Hardware Archive
A wider bipartisan group of U.S. lawmakers is asking the Biden administration about its plans to respond to China’s rising use of RISC-V chip design technology after Reuters last month reported on growing concerns about it in both houses of Congress. Now, a broader group of 18 lawmakers that includes five Democrats is asking the Biden administration for how it plans to prevent China “from achieving dominance in … RISC-V technology and leveraging that dominance at the expense of U.S. national and economic security,” according to a letter the group sent to Raimondo and seen by Reuters. A rather shortsighted take, and without even looking I wouldn’t be surprised if some of these lawmakers have chip factories or whatever in their districts.
There’s an alternative universe where we decided to teach the kernel about every piece of hardware it should run on. Fortunately (or, well, unfortunately) we’ve seen that in the ARM world. Most device-specific simply never reaches mainline, and most users are stuck running ancient kernels as a result. Imagine every x86 device vendor shipping their own kernel optimised for their hardware, and now imagine how well that works out given the quality of their firmware. Does that really seem better to you? It’s understandable why ACPI has a poor reputation. But it’s also hard to figure out what would work better in the real world. We could have built something similar on top of Open Firmware instead but the distinction wouldn’t be terribly meaningful – we’d just have Forth instead of the ACPI bytecode language. Longing for a non-ACPI world without presenting something that’s better and actually stands a reasonable chance of adoption doesn’t make the world a better place. Matthew Garrett with the usual paragraphs of wisdom.
This project allows old x86 computers using a classic BIOS to boot from modern NVMe storage attached via PCI(e). It’s a heavily modified version of iPXE (which usually allows for booting from the network), but instead of the network, this code uses a port of the SeaBIOS NVMe implementation to talk to a local NVMe drive. What a useful idea.
Tom’s Hardware reports: MyDrivers has published a review of Loongson’s 3A6000 quad-core CPU, confirming that the chip’s IPC improvements are real. Benchmarks reveal that the 3A6000 enjoys an impressive 60% performance uplift in single-core performance and an even more impressive 2x performance multiplier in multi-core performance over its 3A5000 predecessor. With these improvements, the 3A6000 features performance comparable to a Core i3-10100F, with the IPC performance of a Zen 3 chip. Of course, both Intel’s Comet Lake 10th Gen architecture and AMD’s Zen 3 architecture are now coming up on three years old. They’re nowhere near the top of our list of the best CPUs for gaming or other purposes. But it still represents a step in the right direction. Chinese chipmakers are improving quite fast, but unless they can somehow get access to the latest machinery from the Dutch company ASML, which makes virtually all of the machines capable of producing the chips with the smallest nanometers and is the linchpin in the entire semiconductor industry, they won’t be able to overtake or even match what TSMC and Intel are doing. That being said, I love weird processors, and I’d love to get my hands on one of these to play around with.
I’ve recently been working on putting together a CI system for postmarketOS that will allow us to do proper automated integration testing. That is to say – when someone opens a merge request that modifies our initramfs (for example), we should be able to click a button and some minutes later know that this change doesn’t break any of our important usecases. QEMU absolutely can (and will) get us most of the way there, but at some point we need to just run the same software that we’re running on end user devices. Furthermore, QEMU can’t tell us anything about changes in the kernel that might affect our devices, and manually testing during kernel upgrades, frankly, sucks. So we need a fancy board farm, this is one of those things where folks with the right technical background could build something over the course of a week. But for someone like me it’s full of trial and error and hidden complexity… It’s easy enough to do this with one device – just hack something together, but to be successful we need something reliable and adaptable, that we can adjust to fit our needs in the future, and the wide range of devices we support. Now this is an article you won’t come across very often, as the number of people setting up something like this who can actually talk openly about it – someone doing this for a closed company probably can’t – is probably quite small. A great read.
