Peering Into The Murky Depths Of Alder Lake

The winds of change are in the air for CPUs. Intel has long lorded over the computing world, and they remain a force to contend with, but many challengers gather at their gates. AMD, ARM, IBM, and other X86 designs sense a moment of weakness. In response, Intel released their Alder Lake platform with high-performance and high-efficiency cores, known as Golden Cove and Gracemont, respectively. [Clamchowder] and [cheese] have written up as many details as they were able to suss out about Gracemont.

ARM has done a multi-multi core design (big.LITTLE) for several years where they have a mix of high-power, high-performance cores and smaller, low-power cores. This allows the scheduler to make tradeoffs between power and performance. Typically the smaller cores in an ARM design are simpler in-order processors, having more in common with a microcontroller than with a full-scale desktop core. Many people have made an obvious comparison with the apparent similarities between ARM’s approach and Intel’s new offerings as Gracemont is based on Intel’s old Atom core, a low-power single issue, in-order processor. Continue reading “Peering Into The Murky Depths Of Alder Lake”

Apple Falling Division

[Paul Curtis] over at Segger has an interesting series of blog posts about calculating division. This used to be a hotter topic, but nowadays many computers or computer languages have support for multiplication and division built-in. But some processors lack the instructions and a library to do it might be less than ideal. Knowing how to roll your own might allow you to optimize for speed or space. The current installment covers using Newton’s algorithm to do division.

Steve Martin had a famous bit about how to be a millionaire and never pay taxes. He started out by saying, “First… get a million dollar. Then…” This method is a bit like that since you first have to know how to multiply before you can divide. The basic premise is twofold: Newton’s method let you refine an estimate of a reciprocal by successive multiplications and then multiplying a number a reciprocal is the same as dividing. In other words, if we need to divide 34 by 6, you could rewrite 34/6 to 34 * 1/6 and the answer is the same.

Continue reading “Apple Falling Division”

ESP32 Internet Radio Is No Game

More than once, we’ve looked at a cool board like the TTGO T-Display and thought, “What can we build with this?” If you’re [Danko Bertović], the answer is the tiny Internet radio you can see in the Volos Projects video below.

Of course, the core Internet streaming code would be useful with any ESP32, but the display makes for a good-looking unit. The code is available on GitHub. With judicious use of network and audio libraries, the player only takes a few hundred lines of code. Pretty impressive considering it even shows a visualization on the tiny display screen.

What we’d really like to see is a nice case, power supply, and speaker option to make a tiny and portable unit. With a 3D printer, it is easy to make very professional-looking projects, as we often see. On the other hand, it does look better than the breadboard version you can see towards the end of the video. It is, though, a neatly done breadboard.

If you want a larger screen, you might enjoy the ESP32 internet radio we looked at before. Probably our favorite case for an Internet radio was this globe.

Continue reading “ESP32 Internet Radio Is No Game”

ESP32-Cam Makes A Dandy Motion Detector

Halloween is right around the corner and just about every Halloween project needs some kind of motion sensor. Historically, we’ve used IR and ultrasonic sensors but [Makers Mashup] decided to use an ESP32-Cam as a motion sensor in his latest animatronic creation. You can see a video of the device and how it works below.

The project is a skull that follows you around with a few degrees of motion on a stepper motor. There’s a 3D-printed enclosure to make the hardware assembly easy. The base software was borrowed from [Eloquent Arduino].

Continue reading “ESP32-Cam Makes A Dandy Motion Detector”

ARM’s Chinese Venture Goes Sour

We’re used by now to many of the more capable microcontrollers and systems-on-chip that we use having an ARM core at their heart. From its relatively humble beginings in a 1980s British home computer, the RISC processor architecture from Cambridge has transformed itself into the go-to power-sipping yet powerful core for manufacturers far and wide. This has been the result of astute business decisions over decades, with ARM’s transformation into a fabless vendor of cores as IP at its heart. Recent news suggests that perhaps the astuteness has been in short supply of late though, as it’s reported that ARM’s Chinese subsidiary has gone rogue and detatched from the mothership taking the IP with it.

It seems that the CEO of the Chinese company managed to retain legal power when sacked by the parent company over questionable ties with another of his ventures, and has thus been able to declare it independent of its now-former parent. It still has the ARM IP up to the moment of detatchment and claims to be developing its own new products, but it seems likely that it won’t receive any new ARM IP.

