Hardware Archive
Smartwatches at the turn of the century were a more motley assortment than today’s, with an even wilder range of functionality. If you had a few hundred dollars or so, there were some interesting options, even back then. But if all you had was $85 (in 2024 dollars about $150), you still weren’t left out, because in 2001 you could get the Web-@nywhere (the “Worldwide Web Watch”). Load up the software on your PC and slap it in its little docking station, and you could slurp down about 93K of precious Web data to scroll on the 59×16 screen — 10 characters by 2 characters — to read any time you wanted! That is, of course, if the remote host the watch’s Windows 9x-based client accessed were still up, on which it depended for virtually anything to download and install. Well, I want 95,488 bytes of old smartwatch tiny screen Web on my wrist, darn it. We’re going to reverse-engineer this sucker and write our own system using real live modern Web data. So there! ↫ Old Vintage Computing Research Y’all know the drill by now – I’m a sucker for these kinds of stories. What a great, extremely detailed read, with code to boot.
Roughly a year ago I moved into my new apartment. One of the reasons I picked this apartment was age of the building. The construction was finished in 2015, which ensured pretty good thermal isolation for winters as well as small nice things like Ethernet ports in each room. However, there was one part of my apartment that was too new and too smart for me. It is obviously a touchscreen of some sort, but there was zero indication as to what it controls. The landlord had no idea what this is. There are no buttons or labels on the thing, just a tiny yellow light to let you know it has the power. ↫ Nikita Lapkov What follows is an investigation into what it is, how to get it working, and, of course, how to hack it and make it more useful.
I always liked small laptops and phones – but for some reason they fell out of favor of manufacturers (“bigger is more better”). Now if one wanted to get tiny laptop – one of the few opportunities would have been to fight for old Sony UMPC’s on ebay which are somewhat expensive even today. Recently Raspberry Pi/CM4-based tiny laptops started to appear – especially clockwork products are neat, but they are not foldable like a laptop. When in summer of 2023 Sipeed announced Lichee Console 4A based on RISC-V SoC – I preordered it immediately and in early January I finally received it. Results of my testing, currently uncovered issues are below. ↫ Mikhail Svarichevsky I want one of these.
At CES 2024, ASUS unveiled a new standard for motherboards, graphics cards, and cases. Called BTF (short for Back-to-The-Future), it offers much cleaner cable management with power connectors at the back of a motherboard. More importantly, it fully ditches the ill-fated 12VHPWR plug in favor of a much tidier (and probably safer) 600W PCIe connector. ASUS claims computers with BTF components are easier to assemble since all plugs and connectors are located at the back side of the motherboard tray without other components obstructing access to power, SATA, USB, IO, and other connectors. Therefore, “you won’t have to reach as far into the depth of your chassis to plug things in.” BTF should also make cable management much more elegant, resulting in a tidy, showcase-ready build. ↫ Taras Buria at NeoWin The interior of PCs effectively hasn’t changed since the ’80s, and it feels like it, too. Many of the connectors and plugs are unwieldy, in terrible places, hard to connect/disconnect, difficult to route, and so on. A lot more needs to be done than putting the connectors on the back of the motherboard and integrating GPU power delivery into the PCIe slot, but even baby steps like these are downright revolutionary in the conservative, change-averse, anti-user world of PC building. I don’t say this very often, but basically, look at the last Intel Mac Pro. That’s what a modern PC should look and work like inside.
Confirming a previous leak, Lenovo officially announced the ThinkBook Plus Gen 5 Hybrid during its CES 2024 product reveals. It combines a Windows 11 notebook with a 14-inch OLED 2.8K touchscreen display that can detach from the keyboard and be used as an stand-alone Android 13 tablet. ↫ John Callaham I’m not even sure why I’m posting this, other than that it perfectly illustrates the problems Windows on one side, and Android on the other, face in providing the full device spectrum to users. Windows only really works on desktops and laptops, while Android only really works on smartphones and tablet. As such, Frankenstein devices like these have to be made to cover the entire spectrum. I kind of want one.
