Some SPI Flash Chip Nuances Worth Learning

Some hackers have the skills to help us find noteworthy lessons in even the most basic of repairs. For instance, is your computer failing to boot? Guess what, it could just be a flash chip that’s to blame — and, there’s more you should know about such a failure mode. [Manawyrm] and [tSYS] over at the Kittenlabs blog show us a server motherboard fix involving a SPI flash chip replacement, and tell us every single detail we should know if we ever encounter such a case.

They got some Gigabyte MJ11-EC1 boards for cheap, and indeed, one of the BIOS chips simply failed — they show you how to figure that one out. Lesson one: after flashing a SPI chip, remember to read back the image and compare it to the one you just flashed into it! Now, you might be tempted to take any flash chip as a replacement, after all, many are command-compatible. Indeed, the duo crew harvested a SPI chip from an ESP32 board, the size matched, and surely, that’d suffice.

That’s another factor you should watch out for. Lesson two is to compare the SPI flash commands being used on the two chips you’re working with. In this case, the motherboard would read the BIOS alright and boot just fine, but wasn’t able to save the BIOS settings. Nothing you couldn’t fix by buying the exact chip needed and waiting for it to arrive, of course! SPI flash command sets are fun and worth learning about — after all, they could be the key to hacking your “smart” kettle. Need a 1.8 V level shifter while flashing? Remember, some resistors and a NPN transistor is more than enough.

A series of wooden rectangles are arranged vertically around the edges of a dark wooden base, reminiscent of a very tall radial fan. Light glows from the base up the slots between the vanes. a cord runs from behind the dark base to a small puck of the same color. The setup sits on a light grey table in front of a light grey wall.

A Beautiful Lamp-Inspired PC Case

Sometimes you see something super cool and think of how it would be really neat if applied in a totally different context. [MXC Builds] saw an awesome lamp from [karacreates], but decided it would be better as a PC case.

We love seeing how different techniques can be used in conjunction to make something that no one method could produce on its own, and for this build, we see [MXC Builds] use 3D printing, laser cutting, CNC, sewing, soldering, and traditional woodworking techniques.

A large part of the video is spent on the CNC process for the walnut base and power button enclosure for the build. As with any project, there are a few places requiring some creative use of the tools on hand, like the walnut piece for the base being too tall for the machine’s usual z-calibration puck or any of [MXC Builds]’s bits to do in one pass, and it’s always interesting to see how other makers solve these issues.

If you’re looking for other beautiful casemods, how about a transparent PS2 or this Art Deco number? Before you go, may we bend your ear about how PC Cases are Still Stuck in the Dark Ages?

Continue reading “A Beautiful Lamp-Inspired PC Case”

New Release Of Vision Basic: Hot New Features!

As the Commodore 64 ages, it seems to be taking on a second life. Case in point: Vision BASIC is a customized, special version of the BASIC programming language with a ton of features to enable Commodore 64 programs to be written more easily and with all sorts of optimizations. We’ve tested out both the original 1.0 version of Vision BASIC, and now with version 1.1 being released there are a whole host of tweaks and updates to make the experience even better!

One of the only limitation of Vision BASIC is the requirement for expanded RAM. It will not run on an unexpanded C64 — but the compiled programs will, so you can easily distribute software made using Vision on any C64. A feature introduced in version 1.1 is support for GeoRAM, a different RAM expansion cartridge, and modern versions of GeoRAM like the NeoRAM which has battery-backed RAM. This allows almost instantaneous booting into the Vision BASIC development environment.

Continue reading “New Release Of Vision Basic: Hot New Features!”

IBM’s 1969 Educational Computing

IBM got their PCs and PS/2 computers into schools in the 1980s and 1990s. We fondly remember educational games like Super Solvers: Treasure Mountain. However, IBM had been trying to get into the educational market long before the PC. In 1969, the IBM Schools Computer System Unit was developed. Though it never reached commercial release, ten were made, and they were deployed to pilot schools. One remained in use for almost a decade! And now, there’s a new one — well, a replica of IBM’s experimental school computer by [Menadue], at least. You can check it out in the video below.

