bunnie Huang

Ask yourself why you want something that is made in the US. I have long had the same bias, and I am trying to understand it myself. Is it perceived quality of the final object? Do you feel that the environmental impact of shipping is worth overcoming? Do you want to keep money in local communities? Do you just think Americans make things right? What’s your goal here. Also read about the too numerous to count HW kickstarter projects that 10x their funding goal and then still never deliver. Also go read everything [Bunnie Huang](https://www.bunniestudios.com/) has written.

I'm a big fan of Adafruit, which is a great place to buy electronic components that all have projects, tutorials, active forums, etc. Here's the "learn" page on the Adafruit web site: https://learn.adafruit.com/ And the Adafruit Youtube channel: https://www.youtube.com/adafruit And it's definitely worth just dipping in and out of -- and eventually reading all the way through -- "The Art of Electronics." It's a terrific book, and manages to be both readable and super in-depth about every topic you could possibly want to know about when you're trying to figure out what's what. There are also a number of inspiring blogs by electrical engineers who are also great writers. Which ones to follow sort of depend on what specialties you are most interested in, but Bunnie Huang's is a great one to start with: https://www.bunniestudios.com/

There are a bunch of interesting accounts which basically boil down to the following: you can do in Shenzhen in a few days that which would take weeks to months in the US or EU (well, most of western civilization). Take "Ten Days in Shenzhen" at https://www.karimyaghmour.com/blog/2018/01/ten-days-in-shenzhen.html or Bunnie Huang's blog at https://www.bunniestudios.com/ . It stems from the fact that everything is local. You don't have to wait for product to get shipped across the country for each individual step in manufacturing when you can just walk / taxi / bus it between the manufacturers in a day or two. Iterating is fast. TSMC has a similar work ethic. They have process engineers running experiments 24/7 that each build on the results of the previous shift. When's the last time Intel could hire folks with that kind of dedication in Oregon? Netflix has a documentary called The Factory which touches on the kind of cultural differences between workers in the US vs China that is rather enlightening.

Bunnie Huang (of MIT fame, original cracker of Xbox, proponent of open hardware [see: Chumby, Novena, etc]) has a really interesting talk given at the Media Labs in '15 at his alma mater[1]. It's about 1:15 but fully worth the watch including the QA, if you have any interest in 'open hardware'. That Rockchip run's on a fairly complicated 4core ARM. So which while the instruction set may be open, so is Intels. They'll happily ship to you their 3 volume multi-thousand page ISA. (Open is when you get to see any of final Altium/Cadence files that were used for the tape-out to make production wafers). At best you might get a few block diagrams. Good luck getting anything more than that. The closest you'll get to something open might be collaborative open standards a la the OpenPOWER (still, not that open) or RISC-V (which is probably still fab'ing at what, 65nm? 45?) Anyways, that Rockchip is far from open. watch the talk to see his rationale behind choosing Freescale as his microprocessor of choice. He remarks on how $2.25 SoC picked up on the streets of Shenzhen have the full capabilities to run on GSM, which is really pretty remarkable. Obviously they're stealing IP and probably using grey-market tech/seconds that failed QA, but when he talks about the innovation _within_ that grey-market its fascinating. Even with his PhD in EE from MIT, experience in taking consumer products to market multiple times, fluency in Mandarin to aide in his ability to navigate the markets of…

Not until software lottery tickets quit returning 10x+. Who, in their right mind, would invest in hardware? NRE kills--especially in mechanical engineering which actually seems to be regressing. Machining companies now charge a fortune for things which my grandfather used to do by hand in his garage in about an hour. Not to mention how long it takes to get an injection mold (14 weeks is not uncommon). For the price of a couple injection molds, I can flog a gaggle of stupid 20-somethings living in squalor for about a year. No contest as to which has a better probability of cashing out. In addition, who is going to buy out a hardware startup? Hardware really doesn't fit the Yahooglezoft strategy, so now your exits are very limited. Even worse, hardware acquirers tend to be rational about things like costs and profit. Finally, everybody who does hardware tries to create the cloud back end and then get anybody else to do the hardware while they extract the rent. So, longer timelines to cash out with higher risk. Yeah, don't think so. Now, the upside is that if you do manage to get a product out, you have an actual barrier to entry. 2 guys and a dog in the Ukraine can't just clone you. While I love the manufacturing incubators, they don't solve the fundamental problem of not knowing how to run machines. The existence of a CNC machine or a TIG welder doesn't magically make you a machinist or a welder. Being able to run a CNC machine doesn't mean you know how to…

It's great, but I think the critique from the other day was also pretty valid. and offered an alternative. https://www.bunniestudios.com/blog/2026/bio-the-bao-i-o-coprocessor/ I think, for my use, just having the ability to write to DMA registers would have been a big advantage. It feels wasteful to have A DMA waiting on a FIFO just to write what it gets to DMA registers to do the transfer you actually wanted. Looking at the Architecture diagram It seems like it could have allowed that and stayed on the same side of the AHB5 splitter.