As I work on moss and research modern processor design patterns and techniques, I am also looking for patterns and techniques from the past that, for one reason or another, have not persisted into our modern machines. While on a run this week, I was listening to an old Oxide and Friends episode where Bryan, Adam, and crew were reminiscing on the SPARC instruction set architecture (ISA). SPARC is a reduced instruction set computer (RISC) architecture originally developed by Sun Microsystems, with the first machine, the SPARCstation1 (a.k.a. Sun 4/60, a.k.a Campus), being delivered in 1987. It was heavily influenced by the early RISC designs from David Patterson and team at Berkeley in the 1970s and 1980s, which is the same lineage from which RISC-V has evolved. Given the decision to base moss on the RISC-V RV64I ISA, I was interested to learn more about the history and finer details of SPARC. ↫ Daniel Mangum The sad thing is that SPARC is pretty close to dead at this point, with the two major players in the high-end – Oracle and Fujitsu – throwing in the towel half a decade ago. There’s some lower-end work, such as the LEON chips, but those efforts, too, seem to be going nowhere at the moment. Definitely sad, since I’ve always been oddly obsessed with the architecture, and hope to still somehow get my hands on the last UltraSPARC workstation ever built (the Sun Ultra 45, which is, sadly, incredibly expensive on the used market). There’s also a whole boatload of servers on the used market with fancier, newer SPARC processors, but as far as I know, none of those support any form of even barely usable graphics, making them useless for weird people like me who want to run a desktop on them.
Pastel cities trapped in a timeless future-past. Empty apartments drenched in nostalgia. Classic convertibles speeding into a low-res sunset. Femme fatales and mutated monsters doing battle. Deep, dark dungeons and glittering star ships floating in space. All captured in a eerie palette of 4096 colours and somehow, you’re sure, from some alternate 1980s world you can’t quite remember… ↫ Biz Davis The PC-98 is exotic, and a little bit mysterious. Of course, thanks to the internet, abundant emulation options, detailed YouTube videos, and more, all the information is out there – but I still find that the PC-98 carries with it an air of mystery.
I also had another set of addressable lights on my desk. While decorating my office for Christmas, I decided to invest some time in connecting them to Home Assistant using the BJ_LED code as a template. It should have been straightforward, right? Well, yes, but also no. ↫ Will Cooke We all love a good reverse-engineering story, especially if it involves bricking Christmas lights.
We’re going to cover the Cortex A57 as implemented in the Nintendo Switch’s Nvidia Tegra X1. The Tegra X1 targets a wide range of applications including mobile devices and automobiles. It focuses on providing high GPU performance in a limited power envelope, making it perfect for a portable gaming console like the Switch. Tegra X1 consumes 117,6 mm2 on TSMC’s 20 nm (20 SoC) process and uses a quad core A57 cluster to provide the bulk of its CPU power. Each Cortex A57 core consumes just under 2 mm2 of area, and the quad core A57 cluster takes 13.16 mm2. ↫ Clamchowder at Chips and Cheese An old SoC still doing excellent work in the Switch.
This is a website dedicated to a project of mine, Sol-1. Sol-1 is a homebrew CPU and Minicomputer built from 74HC logic. ↫ Paulo Constantino Sol-1 has user and kernel priviledge mode, a maximum of 256 processes in parallel, paged virtual memory, serial ports, parallel ports, IDE interface, realtime clock, a DMA channel, and much more. There’s also an accompanying operating system called Solarium.
When I started taking apart the Fairphone 5, I didn’t really expect any surprises. Having dis- and reassembled the previous model several times, I had some experience with Fairphone’s approach to building a smartphone: Modularity paired with easy access to all major components. It’s a winning formula for a repairable smartphone they have iterated on several times now. So, what’s actually different this time around—apart from a new and shiny OLED screen and beefed up cameras? ↫ Manuel Haeussermann for iFixit Spoiler: it’s still a 10/10 for repairability, but with new niceties to make the process even more pleasant.
HP knows people have grown to hate printers. It even knows that people hate HP printers. But based on a new marketing campaign the company launched, HP is OK with that—so long as it can convince people that there are worse options out there. The marketing campaign hitting parts of Europe aims to present HP as real and empathetic. The tagline “Made to be less hated” seems to acknowledge people’s frustration with printers. But HP’s a top proponent of the exact sort of money-grabbing, disruptive practices that have turned people against printers. ↫ Scharon Harding at Ars Technica I need to print something maybe a few times a year, and I still hate dealing with my printer more than any other tech item in my house. Everything about them is bad, and no cutesy marketing campaign centrered on them being bad is going to change that.
