Hardware Archive

SoftBank could sell or spin off Arm way sooner than it planned

SoftBank is reportedly assessing spin-off options for its semiconductor firm, Arm Holdings. The Wall Street Journal reports from its sources that those options include having an initial public offering or a sale. The Japanese tech conglomerate picked up Arm back in 2016 for $32 billion and currently shares some ownership with investors in the SoftBank Vision Fund. The moves are being considered as SoftBank fends off challenges from activist investment house Elliott Management over major losses for its Vision Fund, including WeWork’s attempted IPO. SoftBank is supposedly targeting $41 billion in immediate fundraising through share buybacks and divestitures. Depending on Arm’s current prospectus, a sale could be more likely to happen than an IPO. I’m linking to the AndroidPolice item since the original article is stuck behind a paywall. Whoever intends to acquire ARM better have a very good story to tell antitrust regulators, because I doubt Intel, Apple, Google, or any of the other major technology companies will be allowed to acquire it. I wonder who else could be a potential buyer – maybe another investment fund?

Review: System76’s Lemur Pro

If you’re a Linux user on the hunt for a new laptop, there’s quite a bit of preparation and research you must do on top of the regular research buying such an expensive piece of equipment already entails. Reading forum posts from other Linux users with the laptop you’re interested in, hunting for detailed specifications to make sure that specific chip version or that exact piece of exotic hardware is fully supported, checking to see if your favourite distribution has adequate support for it, and so on. There is, however, another way. While vastly outnumbered, there are laptops sold with Linux preinstalled. Even some of the big manufacturers, such as Dell, sell laptops with Linux preinstalled, but often only on older models that have been out for a while, or while not fully supporting all hardware (the fingerprint reader and infrared camera on my XPS 13 were not supported by Linux, for instance). For the likes of Dell, Linux in the consumer space is an afterthought, a minor diversion, and it shows. If you want the best possible out-of-the-box Linux experience, you’ll have to go to one of the smaller, more boutique Linux-only OEMs. One of the more prominent Linux OEMs is System76, who have been selling various laptops and desktops with Linux preinstalled for more than decade now. Recently, they launched their new ultraportable, the Lemur Pro, and they kindly loaned one to us for review. Full disclosure: System76 sent us the laptop as a loan, and it will be returned to them after publication of this review. They did not read this review before publication, and placed zero restrictions on anything I could write about. Specifications The Lemur Pro configuration System76 sent to us comes in at $1492, and packs a 4C/8T 10th Gen Intel Core i7-10510U, with frequencies of 1.8 up to 4.9 GHz and 8MB Cache. It came with 16GB of RAM, of which 8 is soldered onto the motherboard, and 8 is seated in the single RAM expansion slot. Storage-wise, it is equipped with a 500GB SSD in one of its two user-accessible M.2 slots – a Samsung 970 Evo Plus. The 14.1″ display has a resolution of 1920×1080 with a matte finish, with a maximum refresh rate of 60Hz. The display is powered by the integrated GPU, and there’s no option for a discrete GPU. The battery is a 73 Wh unit, and is entirely user-replacable. Bucking a trend in the industry, the Lemur Pro is reasonably equipped when it comes to ports: one USB 3.1 Type-C Gen 2 port, two USB 3.0 Type-A ports, a MicroSD Card Reader, a full-size HDMI port, a barrel connector for the included charger (USB-C charging is also supported), a combined headphone/microphone jack, and the usual Kensington lock. The USB-C port can also be used as a display port with DisplayPort 1.2. Hardware The design of the Lemur Pro is unassuming, mostly black, and free of the kind of design frivolities other laptops tend to suffer from. There’s no RGB here, no flames painted on the lid to make it go faster, no screaming logos or gamer accents – just a black laptop with a System76 logo on the lid. That’s it. It is incredibly light, weighing a mere 0.99 kg – for comparison, a MacBook Air weighs 1.29 kg, so the Lemur Pro is considerably lighter. This does come at a price, however, and the Lemur Pro just doesn’t feel as strong and sturdy as similar laptops with a bit more heft to it. There’s an amount of flex in the display lid, bottom cover, and keyboard cover that you just won’t see in a MacBook Air or an XPS 13. It’s a trade-off you have to make – if you really value the extreme kind of portability the Lemur Pro provides, it means giving in somewhere else. I’m disappointed System76 does not provide a high refresh rate display on the Lemur Pro, in the very least as an option. Once you’ve gotten used to 144Hz (or even higher) on your computer displays, using a 60Hz display feels like a major step back. I understand the battery life concerns, but I’m definitely more than willing to give up a little bit of battery life if it meant a buttery-smooth 144Hz UI. Aside from the lack of a high refresh rate option, the display is excellent – it’s bright and the colours look normal, but note that I’m not a colour expert, so I can’t make any claims about colour accuracy. For my general use, however, I didn’t run into any issues. Speaking of battery life – this is one of the major strong points of the Lemur Pro. System76 advertises a maximum battery life of 14 hours, and while these kind of figures are usually complete nonsense, I think they’re not far off the mark here. Since we do not (yet) have a long history of laptop reviews, we do not have any consistent methodology for measuring battery life, so anything I say here is very subjective and difficult for you as a reader to parse. That being said, with casual use – meaning, browsing, writing, Twitter and e-mailing while watching YouTube videos – I could definitely hit the 10 hour mark at the balanced power setting. Switching to the power saver setting yielded me even more hours of battery life, but it did cause a notable hit in performance, especially for video. Simple 1080p YouTube video – either played in Firefox or locally – would stutter and lag, but everything else seemed to perform just fine. My guess is that the power saver setting targets the integrated Intel GPU quite aggressively, but honestly, for several hours of additional battery life, I think it’s worth it. The battery life is especially remarkable since getting proper battery life out of laptops designed for Windows running Linux is often a major hassle, and no matter what you do, Linux battery life on laptops not designed for Linux always lags

