industry to ensure good x86 PC backward compatibility. AMD, Nvidia and Asus deserve to be commended on their efforts here.

I'm also impressed that the modern Nvidia Geforce 4060 Ti still supports some legacy video BIOS modes to a usable level although this is not complete. I didn't document in this blog post but brief tests with other VESA modes and resolutions didn't work so well. I wonder how long more this amount of x86 PC backward-compatibility will continue to last though... It definitely feels like the end is near.
Their blog post includes a video about their system. (And yes, it plays Doom.) But their ultimate goal is to use it to play modern games like Cyberpunk 2077 and Flight Simulator 2020 (as well as the upcoming Flight Simulator 2024) "at reasonably good settings and performance." (And also to experiment with light machine-learning workloads, do basic video editing, run virtual machines.)

After successfully building their DOS-running system, they asked ChatGPT what it thought. Would the system's specs be powerful enough to handle the 30-year-old operation system? And ChatGPT confidently replied:

"Neither the Ryzen 5 7600 nor the GeForce RTX 4060 Ti is designed to run DOS natively. DOS is an older operating system that was primarily used on x86 architecture from the late 20th century, and modern hardware like the Ryzen 5 7600 and GeForce RTX 4060 Ti are not compatible with DOS due to their 64-bit architecture and lack of necessary drivers to interface correctly with DOS, which relied on much older technology..."

yeokm1's blog post concludes: "I think I just proved ChatGPT wrong :P"

AMD has released BIOS updates to boost performance and reduce latency for its Ryzen 9600X and 9700X processors. The updates come a month after disappointing Zen 5 desktop CPU reviews and coincide with Windows 11 optimizations for AMD chips. The new AGESA PI 1.2.0.2 firmware addresses high core-to-core latency issues and introduces a 105-watt cTDP option, promising up to 10% performance gains for multithreaded workloads.

Long-time Slashdot reader dmitrygr writes: Debian Linux booted on a 4-bit intel microprocessor from 1971 — the first microprocessor in the world — the 4004. It is not fast, but it is a real Linux kernel with a Debian rootfs on a real board whose only CPU is a real intel 4004 from the 1970s.
There's a detailed blog post about the experiment. (Its title? "Slowly booting full Linux on the intel 4004 for fun, art, and absolutely no profit.")

In the post dmitrygr describes testing speed optimizations with an emulator where "my initial goal was to get the boot time under a week..."

Intel has released microcode update 0x12B for its 13th and 14th generation Core processors, addressing persistent stability issues stemming from voltage irregularities. The update targets a specific clock tree circuit within the CPU's IA core that was causing elevated voltage requests during idle and light workloads.

The company identified four key factors contributing to voltage instability: motherboards exceeding Intel's power specifications, an Enhanced Thermal Velocity Boost algorithm allowing sustained high performance at elevated temperatures, frequent high voltage requests from the processor, and problematic microcode demanding elevated core voltages during low-activity periods. While previous update 0x129 addressed some concerns, the new 0x12B update aims to resolve the root cause of the "Vmin shift" problem, where voltage spikes lead to increased power requirements and potential degradation over time. Intel is working with motherboard manufacturers to roll out BIOS updates incorporating the new microcode.

The JavaScript runtime Bun is a Node.js/Deno alternative (that's also a bundler/test runner/package manager).

And Bun 1.1.28 now includes experimental support for ">compiling and running native C from JavaScript, according to this report from The New Stack: "From compression to cryptography to networking to the web browser you're reading this on, the world runs on C," wrote Jarred Sumner, creator of Bun. "If it's not written in C, it speaks the C ABI (C++, Rust, Zig, etc.) and is available as a C library. C and the C ABI are the past, present, and future of systems programming." This is a low-boilerplate way to use C libraries and system libraries from JavaScript, he said, adding that this feature allows the same project that runs JavaScript to also run C without a separate build step... "It's good for glue code that binds C or C-like libraries to JavaScript. Sometimes, you want to use a C library or system API from JavaScript, and that library was never meant to be used from JavaScript," Sumner added.