This ambition to escape dependence on foreign technology rests on the shoulders of Huawei and SMIC. The successful launch of the Kirin 9000S injected new vigor into the semiconductor industry, with executives reporting that chip start-ups are seeing a surge in funding. But Huawei’s long-term ambitions are not limited to the markets in China’s orbit. The original nickname for the Kirin 9000S—Charlotte—is a symbol of these hopes. It was named not for an individual, but for the city in North Carolina. Other mobile semiconductors in development are also named internally for US cities, insiders say. Using American names, says one Huawei employee, reflects “our desire to one day reclaim our place in the global supply chain.” It’s amazing how without any official support and using cobbled-together outdated lithography machines, Huawei and SMIC have managed to make a reasonably competitive smartphone SoC. As I keep saying – Chinese chip makers have the full financial might of the Chinese state behind them, and they’ll stop at nothing to reduce their dependence on ASML, TSMC, Intel, AMD, and so on. And they’re making progress.
Chinese chip designer Loongson has finally launched its loong teased “next-generation” 3A6000-series processors based on the LoongArch microarchitecture. IPC tests showed the 3A6000 matching Intel’s Raptor Lake i5-14600K in IPC (instructions per clock), with both chips clocked at 2.5GHz. As well as the headlining x86 compatible processor came the announcement of numerous partner desktop, laptop, and all-in-one machines — plus a consumer-grade motherboard from Asus. It was also entertaining to see a recorded overclocking session, which took an LN2-cooled 3A6000 chip to the current maximum 3 GHz. Many of us are being dismissive now, but give it a few more generations and Chinese PC users won’t be depending on Intel or AMD anymore – and that’s pretty impressive.
I got my hands on a Monoprice Blackbird 4K Pro HDCP 2.2 to 1.4 Converter. According to the marketing copy it “is the definitive solution for playback of new 4K HDCP 2.2 encoded content on 4K displays with the old HDCP 1.4 standard.” Stuffed after a delicious Thanksgiving meal, I decided to take it apart after the guests had left. It’s a simple single-function device, so I didn’t expect much, but maybe there’s some things to be learned? Turns out there’s a lot to learn, and it’s also incredibly interesting. The note at the end about the legality of this device is also interesting.
Frore Systems is a startup with $116 million in funding, and I’ve shown you its first product before: the AirJet Mini is a piezoelectric cooling chip that weighs just nine grams and is thinner than two US quarters stacked together. Each nominally consumes one watt and can remove 4.25 additional watts of heat. Here’s the question: what would happen if Frore used those AirJets to cool a laptop that normally doesn’t have a fan at all? What the company discovered — and I saw firsthand — is that Apple’s M2 chip can run faster, for longer, with Frore’s tech on board. Without it, a 15-inch M2 MacBook Air was like a runner that can’t sprint indefinitely without running out of breath. But with three AirJet Minis, the same laptop got a permanent second wind. Frore’s AirJet coolers have been featured on YouTube channels like LTT as well, and there’s no doubt in my mind these will be the future of laptop cooling, especially in the thinner segment of the laptop market. At least in thin laptops, AirJets are better in virtually every way than fans, and provide far superior cooling compared to fanless designs without adding bulk or noise. The only thing that sucks as an enthusiast is that you can’t really modify an existing laptop yourself. Either this company gets gobbled up by an OEM, or their products will make their way in almost every thin laptop.
Sunway’s new supercomputer therefore feels like a system designed with the goal of landing high on some TOP500 lists. For that purpose, it’s perfect, providing a lot of throughput without wasting money on pesky things like cache, out-of-order execution, and high bandwidth memory. But from the perspective of solving a nation’s problems, I feel like Sunway is chasing a metric. A nation doing well in advanced technology might have a lot of supercomputer throughput, but more supercomputer throughput doesn’t necessarily mean you’ll solve technological problems faster. A detailed look at China’s new supercomputer. The conclusion quoted above is very well supported by the data and research concerning this new supercomputer, and the article is a great read.
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.
