A RISC-V chip with built-in neural networks

From the comments by user Zepan:<p>Hi, thanks for your report ! I’m zepan, the shopkeeper of Sipeed Tech. Taobao shop, Sipeed M1 is our first AI moulde ,and the second RISC-V module.<p>I need correct some mistake in your article, and supplement some information.<p>1. The chip name is K210, not K201; and it is RV64GC.<p>2. It have 8MB high speed SRAM, not 6MB.<p>3. There are 5.9MB SRAM can be used for convolutional neural network acceleration, so, it is possible to run small model like tiny-yolo v2,MobileNet, as you see in face detection routine video.<p>4. It isn’t preoder mode before 10.7, but it is too popular that it is sold out in 10.7, and I change it to preorder mode.<p>More information:<p>1. It can be up to 800MHz if you want, and give you more than 0.5TOPS, 240FPS detection speed @ QVGA.<p>2. It is low power consume, unlike other ARM board; It only consume 0.35W when running face detection routine.<p>3. Its IO speed in simple test is up to 300MHz<p>4. It is possible to run openwrt on it as it have MMU, but too tight in ram.<p>5. We’re adding micropython to the board, and soon you can simply run python on it, you know python is the best lang for DL.<p>6. We also have Mic Array board, and another fancy board in development.<p>7. We have best price for those solution, Sipeed M1 module in $6, M1w(with wifi) in $7.5, the simple dev board with LCD, 2M pixel Camera, I2S MIC, Speaker PA, on board downloader, in $15.<p>8. We have model shop soon, you can sell your model on our online modelshop.<p>9. Our telgram link: <a href="https:&#x2F;&#x2F;t.me&#x2F;joinchat&#x2F;IoJz2UoLTnlscC0WrCCNrg" rel="nofollow">https:&#x2F;&#x2F;t.me&#x2F;joinchat&#x2F;IoJz2UoLTnlscC0WrCCNrg</a><p>Our qq group: 878189804 for AI, 826307240 (3k members group) for common discussion.

Here is a bit detail of the neural network chip. [1]<p>Anything aside, as a former embedded dev, I would never use there kinds of voodoo chinese chips. Don’t get me wrong, it’s not because it’s from China. It is because of __zero__ well-written public documentation.<p>Just listing specs in datasheets is meaningless. The datasheet is translated but it does nothing to devs because it has absolutely zero about low level peripherals and its core. You cannot imagine how painful writing hardware-level code is when the documentation doesn’t help. I’m 100% sure even the chinese version documentation has almost zero information too.<p>Look, they have a sort of SDKs and the only example is one-line-long Hello World. [2]<p>[1]; <a href="https:&#x2F;&#x2F;github.com&#x2F;kendryte&#x2F;kendryte-doc-ds&#x2F;blob&#x2F;master&#x2F;en&#x2F;003.md#neural-network-processor-kpu" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;kendryte&#x2F;kendryte-doc-ds&#x2F;blob&#x2F;master&#x2F;en&#x2F;0...</a><p>[2]: <a href="https:&#x2F;&#x2F;github.com&#x2F;kendryte&#x2F;kendryte-freertos-sdk&#x2F;blob&#x2F;develop&#x2F;src&#x2F;hello_world&#x2F;main.c" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;kendryte&#x2F;kendryte-freertos-sdk&#x2F;blob&#x2F;devel...</a>

<i>”Documentation is available, although the datasheet will need to be translated”</i><p>That’s one way to see it. An alternative is <i>”If you’re a young hacker, you better start learning Chinese, just as young hackers fifty-ish years ago didn’t wait for translations, but learnt English”</i>

Since RISC-V is being mentioned here, thought maybe I took the opportunity to ask about it. I would like to learn more about instruction set, I know very little about it so I was planning on purchasing a book called <i>Computer Organization and Design</i>, it comes in three flavors: MIPS, RISC-V, and ARM. Which one would be the best option to read of those given, or which one would be the most relevant where I would get the most value out of it. Can anyone provide a minimal suggestion? Thank you in advance!

Since Nvidia is replacing Falcon with RISC-V there is also neural networks with build in RISC-V.

Once it becomes available again I would really like to order one (or more), but the web page linked from the article for ordering is in Chinese, and I didn&#x27;t see any way to get to an English language ordering page.

This offers an order of magnitude more computation per $ or per watt than anything else in its field.<p>I bet they&#x27;ll be out of stock for a really long time sadly :-(

Please let me know the URL for online purchase of Sipeed M1 K210