It's currently possible to achieve this by compiling to WebAssembly or writing a N-API (napi) addon or V8 C++ API library addon, the team explained. But both are suboptimal... WebAssembly can do this but its isolated memory model comes with serious tradeoffs, the team wrote, including an inability to make system calls and a requirement to clone everything. "Modern processors support about 280 TB of addressable memory (48 bits). WebAssembly is 32-bit and can only access its own memory," Sumner wrote. "That means by default, passing strings and binary data JavaScript WebAssembly must clone every time. For many projects, this negates any performance gain from leveraging WebAssembly."

The latest version of Bun, released Friday, builds on this by adding N-API (nap) support to cc [Bun's C compiler, which uses TinyCC to compile the C code]. "This makes it easier to return JavaScript strings, objects, arrays and other non-primitive values from C code," wrote Sumner. "You can continue to use types like int, float, double to send & receive primitive values from C code, but now you can also use N-API types! Also, this works when using dlopen to load shared libraries with bun:ffi (such as Rust or C++ libraries with C ABI exports)....

"TinyCC compiles to decently performant C, but it won't do advanced optimizations that Clang or GCC does like autovectorization or very specialized CPU instructions," Sumner wrote. "You probably won't get much of a performance gain from micro-optimizing small parts of your codebase through C, but happy to be proven wrong!"

An anonymous Slashdot reader writes: Haiku (the MIT-licensed operating system, inspired by BeOS) has released its fifth beta for Haiku R1.

Some new features include improved UI color management, improved dark mode coloring, Tracker improvements, TUN/TAP support for VPN connections, TCP throughput improvements, performance optimizations, UFS2 (BSD's filesystem) read-only support, new FAT filesystem driver, improved hardware support, improved POSIX compliance, improved performance, and more.
Slashdot has been covering the fate of the BeOS since 2000 (as well as the short-lived derivative project ZETA — and Haiku).

And now "With a history of over two decades and previously known as OpenBeOS, today's Haiku is pushing forward..." writes the site NotebookCheck: Haiku is a spiritual successor to BeOS, with a focus on a clean and user-friendly design paired with low system requirements. The minimum system requirements are still an Intel Pentium II/AMD Athlon CPU or better, at least 384 MB RAM, an 800x600 screen, and at least 3GB storage. It works on both 32-bit and 64-bit x86 PCs, and the 32-bit version can run many unmodified BeOS applications. It might be the best desktop open-source operating system not based on Linux or Unix... It works well in a virtual machine like VirtualBox or UTM.

Nvidia is banking on its software expertise and broad GPU ecosystem to stay ahead in the fiercely competitive AI chip market, CEO Jensen Huang said in an interview with Goldman Sachs Wednesday. Huang pointed to NVIDIA's large base of installed GPUs and their software compatibility as key strengths.

Huang highlighted three key elements of Nvidia's competitive moat: a large installed base of GPUs across multiple platforms, the ability to enhance hardware with software like domain-specific libraries, and expertise in building rack-level systems. The CEO said Nvidia's chip design prowess, noting the company has developed seven different chips for its upcoming Blackwell platform.

These comments come as Nvidia faces increasing competition from rivals. Addressing supply chain concerns, Huang said NVIDIA has sufficient in-house intellectual property to shift manufacturing if necessary without significant disruption. The company plans to begin shipping Blackwell-based products in the fourth quarter of fiscal 2025, with volume production ramping up in fiscal 2026, according to Huang.

From the note that Goldman Sachs sent to its clients: 1) Accelerated Computing: Mr. Huang highlighted his long-held view that Moore's Law was no longer delivering the rate of innovation it had in the past and, as such, was driving computation inflation in Data Centers. Further, he noted that the densification and acceleration of the $1 trillion data center infrastructure installed base alone would drive growth over the next 10 years, as it would deliver material performance improvement and/or cost savings.

2) Customer ROI: Mr. Huang noted that we have hit the end of transistor scaling that enabled better utilization rates and cost reductions in the previous virtualization and cloud computing cycles. He explained that, while using a GPU to augment a CPU will drive an increase in cost in absolute terms (~2x) in the case of Spark (distributed processing system and analytics engine for big data), the net cost benefit could be as large as ~10x for an application like Spark given the speed up of ~20x. From a revenue generation perspective, Mr. Huang shared that hyperscale customers can generate $5 in rental revenue for every $1 spent on Nvidia's infrastructure, given sustained strength in the demand for accelerated computing.

