History Of The SPARC CPU Architecture

March 28, 2023 by 31 Comments

[RetroBytes] nicely presents the curious history of the SPARC processor architecture. SPARC, short for Scalable Processor Architecture, defined some of the most commercially successful RISC processors during the 1980s and 1990s. SPARC was initially developed by Sun Microsystems, which most of us associate the SPARC but while most computer architectures are controlled by a single company, SPARC was championed by dozens of players.  The history of SPARC is not simply the history of Sun.

A Reduced Instruction Set Computer (RISC) design is based on an Instruction Set Architecture (ISA) that runs a limited number of simpler instructions than a Complex Instruction Set Computer (CISC) based on an ISA that comprises more, and more complex, instructions. With RISC leveraging simpler instructions, it generally requires a longer sequence of those simple instructions to complete the same task as fewer complex instructions in a CISC computer. The trade-off being the simple (more efficient) RISC instructions are usually run faster (at a higher clock rate) and in a highly pipelined fashion. Our overview of the modern ISA battles presents how the days of CISC are essentially over. Continue reading “History Of The SPARC CPU Architecture” →

Europe’s Proposed Right-To-Repair Law: A Game Changer, Or Business As Usual?

March 27, 2023 by 145 Comments

Recently, the European Commission (EC) adopted a new proposal intended to enable and promote the repair of a range of consumer goods, including household devices like vacuum cleaners and washing machines, as well as electronic devices such as smartphones and televisions. Depending on how the European Parliament and Council vote in the next steps, this proposal may shape many details of how devices we regularly interact with work, and how they can be repaired when they no longer do.

As we have seen recently with the Digital Fair Repair Act in New York, which was signed into law last year, the devil is as always in the details. In the case of the New York bill, the original intent of enabling low-level repairs on defective devices got hamstrung by added exceptions and loopholes that essentially meant that entire industries and types of repairs were excluded. Another example of ‘right to repair’ being essentially gamed involves Apple’s much-maligned ‘self repair’ program, that is both limited and expensive.

So what are the chances that the EU will succeed where the US has not?

Continue reading “Europe’s Proposed Right-To-Repair Law: A Game Changer, Or Business As Usual?” →

Who Needs Gasoline When You’ve Got Sodium?

March 23, 2023 by 58 Comments

YouTuber and serial debunker [Thunderf00t] was thinking about the use of sodium to counteract global warming. The theory is that sodium can be used as a fuel when combusted with air, producing a cloud of sodium hydroxide which apparently can have a cooling effect if enough of it is kicking around the upper atmosphere. The idea is to either use sodium directly as a fuel, or as a fuel additive, to increase the aerosol content of vehicle emissions and maybe reduce their impact a little.

One slight complication to using sodium as a fuel is that it’s solid at room temperature, so it would need to be either delivered as pellets or in liquid form. That’s not a major hurdle as the melting point is a smidge below 100 degrees Celsius and well within the operating region of an internal combustion engine, but you can imagine the impact of metal solidifying in your fuel system. Luckily, just like with solder eutectic mixes, sodium-potassium alloy happens to remain in liquid form at handleable temperatures and only has a slight tendency to spontaneously ignite. So that’s good.

Initial experiments using ultrasonic evaporators proved somewhat unsuccessful due to the alloy’s electrical conductivity and tendency to set everything on fire. The next attempt was using a standard automotive fuel injector from the petrol version of the Ford Fiesta. Using a suitable container, a three-way valve to allow the introduction of fuels, and an inert argon feed (preventing spontaneous combustion in the air), delivering the liquid metal fuel into the fuel injector seems straightforward enough.

[Thunderf00t] started with ethanol, then worked up to pentane before finally attempting to use the feisty sodium-potassium, once the bugs had been shaken out of the high-speed video setup. [Thunderf00t] does stress the importance of materials selection when handling this potential liquid metal fuel, since it apparently just bursts into flames in a violent manner on contact with incompatible materials. Heck, this stuff even reacts with PTFE, which is generally considered a very resistant material. We’re totally convinced we’d not like to see this stuff being pumped from a roadside gas station, at all, but it sure is a fun concept to think about.

Sodium-Potassium alloy doesn’t feature on these pages too often, but here’s a little fountain of the stuff, just because why not?