Ampere’s product list: 80 cores, up to 3.3 GHz at 250 W; 128 core in Q4

With the advent of higher performance Arm based cloud computing, a lot of focus is being put on what the various competitors can do in this space. We’ve covered Ampere Computing’s previous eMag products, which actually came from the acquisition of Applied Micro, but the next generation hardware is called Altra, and after a few months of teasing some high performance compute, the company is finally announcing its product list, as well as an upcoming product due for sampling this year. Ampere’s Altra is a realized version of Arm’s Neoverse N1 enterprise core, much like Amazon’s Graviton2, but this time in an 80-core arrangement. Where Graviton2 is designed to suit Amazon’s needs for Arm-based instances, Ampere’s goal is essentially to supply a better-than-Graviton2 solution to the rest of the big cloud service providers (CSPs). Of the companies that have committed to an N1 based design, so far on paper Ampere is publically the biggest and fastest on the books. Can we have these in workstations please? I know they’re not designed for my kinds of uses, but damn if these aren’t awesome.

The Open Book Project

As a society, we need an open source device for reading. Books are among the most important documents of our culture, yet the most popular and widespread devices we have for reading — the Kobo, the Nook, the Kindle and even the iPad — are closed devices, operating as small moving parts in a set of giant closed platforms whose owners’ interests are not always aligned with readers’. The Open Book aims to be a simple device that anyone with a soldering iron can build for themselves. The Open Book should be comprehensible: the reader should be able to look at it and understand, at least in broad strokes, how it works. It should be extensible, so that a reader with different needs can write code and add accessories that make the book work for them. It should be global, supporting readers of books in all the languages of the world. Most of all, it should be open, so that anyone can take this design as a starting point and use it to build a better book. Whenever someone asks what “putting your money where your mouth is” means, just link them to the Open Book.

Buyer beware – that 2TB-6TB “NAS” drive you’ve been eyeing might be SMR

Storage vendors, including but reportedly not limited to Western Digital, have quietly begun shipping SMR (Shingled Magnetic Recording) disks in place of earlier CMR (Conventional Magnetic Recording) disks. SMR is a technology that allows vendors to eke out higher storage densities, netting more TB capacity on the same number of platters—or fewer platters, for the same amount of TB. Until recently, the technology has only been seen in very large disks, which were typically clearly marked as “archival”. In addition to higher capacities, SMR is associated with much lower random I/O performance than CMR disks offer. This is going to be another one of those stupid things us technology buyers have to look out for when buying storage, isn’t it? Like

Qemu 5.0.0 released

Qemu 5.0.0 has been released, with a massive laundry list of changes, fixes, and improvements for a lot of Qemu’s emulated platforms. The new version will make it your operating system’s repositories soon enough if you use Linux, but if you use a platform where you have to muddle along with and juggle your applications and updates manually like a peasant, like Windows or macOS, you’ll have to wait until someone packages it for you so you can update your binary manually. Of course, you can always build it yourself, too.