In 2021, at the height of cryptocurrency mining, Nvidia released the Nvidia CMP 170HX. Designed as a compute-only card to accelerate Ethereum’s memory-hard Ethash Proof-of-Work mining algorithm with its 1500 GB/s HBM2e memory bus, Nvidia implemented the hardware using the GA100 silicon from their Ampere architecture. Thus, the CMP 170HX is essentially a variant of the all-mighty Nvidia A100, Nvidia’s top-performing datacenter GPU at that time. Naturally, the existence of the CMP 170HX raised many questions, including its potential in applications beyond mining. Today, following the discontinuation of Ethash, these $5000 GPUs from closed mining farms are sold on second-hand markets for $400-$500 in China. It’s time to answer these questions. This article contains a basic performance overview, a hardware teardown, a watercooling installation guide, and a repair log. I’m glad smart people are at least trying to turn otherwise useless hardware designed for one of the most brazenly useless applications in human history into something potentially useful.
Arm has traditionally targeted the low end of the power and performance curve, but just as Intel has been looking to expand into the low power market, ARM is looking to expand into higher power and performance segments. The Cortex X series is at the forefront of this effort. Here, we’ll be looking at the Cortex X2 as implemented in the Snapdragon 8+ Gen 1. This SoC features a single X2 core, alongside four Cortex A510 and three Cortex A710 cores. The Cortex X2 in this SoC typically runs at 2.8 GHz, although lscpu indicates its clock speed can range from 787.2 MHz to 3.187 GHz. An in-depth look at this performance ARM core.
While Qualcomm has become wildly successful in the Arm SoC market for Android smartphones, their efforts to parlay that into success in other markets has eluded them so far. The company has produced several generations of chips for Windows-on-Arm laptops, and while each has incrementally improved on matters, it’s not been enough to dislodge a highly dominant Intel. And while the lack of success of Windows-on-Arm is far from solely being Qualcomm’s fault – there’s a lot to be said for the OS and software – silicon has certainly played a part. To make serious inroads on the market, it’s not enough to produce incrementally better chips – Qualcomm needs to make a major leap in performance. Now, after nearly three years of hard work, Qualcomm is getting ready to do just that. This morning, the company is previewing their upcoming Snapdragon X Elite SoC, their next-generation Arm SoC designed for Windows devices. Based on a brand-new Arm CPU core design from their Nuvia subsidiary dubbed “Oryon”, the Snapdragon X Elite is to be the tip of the iceberg for a new generation of Qualcom SoC designs. Not only is it the heart and soul of Qualcomm’s most important Windows-on-Arm SoC to date, but it will eventually be in smartphones and a whole lot more. But we’re getting ahead of ourselves. For now let’s focus on the Snapdragon X Elite SoC and the Oryon cores underpinning it. Some more in-depth information about Qualcomm’s upcoming Snapdragon X Elite, this time from AnandTech.
The device looks like a conventional computer monitor but opens up like a clam. The screen itself is a common flat panel liquid crystal display or LCD, a nearly translucent screen that is typically lit from behind by powered lights. For Eazeye, the backing lights are replaced with a bright white carbon fiber panel that can tip backwards up to 45 degrees. The panel bounces ambient light from the monitor’s surroundings through the LCD screen, which, under the right lighting conditions, provides enough illumination for the screen to be used like normal. I can see this working quite well in certain environments, like offices and well-lit rooms. It sure is a very interesting idea, and I like the design, too.
For years, Qualcomm has been making Snapdragon chips for Windows PCs, and for years, those chips’ performance have failed to dislodge Intel’s or AMD’s chips to any significant degree. Its latest Snapdragon 8cx Gen 3 (and the closely related Microsoft SQ3) appears in just two consumer PCs, the cumbersomely named Microsoft Surface Pro 9 with 5G and Lenovo’s ThinkPad X13s Gen 1. But that may be changing. Nearly three years ago, Qualcomm bought a company called Nuvia for $1.4 billion. Nuvia was mainly working on server processors, but the company’s founders and many of its employees had also been involved in developing the A- and M-series Apple Silicon processors that have all enabled the iPhone, iPad, and Mac to achieve their enviable blend of performance and battery life. Today, Qualcomm is formally announcing the fruit of the Nuvia acquisition: the Qualcomm Snapdragon X Elite is a 12-core, 4 nm chip that will compete directly with Intel’s Core processors and AMD Ryzen chips in PCs—and, less directly, Apple’s M2 and M3-series processors for Macs. We’ve heard a lot of these claims over the years, and to be honest, I’m a little tired of promises. Show me the goods. Apple did.