Continue reading “Who Needs Gasoline When You’ve Got Sodium?” →

A freshly reballed BGA chip next to a clean PCB footprint

Working With BGAs: Soldering, Reballing, And Rework

March 23, 2023 by 31 Comments

In our previous article on Ball Grid Arrays (BGAs), we explored how to design circuit boards and how to route the signals coming out of a BGA package. But designing a board is one thing – soldering those chips onto the board is quite another. If you’ve got some experience with SMD soldering, you’ll find that any SOIC, TQFP or even QFN package can be soldered with a fine-tipped iron and a bit of practice. Not so for BGAs: we’ll need to bring out some specialized tools to solder them correctly. Today, we’ll explore how to get those chips on our board, and how to take them off again, without spending a fortune on equipment.

Tools of the Trade

For large-scale production, whether for BGA-based designs or any other kind of SMD work, reflow ovens are the tool of choice. While you can buy reflow ovens small enough to place in your workshop (or even build them yourself), they will always take up quite a bit of space. Reflow ovens are great for small-scale series production, but not so much for repairs or rework. Continue reading “Working With BGAs: Soldering, Reballing, And Rework” →

Combining Acoustic Bioprinting With Raman Spectroscopy For High-Throughput Identification Of Bacteria

March 22, 2023 by 6 Comments

Rapidly analyzing samples for the presence of bacteria and similar organic structures is generally quite a time-intensive process, with often the requirement of a cell culture being developed. Proposed by Fareeha Safir and colleagues in Nano Letters is a method to use an acoustic droplet printer combined with Raman spectroscopy. Advantages of this method are a high throughput, which could make analysis of samples at sewage installations, hospitals and laboratories significantly faster.

Raman spectroscopy works on the principle of Raman scattering, which is the inelastic scattering of photons by matter, causing a distinct pattern in the thus scattered light. By starting with a pure light source (that is, a laser), the relatively weak Raman scattering can be captured and the laser light filtered out. The thus captured signal can be analyzed and matched with known pathogens. Continue reading “Combining Acoustic Bioprinting With Raman Spectroscopy For High-Throughput Identification Of Bacteria” →

One Method For Removing Future Space Junk

March 22, 2023 by 18 Comments

When sending satellites into space, the idea is to place them into as stable an orbit as possible in order to maximize both the time the satellite is useful and the economics of sending it there in the first place. This tends to become rather untenable as the amount of space junk continues to pile up for all but the lowest of orbits, but a team at Brown University recently tested a satellite that might help solve this problem, at least for future satellite deployments.

The main test of this satellite was its drag sail, which increases its atmospheric drag significantly and reduces its spaceflight time to around five years. This might make it seem like a problem from an economics standpoint, as it’s quite expensive to build satellites and launch them into space, but this satellite solves these problems by being both extremely small to minimize launch costs, and also by being built out of off-the-shelf components not typically rated for spaceflight. For example, it gets its power solely from AA batteries and uses an Arduino for its operation and other research.

The satellite is currently in orbit, and has already descended from an altitude of 520 km to 470 km. While it won’t help reduce the existing amount of debris in orbit, the research team hopes to demonstrate that small satellites can be affordable and economically feasible without further contributing to the growing problem of space junk. If you’re looking to launch your own CubeSat one day, take a look at this primer which goes over most of the basics.

Why LLaMa Is A Big Deal

March 22, 2023 by 27 Comments

You might have heard about LLaMa or maybe you haven’t. Either way, what’s the big deal? It’s just some AI thing. In a nutshell, LLaMa is important because it allows you to run large language models (LLM) like GPT-3 on commodity hardware. In many ways, this is a bit like Stable Diffusion, which similarly allowed normal folks to run image generation models on their own hardware with access to the underlying source code. We’ve discussed why Stable Diffusion matters and even talked about how it works.

LLaMa is a transformer language model from Facebook/Meta research, which is a collection of large models from 7 billion to 65 billion parameters trained on publicly available datasets. Their research paper showed that the 13B version outperformed GPT-3 in most benchmarks and LLama-65B is right up there with the best of them. LLaMa was unique as inference could be run on a single GPU due to some optimizations made to the transformer itself and the model being about 10x smaller. While Meta recommended that users have at least 10 GB of VRAM to run inference on the larger models, that’s a huge step from the 80 GB A100 cards that often run these models.

While this was an important step forward for the research community, it became a huge one for the hacker community when [Georgi Gerganov] rolled in. He released llama.cpp on GitHub, which runs the inference of a LLaMa model with 4-bit quantization. His code was focused on running LLaMa-7B on your Macbook, but we’ve seen versions running on smartphones and Raspberry Pis. There’s even a version written in Rust! A rough rule of thumb is anything with more than 4 GB of RAM can run LLaMa. Model weights are available through Meta with some rather strict terms, but they’ve been leaked online and can be found even in a pull request on the GitHub repo itself. Continue reading “Why LLaMa Is A Big Deal” →