The internals were based somewhat on the IBM System/360’s technology. Interestingly, it used a touch-sensitive keypad instead of a traditional keyboard. From what we’ve read, it seems this system had a lot of firsts: the first system to use a domestic TV as an output device, the first system to use a cassette deck as a storage medium, and the first purpose-built educational computer. It was developed at IBM Hursley in the UK and used magnetic core memory. It used BCD for numerical display instead of hexadecimal or octal, with floating point numbers as a basic type. It also used 32-bit registers, though they stored BCD digits and not binary. In short, this thing was way ahead of its time.

Continue reading “IBM’s 1969 Educational Computing”

Mainframe Chip Has 360MB Of On-Chip Cache

It is hard to imagine what a mainframe or supercomputer can do when we all have what amounts to supercomputers on our desks. But if you look at something like IBM’s mainframe Telum chip, you’ll get some ideas. The Telum II has “only” eight cores, but they run at 5.5 GHz. Unimpressed? It also has 360 MB of on-chip cache and I/O and AI accelerators. A mainframe might use 32 of these chips, by the way.

[Clamchowder] explains in the post how the cache has a unique architecture. There are actually ten 36 MB L2 caches on the chip. There are eight caches, one for each core, plus one for the I/O accelerator, and another one that is uncommitted.

A typical CPU will have a shared L3 cache, but with so much L2 cache, IBM went a different direction. As [Clamchowder] explains, the chip reuses the L2 capacity to form a virtual L3 cache. Each cache has a saturation metric and when one cache gets full, some of its data goes to a less saturated cache block.

Remember the uncommitted cache block? It always has the lowest saturation metric so, typically, unless the same data happens to be in another cache, it gets moved to the spare block.

There’s more to it than that — read the original post for more details. You’ll even read speculation about how IBM managed a virtual L4 cache, across CPUs.

Cache has been a security bane lately on desktop CPUs. But done right, it is good for performance.

A Nibble Of Core Memory, In An SAO

Core memory, magnetized memory using tiny magnetic rings suspended on a grid of wires, is now more than five decades obsolete, yet it exerts a fascination for hardware hackers still. Not least [Andy Geppert], who’s made a nibble, four bits of it, complete with interactive LED illumination to show state. Best of all, it’s on a badge Simple Add-On (SAO) for fun and games at your next hacker con.

Aside from it being a fun project, perhaps the most interesting part comes in the GitHub repository, where can be found the schematic for the device. He’s built all the drive and sense circuitry himself rather than finding an old-stock core memory driver chip, which gives those of us who’ve never worked with this stuff the chance to understand how it works. Beyond that it takes input from the Stemma or SAO ports to a GPIO expander, which provides all the lines necessary to drive it all.

To show it in action he’s posted a video which we’ve placed below. If you’re hungry for more, it’s not [Andy]’s first outing into core memory.

Continue reading “A Nibble Of Core Memory, In An SAO”

An adorable mini rack for NUCs, plus a 5-port switch.

A Mini NUC Rack For Your Desktop

We (well, some of us) are complete suckers for things that are both much smaller and much larger than life. And if that thing actually does what its supposed to? Squee! So naturally, we rushed to bring you news of this mini NUC rack designed by [Jeremy Weatherford].

Inspiration comes from a lot of places, often times from stuff that lives on your desk. [Jeremy] had a pile of NUCs and thought they resembled a mini rack already, so why not build them one to live in? It was the perfect excuse to learn CAD, so off [Jeremy] went. Although this is a mini rack, the parts were too big to print. Another opportunity presented itself, and [Jeremy] tried out an online service to get the acrylic cut.

Assembly may have been fiddly with super glue all over the nice black acrylic, but [Jeremy] learned an important tip: excess glue can be removed with vegetable oil. Once it was built, he decided to make it into a control system lab and even found a perfect little five-port switch to top it off. The logo plate, of course, is the icing on this cake.

If you prefer your tower of mini-computers to be extruded, we covered a clever design from [Jay Doscher] back in May.