What will be the effect of this at our level? Perhaps we have already seen it, as more Chinese chips such as the cheaper STM32 clones are likely to get low-end ARM cores as a result. It seems likely that newer ARM IP will remain for now in more expensive non-Chinese chip families, but in the middle of a semiconductor shortage it’s likely that we wouldn’t notice anyway. Where it will have a lasting effect is in future Chinese joint ventures by non-Chinese chip companies. Seeing ARM’s then-owner Softbank getting their fingers burned in such a way is likely to provide a disincentive to other companies considering a similar course. Whether ARM will manage to resolve the impasse remains to be seen, but it can hardly be a help to the rocky progress of their Nvidia merger.

Arm Researchers Announce The PlasticArm

If the Cortex family of embedded microprocessors aren’t flexible enough for your designs, an article published this week (click here for the PDF version) in the journal Nature might be of interest. We’re not talking flexibility in terms of features, but real, physical flexibility of the microprocessor itself. A research team from Arm Ltd. has developed the PlasticArm, which is a 32-bit processor derived from the Cortex-M0+ family.

They accomplished this by constructing a CPU from metal-oxide thin-film transistors (TFT) on a polyimide substrate, the resultant chip being called a natively flexible microprocessor. While much of the hype focuses on the flexibility aspect, we think the real innovation here is the low cost. The processes used to deposit transistors onto silicon wafers is much more expensive than those on this flexible substrate.

Don’t get too excited just yet, because there were some compromises made along the way. Modern microprocessor silicon dies are measured in the tens of microns, but the PlasticArm total die size is a comparatively whopping 9 mm square. The researchers were appropriately focused on the core CPU, and the auxiliary building blocks such as ROM and RAM seem almost an afterthought. With only 456 bytes of program store and 128 bytes of RAM, only the tiniest of applications are suited to this chip. Other compromises were made, such as no internal registers — they are mapped to the external RAM — and the CPU runs a lot slower than we’re used to, topping out at 29 kHz (note: k not M).

There are certainly some challenges with this new technology, and we won’t be designing with these chips any time soon. But it has the potential to offer benefits in certain niche applications where low-cost and/or flexibility is more important than processor speed and performance.

 

Is 32-bits Really Dead?

While some of us are still clinging onto our favorite 8-bit microprocessors, ARM announced they will be killing off the 32-bit architecture in 2022 and/or 2023. Over on the GaryExplains YouTube channel, [Gary Sims] posted a great review of the current 32- vs 64-bit state-of-affairs — not just for ARM but for Intel and AMD processors as well. And it’s a dismal outlook for you 32-bit fans.

ARM announced last Fall that there would be no more 32-bit support as of 2022, then this March they made a similar announcement but with a 2023 deadline. [Gary] tries to parse these statements, and takes an educated guess at what the disparity means (spoiler alert — he predicts that one more 32-bit core will soon be released).

[Gary] clearly breaks down the 32-bit situation by operating systems such as Linux, Windows, MacOS, Android, and iOS, and how all of these have been transitioning to 64-bits over recent years. He does a thorough job, and concludes that the transition is already well underway. And while Linux and Windows have not completely dropped 32-bit support, the writing is on the wall.

Take note, however, that this discussion regards the Cortex-A family of cores found in smart phones, tablets, computers, and powerful embedded applications like autonomous vehicles. The popular 32-bit Cortex-M family of low-cost / low-power cores that are used in so many embedded system designs will remain 32-bits for the foreseeable future.

After watching [Gary]’s presentation, if you want to learn more, check out the writeup that [Maya Posch] did on the details of the latest ARMv9 ISA a few weeks ago. Also watch this 8-bit vs 32-bit presentation by our Editor-in-Chief [Mike Szczys]. Despite being from five years ago, it is still quite applicable today. What about 16-bit MCUs — the old Intel/AMD embedded 80186 processor, the 8051 follow-ons like the 80C196, 80C251, or 8051XA, the 6502 follow-ons like the 65C816, Zilog’s Z8000, the Renesas M16C, etc. — is anyone using them anymore? If so, or if you’re using a 4-bit MCU these days, let us know in the comments below.

Continue reading “Is 32-bits Really Dead?”