Intel launched its new Core Ultra 200V-series processors on Tuesday, promising significant improvements in power efficiency, performance, and battery life over competitors and previous generations. The company claims the chips offer "historic x86 power efficiency" and the "world's fastest mobile CPU cores." The processors, available for pre-order in OEM systems and shipping September 24, feature four Lion Cove P-cores and four Skymont E-cores with boost speeds up to 5.1 GHz.

Intel says the chips deliver up to 20.1 hours of battery life, Tom's Hardware reports, outperforming Qualcomm's Snapdragon X Elite by nearly two hours and AMD's chips by almost four hours. Intel asserts a 30% faster gaming performance than competing processors and highlighted compatibility issues with Qualcomm's chips, noting that nearly two dozen games used for benchmarking failed to run on X Elite chips. The company claims up to 64% advantage in single-threaded performance over Qualcomm Snapdragon X Elite and up to 33% over AMD Strix Point HX370.

Raspberry Pi, the British computer manufacturer, unveiled a new 2GB variant of its flagship Raspberry Pi 5 single-board computer on Monday, priced at $50. Raspberry Pi CEO Eben Upton said the company aims to "bring high-performance general-purpose computing to the widest possible audience" with the new offering. The 2GB Raspberry Pi 5 utilizes a cost-optimized D0 stepping of the BCM2712 application processor, which removes non-essential functionality to reduce manufacturing costs.

Upton stated the chip is "functionally identical" to users compared to higher-memory variants. While the reduced RAM may limit multitasking capabilities, Raspberry Pi's optimized OS allows for efficient resource usage. The company expects the 2GB model to suffice for many users' needs, while power users may opt for 4GB or 8GB versions priced at $60 and $80 respectively. The Raspberry Pi 5, launched in October 2023, features a quad-core Arm Cortex-A76 CPU running at 2.4GHz, dual 4K display output, and support for PCIe SSDs. Upton noted the latest model is "about 150 times as powerful" as the original Raspberry Pi from 2012.

Slashdot reader yeokm1 is the Singapore-based embedded security researcher whose side projects include installing Linux on a 1993 PC and building a ChatGPT client for MS-DOS.

Today he writes: When one thinks of modern technologies like Thunderbolt, 2.5 Gigabit Ethernet and modern CPUs, one would associate them with modern operating systems. How about DOS?

It might seem impossible, however I did an experiment on a relatively modern 2020 Thinkpad and found that it can still run MS-DOS 6.22. MS-DOS 6.22 is the last standalone version of DOS released by Microsoft in June 1994. This makes it 30 years old today.

I'll share the steps and challenges in locating a modern laptop capable of doing so — and the challenge of making the 30-year-old OS work on it with audio and networking functions. This is likely among the final generation of laptops able to run DOS natively.

AMD increased its market share in data center and laptop CPU segments during Q2 2024, according to a new report from Mercury Research. The company captured 24.1% of the data center CPU market, up 0.5% from the previous quarter and 5.6% year-over-year. In laptops, AMD's share rose to 20.3%, a 1% increase quarter-over-quarter and 3.8% year-over-year. The company's revenue share in laptops reached 17.7%, indicating lower average selling prices compared to Intel.

Intel maintained its overall lead, controlling 78.9% of the client PC market. In desktops, Intel gained 1% share, now holding 77% of the market. AMD's data center revenue share hit 33.7%, suggesting higher average selling prices for its EPYC processors compared to Intel's Xeon chips. AMD earned $2.8 billion from 24.1% unit share, while Intel made $3.0 billion from 75.9% unit share.

After months of back and forth, Intel has finally agreed to extend the warranty on all affected 13th- and 14th-generation desktop CPUs by an additional two years. This extension increases the warranty period for new boxed Intel CPUs from three to five years. For CPUs pre-installed in systems, Intel directs users to contact their PC's manufacturer for support, maintaining its established channels for warranty claims. The Verge adds: Intel has said that a primary cause of the instability issues for the desktop CPUs was due to an "elevated operating voltage" and that it was working on a patch for mid-August that addresses the root cause of that. But the patch apparently won't fix any damage that's already happened, meaning the best way to fix a damaged chip is to replace it.