Nvidia and AMD could sell PC chips as soon as 2025, one of the people familiar with the matter said. Nvidia and AMD would join Qualcomm, which has been making Arm-based chips for laptops since 2016. At an event on Tuesday that will be attended by Microsoft executives, including vice president of Windows and Devices Pavan Davuluri, Qualcomm plans to reveal more details about a flagship chip that a team of ex-Apple engineers designed, according to a person familiar with the matter. Nvidia is such a natural partner for Microsoft when it comes to ARM chips, I’m surprised it’s taking them this long to jump back into the ring after the failed Surface RT. AMD making ARM chips is fascinating and surprising, though, but I guess they don’t feel they can compete on performance-per-watt with x86.
TalosSpace has more details on the upcoming, recently announced OpenPOWER machines from Raptor. I asked Timothy Pearson at Raptor about the S1’s specs, and he said it’s a PCIe 5.0 DDR5 part running from the high 3GHz to low 4GHz clock range, with the exact frequency range to be determined. (OMI-based RAM not required!) The S1 is bi-endian, SMT-4 and will support at least two sockets with an 18-core option confirmed for certain and others to be evaluated. This compares very well with the Power10, which is also PCIe 5.0, also available as SMT-4 (though it has an SMT-8 option), and also clocks somewhere between 3.5GHz and 4GHz. S1 embeds its own BMC, the X1 (or variant), which is (like Arctic Tern) a Microwatt-based ISA 3.1 core in Lattice ECP5 and iCE40 FPGAs with 512MB of DDR3 RAM, similar to the existing ASpeed BMC on current systems. X1 will in turn replace the existing Lattice-based FPGA in Arctic Tern as “Antarctic Tern,” being a functional descendant of the same hardware, and should fill the same roles as a BMC upgrade for existing Raptor systems as well as the future BMC for the next generation systems and a platform in its own right. The X1 has “integrated 100% open root of trust” as you would expect for such a system-critical part. This all sounds like exactly the kind of things I wanted to hear, and these details make me sufficiently excited about the near future of Raptor’s OpenPOWER workstations. The only little bit of less pleasant news is that the machines won’t be available until late 2024, so we’ve got a little wait ahead of us.
Well, this is a pleasant surprise and a massive coincidence. Besides that BMC-focused press release, Raptor Computing Systems tweeted out that they are working on “next generation of high performance, fully owner controlled systems! Built using the open POWER ISA 3.1, these new machines will be direct upgrades for existing POWER9 systems.” Power ISA 3.1 aligns with new functionality IBM introduced in Power10. This is fantastic news, and it seems they’re sidestepping the IBM POWER10 binary blobs issue by relying on a different chip vendor altogether, Solid Silicon, who announced an OpenPOWER CPU that will be used in Raptor’s upcoming systems, the S1. It seems unlikely to me that the S1 will be an entirely new, unique processor, so perhaps it’s a slightly modified IBM POWER10 design without the binary blobs. I’m incredibly excited about this news, and can’t wait to hear what they’re planning.
It doesn’t have a name yet, but Qualcomm says it’s developing a “RISC-V Snapdragon Wear” chip in collaboration with Google. The company says it plans to “commercialize the RISC-V based wearables solution globally including the US.” For Google and Qualcomm, this chip represents everyone’s first swing at a commercial RISC-V Android project, and as far as we can tell, it’s the first announced mass-market RISC-V Android chip ever. Qualcomm says the groundwork it and Google lay out “will help pave the way for more products within the Android ecosystem to take advantage of custom CPUs that are low power and high performance.” This is the biggest endorsement of RISC-V yet, and could catapult the platform to mainstream popularity pretty quickly. I do hope Qualcomm isn’t going to wrap their chip in a load of proprietary nonsense making them needlessly complex to support now and in the future, but I won’t be surprised if that hope turns out to be futile.