Like all open-source, it's an ongoing effort. Bunnie has a comparison, https://www.bunniestudios.com/blog/2017/litex-vs-vivado-first-impressi… > Thanks to the extensive work of the MiSoC and LiteX crowd, there’s already IP cores for DRAM, PCI express, ethernet, video, a softcore CPU (your choice of or1k or lm32) and more.. LiteX produces a design that uses about 20% of an XC7A50 FPGA with a runtime of about 10 minutes, whereas Vivado produces a design that consumes 85% of the same FPGA with a runtime of about 30-45 minutes.. LiteX, in its current state, is probably best suited for people trained to write software who want to design hardware, rather than for people classically trained in circuit design who want a tool upgrade.

A tangential topic studies how you can actually trust the hardware. Andrew "Bunnie" Huang has done a lot of great work in the area, first with Precursor, and lately with Infra-Red, in situ (IRIS) inspection of silicone. * https://www.bunniestudios.com/blog/2020/introducing-precursor/ * https://www.bunniestudios.com/blog/2024/iris-infra-red-in-situ-project…

It's less "how do you", and more "can you?". And according to the Novena open-hardware laptop guy, you can't: https://www.bunniestudios.com/blog/?cat=28 >Based on these experiences, I’ve concluded that open hardware is precisely as trustworthy as closed hardware. Which is to say, I have no inherent reason to trust either at all.

> Chinese generic products are often semi-open source amongst design/engineer and manufacturing people. As greatly commented a decade ago by Bunnie Huang, calling it "gongkai" [0] [0] https://www.bunniestudios.com/blog/?p=4297

a related classic from 2010: fake microsd cards https://www.bunniestudios.com/blog/?page_id=1022

I'd love to try this some time. Fritzchens Fritz has some other great examples of IR die shots too! [^1]: https://www.flickr.com/photos/130561288@N04/52465673068/ edit: Oh, bunnie also mentions this in a previous post! See https://www.bunniestudios.com/blog/?p=6656

The author of that quote, Andrew "bunnie" Huang, has an excellent blog which is also worth reading: https://www.bunniestudios.com/blog/ You can grab the download directly from NoStarchPress: https://nostarch.com/hardwarehackerpaperback

Bunnie Huang [1] has written/spoke on this topic. I think he mentions it in the Shenzen documentary [2]. Quote from the documentary page: > We examine the unique manufacturing ecosystem that has emerged, gaining access to the world’s leading hardware-prototyping culture whilst challenging misconceptions from the west. The film looks at how the evolution of “Shanzhai” – or copycat manufacturing – has transformed traditional models of business, distribution and innovation, and asks what the rest of the world can learn from this so-called “Silicon Valley of hardware". 1: https://www.bunniestudios.com 2: https://www.youtube.com/watch?v=SGJ5cZnoodY

And, there's a couple of great write-ups about them [0][1]. [0]: https://www.bunniestudios.com/blog/?p=3554 [1]: https://i.blackhat.com/us-18/Wed-August-8/us-18-Valadon-Reversing-a-Ja…

No love for Precursor? https://www.bunniestudios.com/blog/?p=5921

Amazing that we went from this potato[1] for $12 in 2013 to that Redmi 2 in 9 years. Now, you said that you buy the Redmi second-hand, but still - the leap is amazing. [1] https://www.bunniestudios.com/blog/?page_id=3107

A thoughtful tour of a Lego minifig as sliced by a CT scanner. This is kind of the reverse of 3D printing - rather than build something layer by layer, the assembly is scanned and visualized in slices, allowing a peek into the way parts mate. Differences in density can be automatically color coded - in this case the dye of the figurine provides a convenient separation between differently-colored parts. I never knew that the hands of minifigs were ribbed to create a rotating press fit! This reminds me of the fun I have reading [name that ware](https://www.bunniestudios.com/blog/?cat=54) by Bunny Huang. There's a sense of discovery and delight in analyzing the choices made by designers of physical things. Does anyone know the site library used to pull off this scroll-animation? It works excellently on mobile and I recall it being used heavily on Apple.com to showcase product cut-aways.

https://www.bunniestudios.com/ should have a lot of blog posts with real expertise and research behind them, he’s been blogging about doing this for ages.