ARM development for the office: unboxing an Ampere eMag workstation

One of the key elements I’ve always found frustrating with basic software development is that it can often be quite difficult to actually get the hardware in hand you want to optimize for, and get a physical interaction that isn’t delayed by networking or logging in or anything else. Having a development platform on the desk guarantees that direct access, and for the non-x86 vendors, I’ve been asking for these for some time. Thankfully we’re now starting to see some appear, and Avantek, one of the Arm server retailers, have built an Ampere eMag workstation out of a server board, with some interesting trickery to get it to fit. They sent us one to have a look at. This is only the unboxing and short first impressions, but I am unreasonably excited about what are effectively bog-standard PCs, but with an ARM processor. I can’t wait for these machines to come down in price, because this is the first time in a long, long time that we’ve seen what could become a serious challenge to x86 in its traditional space: desktops and laptops. Once AnandTech publishes its actual review, I’ll be on top of that, too.

Is MIPS dead? Lawsuit, bankruptcy, maintainers leaving and more

But in recent months, MIPS related news has not been so good. First, Wave Computing decided to end MIPS Open Initiative in November 2019, then Paul Burton and Ralf Baechle removed themselves from the Linux kernel MIPS maintainer list in February 2020, as their work with MIPS ended leaving Thomas Bogendoerfer as the only maintainer. But this month, things turned for the worse, with CIP United filing a lawsuit against MIPS LLC on April 6, before announcing the company had obtained exclusive MIPS license rights for mainland China, Hong Kong, and Macau on April 11, 2019 (Click on News in top menu of this link). In my mind, I had written off MIPS as a thing of the past anyway – but it’s still sad to see if tangled up in legal disputes and lack of attention.

Marvell cranks up cores and clocks with “Triton” ThunderX3

Arm server chip upstart Ampere Computing made a big splash with its 80-core “Quicksilver” Altra processor two weeks ago, and Marvell, which is the volume leader in Arm server chips with its “Vulcan” ThunderX2 processors (largely inherited from its acquisition of Broadcom’s Arm server chip assets), is hitting back with some revelations about its future “Triton” ThunderX3 chip and its roadmap out beyond that. Competition in the ARM server space is really heating up.

Looking inside a vintage Soviet TTL logic integrated circuit

This blog post examines a 1980s chip used in a Soyuz space clock. The microscope photo below shows the tiny silicon die inside the package, with a nice, geometric layout. The silicon appears pinkish or purplish in this photo, while the metal wiring layer on top is white. Around the edge of the chip, the bond wires (black) connect pads on the chip to the chip’s pins. The tiny structures on the chip are resistors and transistors. That’s just cool.

Ampere Altra launched with 80 Arm cores for the cloud

Today we have the launch of the Ampere Altra Arm CPUs. This is a completely new design built specifically for cloud providers. It has up to 80 cores and is designed to go head-to-head with AMD EPYC 7002 “Rome” series processors as well as 2nd Gen Intel Xeon Scalable Refresh parts. Ampere is launching today but we do not have a test system as we have had for all of the other major server chip launches recently. It sounds like Ampere is shipping mass production units soon, but not at the time we were briefed. Hopefully, we can get more hands-on insights in the near future. In this article, we are going to discuss the architecture based on the documents we have and our discussion with the company. We are going to check performance claims and help our readers critically analyze what they are being shown. We are then going to discuss systems disclosed with the chips before getting to our final thoughts. This seems like an impressive piece of engineering, but we’ll definitely need test systems and proper reviews to test the claims about the chip. While I’m very happy AMD is back in the game and we’ve got some real competition to Intel, we as consumers would benefit even more from proper ISA competition, something we haven’t seen in a long time.

Open source hardware: the rise of RISC-V

While open source software is taking over the world, a push for open source hardware has been quietly building. The RISC-V Foundation has been pushing its open sourced instruction set architecture for chips based on the long-established paradigms for reduced instruction set computing. And one of its most vocal advocates is Calista Redmond, the chief executive of the RISC-V Foundation, which is working to promote its adoption. This is a slow burn. RISC-V won’t change the world overnight, but will slowly but surely seep into every corner of the computing industry – and looking at the incompatible, closed-source mess that is the ARM world, we really need RISC-V on all those millions of embedded devices we use every day.

What made the 1960s CDC6600 supercomputer fast?

Besides the architectural progress, the CDC6600 was impressive for its clock speed of 10 MHz. This may not sound much, but consider that this was a physically very large machine entirely built from discrete resistors and transistors in the early 60ies. Not a single integrated circuit was involved. For comparison, the PDP-8, released in 1965 and also based on discrete logic, had a clock speed of 1.5 MHz. The first IBM PC, released 20 years later, was clocked at less than half the speed of the CDC6600 despite being based on integrated circuits. The high clockrate is even more impressive when comparing it to more recent (hobbyist) attempts to design CPUs with discrete components such as the MT15, the Megaprocessor or the Monster6502. Although these are comparatively small designs based on modern components, none of them get to even a tenth of the CDC6600 clock speed. Detailed look at the speed of the CDC6600.