System76, the leading US-based Linux computer and keyboard manufacturer, made several new changes to their desktop line in order to optimize AI workloads and other fields reliant on heavy component use. The main focus: An airflow optimization that prevents throttling, putting their desktops at the top of performance charts. A new starter desktop, Thelio Spark, will also debut as a productivity desktop for everyday users. System76 seems to have redesigned the thermal solution on the machines, and judging by the various photos I’ve seen on Mastodon, they look good. System76 also sells the cases for the Thelio separately – they’re slightly different, though – and the company is sending me that case for review, and I’m curious to finally take a closer look. The Thelio Spark, the new kid in town, brings the Thelio line to a more affordable audience, with more affordable specifications. Of course, you’re always going to be paying a prebuilt tax, as well as the custom case tax, but if you want to ensure a plug-and-play Linux experience that isn’t just parts in a random case you can get anywhere else, there aren’t a lot of other options in the market.
Cold War–era computing has a poor reputation. The picture is one of a landscape littered with uninspired attempts to copy American IBM PCs, British ZX Spectrums, and other Western computers. But then there was Yugoslavia’s Galaksija, a very inspired bid to put a computer into the hands of regular comrades. The Galaksija is a Z80-based, 8-bit DIY machine, cleverly designed so that its bill of materials meshed exactly with what a Yugoslavian was able to import from Western Europe. During its brief heyday, thousands were built, leading to commercially assembled Galaksijas finding their way into homes and schools across the country. And now you can try this scrappy machine for yourself. There’s a huge world of computing to discover in former USSR countries, former USSR satellite states, and other countries that delicately straddled the west and east such as former Yugoslavia, many of which most people in the west have never heard of. While many of them may not have been competitive with what the Americans and Europeans were building, that doesn’t mean they’re not interesting or that there’s nothing to learn from the approaches the engineers took.
Most MiniDisc aficionados are aware of unit hacking to gain access to new features. The unit that perhaps benefits the most from this is the Sony MZ-N510, which also comes in the N520 and NF610 variants. The 2001 model R700 can be hacked to add many features of its upscale brother, the R900, as well as the Type-R codec, which renders the R700 capable of performing real-time SP recordings with Sony’s last evolution of ATRAC1. I bet the market for hacking the best music format of all time is small these days, but this is still incredibly cool.
Lenovo is forecasting that the vast majority of its devices will be repairable by 2025 – as will the repair parts themselves – but it is not intending to specify where customers should have their kit fixed. “On repairability, we have a plan that by 2025 more than 80 percent of the repair parts will be repaired again so that they they enter into the circular economy to reduce the impact to the environment.” He added: “More than 80 percent of our devices will be able to be repaired at the customer, by the customer or by the channel and we are enabling this with a design for serviceability kind of approach.” That’s excellent news, and I hope it’s a promise they’ll keep. The right to repair movement is scoring win after win lately, and it seems the tide has really turned on this one. It’s not just nerds anymore – regular people, common media, and even larger companies are beating the drum now.
The X220 ThinkPad is the greatest laptop ever made and you’re wrong if you think otherwise. No laptop hardware has since surpassed the nearly perfect build of the X220. New devices continue to get thinner and more fragile. Useful ports are constantly discarded for the sake of “design”. Functionality is no longer important to manufacturers. Repairability is purposefully removed to prevent users from truly “owing” their hardware. It’s a mess out there. But thank goodness I still have my older, second-hand X220. I don’t agree with the author, but he’s also not wrong. Luckily, things do seem to be improving somewhat, thanks to Framework being a decent success. Other OEMs are starting to make some noise about repairability, as are lawmakers around the world. We might be getting a new X220.
Today, we’re delighted to announce the launch of Raspberry Pi 5, coming at the end of October. Priced at $60 for the 4GB variant, and $80 for its 8GB sibling (plus your local taxes), virtually every aspect of the platform has been upgraded, delivering a no-compromises user experience. Raspberry Pi 5 comes with new features, it’s over twice as fast as its predecessor, and it’s the first Raspberry Pi computer to feature silicon designed in‑house here in Cambridge, UK. While I personally think there are more interesting alternatives to the Pi, there’s no doubt the Pi is the most compatible and most popular of these small board computers, and a big upgrade like this is definitely welcome – assuming they can actually stock these at fair prices at the end of October, when the fifth iteration of the Pi actually launches.