You might like Bunnie Huang's Precursor... https://www.crowdsupply.com/sutajio-kosagi/precursor "Powered by an FPGA-hosted, soft-core System-on-Chip (SoC), it gives developers the freedom to inspect, verify, and customize nearly every aspect of its operation. Help us take those critical first steps toward a world in which silicon-level trustworthiness is attainable." So RISC-V, but with the cores and peripherals all software defined on an FPGA, so you can audit the code and verify what you upload as "software defined hardware". They've blogged profusely about the design decisions and the philosophies behind them: https://www.bunniestudios.com/blog/?cat=71

The assumptions this piece makes about what makes hardware trustworthy are simply not valid and would be dangerous to follow if you really have reason to care about this issue. The ship sailed on trusting any hardware years ago. See this excellent explanation: https://www.bunniestudios.com/blog/?p=5706

Err...the only thing "modern" about this is you can purchase/fabricate one using new material and components in active production. If a modern cell phone integrated those SMT packages, it'd be the size of a brick, and the PCB itself is vanilla 2-layer Tg130 FR4...couldn't even be bothered with a controlled impedance layout, let alone length-match bus traces. Recently, bunnie talked about one "relatively modern" PCB design element---microvias---and explains why you just don't see them much in OSHW designs[1]...an excellent read if you get the chance. JLC is a well known PCB fab house because their low-volume prices are dirt cheap, not because their process is modern. [1] https://www.bunniestudios.com/blog/?p=6011

As someone who attended a private engineering school in Boston (although not MIT) I see a move from scheme to python as regressive in almost every way. Scheme levels the playing field among incoming students, both those who've barely programmed and who've been programming since they were 14. It also provides a special kind of mind bender that challenged me in ways I still use today. Granted I work at a company that employs multiple functional programming languages. Good engineers in time don't use python because they realize it attracts tech debt and cruft. I love python, and still use it for quick hacks and machine learning things. However, I'll defer to the one and only Bunnie Huang regarding python v other languages [0]"fwiw, Python is a great glue, but a terrible structural material. Do not build skyscrapers out of Python" [0] https://www.bunniestudios.com/blog/?p=5863

Agree. There are a number of Go-based clients, e.g. acmetool and lego. As Andrew "bunnie" Huang said recently: "Python is a great glue, but a terrible structural material. Do not build skyscrapers out of Python" (https://bunniestudios.com/blog/?p=5863).

Bunnie Huang gave an excellent talk on this subject at BlueHat IL 2019 that goes very deep, including showing technical details of how attacks like this can be pulled off. Well worth the 45 minutes - https://www.bunniestudios.com/blog/?p=5519

Odd to see such a long article with no mention of MMC, TF, or xD. Also worth noting that these cards (with the exception of xD, which is just a packaged raw NAND flash) have a complete embedded system in them too: https://www.bunniestudios.com/blog/?p=3554

bunnie wrote a great article about fake MicroSD cards back in 2010: https://www.bunniestudios.com/blog/?page_id=1022

I think the fact that fake components regularly make their way into the supply chain should be more of a concern than the supermicro story. So, supply chain security is poor, and getting backdoored parts into the supply chain is a realistic possibility. Bunnie has a good recent talk on this. https://www.bunniestudios.com/blog/?p=5519 The question in my mind, is is it worth the risk of detection. Particularly when other methods are less traceable, more easily deniable.

True. You could to some extent. I think part of the story is the opportunity though. The domain name, the boxes and the time are all things that grant him the opportunity to do it. He probably could have sold them in plastic bags on eBay (someone probably already does), but it wouldn't be same thing. We all seem to agree like things like this is a good thing, but then we undermine people's ability to do it by favoring things like real estate, domains and "middle man businesses" instead. I guess bunnie said it best: "supply chains are made out of people" https://www.bunniestudios.com/blog/?p=4266

It's extremely hard to document because the chips tend not to have unique identifiers, and since they all come from the same factory they're all "real". The best investigation I've seen is https://www.bunniestudios.com/blog/?page_id=1022 on fake SD cards.