USB-IF announces publication of USB4 specification

USB Implementers Forum, the support organization for the advancement and adoption of USB technology, today announced the publication of the USB4 specification, a major update to deliver the next-generation USB architecture that complements and builds upon the existing USB 3.2 and USB 2.0 architectures. The USB4 architecture is based on the Thunderbolt protocol specification recently contributed by Intel Corporation to the USB Promoter Group. It doubles the maximum aggregate bandwidth of USB and enables multiple simultaneous data and display protocols. As the USB Type-C connector has evolved into the role as the external display port of many host products, the USB4 specification provides the host the ability to optimally scale allocations for display data flow. Even as the USB4 specification introduces a new underlying protocol, compatibility with existing USB 3.2, USB 2.0 and Thunderbolt 3 hosts and devices is supported; the resulting connection scales to the best mutual capability of the devices being connected. How many years until USB4 (or later) replaces HDMI and DisplayPort? Since everything is data packets now – analog is a thing of the past – do we really need to have separate video cables? I’d love to one day build a PC that just has an array of USB-C ports on the back, with the ability to plug anything – monitor, keyboard, mouse, serial port adapter for the Windows CE mini laptop I’ve been craving to buy for ages, you know, the usual stuff – into any of the ports. One can dream.

Opening up the Baseboard Management Controller

In 2011 Facebook announced the Open Compute Project to form a community around open-source designs and specifications for data center hardware. Facebook shared its hardware specs, which resulted in 38 percent less energy consumption and 24 percent cost savings compared with its existing data centers. What Facebook and other hyperscalers (Google, Microsoft, et al.) donate to the Open Compute Project are their solutions to the agonizing problems that come with running data centers at scale. Since then, the project has expanded to all aspects of the open data center: baseboard management controllers (BMCs), network interface controllers (NICs), rack designs, power busbars, servers, storage, firmware, and security. This column focuses on the BMC. This is an introduction to a complicated topic; some sections just touch the surface, but the intention is to provide a full picture of the world of the open-source BMC ecosystem, starting with a brief overview of the BMC’s role in a system, touching on security concerns around the BMC, and then diving into some of the projects that have developed in the open-source ecosystem. A good overview.

A look at Celerity’s second generation 496-core RISC-V Mesh NoC

For those who are not familiar with Celerity, this is a multi-university effort that has resulted in an open-source manycore RISC-V tiered accelerator chip. The project is part of the DARPA Circuit Realization At Faster Timescales (CRAFT) program which wants to drive the design cycle for custom integrated circuits to weeks and months from years. The Celerity team first presented the chip at Hot Chips 29. Last year, at VLSI 2019, Celerity was back to talk about the PLL and the NoC of its second-generation chip. The presentation was given by Austin Rovinski from the University of Michigan. I can read the words, but much of this is far too complicated for me to give any meaningful comment.

Farmers are buying 40-year-old tractors because they’re actually repairable

When a brand new John Deere tractors breaks down, you need a computer to fix it. When a John Deere tractor manufactured in 1979 breaks down, you can repair it yourself or buy another old John Deere tractor. Farming equipment—like televisions, cars, and even toothbrushes—now often comes saddled with a computer. That computer often comes with digital rights management software that can make simple repairs an expensive pain in the ass. As reported by the Minnesota StarTribune, Farmers have figured out a way around the problem—buying tractors manufactured 40 years ago, before the computers took over. I wonder if we’ll ever reach that state with computers – a point where they become so locked-down, unrepairable and impossible to fix that we will be forced to keep older hardware around just to retain control over our own devices.

RISC-V stumbling blocks

Recently, I’ve started to explore RISC-V. I experienced the journey as pretty refreshing, particularly because I’ve been working on x86 low-level software almost exclusively for about 10 years. In this post, I want to quickly go over some high-level stumbling blocks I noticed as I was starting out. I’m probably going to write about more technical differences in subsequent blog posts. While reading the rest of this post, please keep in mind: RISC-V is simple! If you managed to do any low-level coding on x86, you will find your way around RISC-V with little effort. It’s easy to forget how crazy some parts of x86 are. Just see my past posts. The more interest in RISC-V, the quicker we can expect a RISC-V laptop to run Linux on. I’m all for it.