Outside of hello worlds, I’m always struggling to find a use case for such small boards. Any ideas?
I've started to dabble in this only very recently. I have two iCEBreaker boards [0] in the mail right now that I'm excited for.<p>From what I understand, one of the big advantages of the up5k chip (the same chip on the webfpga and the icebreaker) is that there's an open toolchain that runs in seconds (at this chip size), which is not the case in the closed/Vivado/"typical fpga" part of the world. (feel free to correct me)<p>Purely in simulation this week, I had Claude (Fable) design a working Tetris out of 1400 relays - originally, this was for a Minivac web simulator that I/Fable had built [1]. Then, it compiled the relay netlist to Verilog and ran it on the simulated chip, comparing this (bezerk) port relay-for-relay against the original simulator. All matching (total llm time: less than an hour). I had it also make a viz of the netlist blinking through one simulated run [2] - sort of like visual6502.org..!<p>Today, I (Fable) ported tinystories [3], a very small language model to this exact same chip. It's currently "golfing" to get it to fit onto the real chip size (it's ~10% over, I'm hopeful). EDIT: it just golfed it! ~~7 tokens/sec at 12Mhz in simulation!<p>None of these are "use cases", but these boards seem small/cheap enough that you can do new deranged things with them, which is very very exciting to me. I see a great future.<p>[0] <a href="https://1bitsquared.com/products/icebreaker" rel="nofollow">https://1bitsquared.com/products/icebreaker</a><p>[1] <a href="https://minivac.greg.technology/tetris/" rel="nofollow">https://minivac.greg.technology/tetris/</a><p>[2] <a href="https://necessary-doom-48128.ondis.co/" rel="nofollow">https://necessary-doom-48128.ondis.co/</a><p>[3] <a href="https://arxiv.org/abs/2305.07759" rel="nofollow">https://arxiv.org/abs/2305.07759</a>
A trick to fit language models of this size it to remove the word to embeddings from the NN, and have a database to look up a dictionary of words and their embeddings. This way the LLL only need the "core" and you do the reverse for loopup again (kind of Text->RAG->LLM->RAG->Text). Have an example here: <a href="https://punnerud.github.io/pyspell/" rel="nofollow">https://punnerud.github.io/pyspell/</a><p>The example have a limited language of around 1000 words, but make it possible to do Python (like) programming with LLM on an ESP32.
wow, that tailscale-in-browser demo is wild..!<p>re: language model - here, the embedding lookup, layers, logits all run on-chip and loops its own output back.<p>the only off-chip piece is the id to string table, i.e. the chip (once I get it!) will speak token ids and my laptop will print them as letters
The demo use WASM/browser when available, but also supports running only on ESP32. Figured out I could offload and give the users a better experience when available.<p>Thanks for noticing the Tailscale demo. Created a Tailscale-Rust client that supports NAT-traversial, encryption etc. and manage to run on ESP32 (together with the local LLM). Just the Tailscale part can be found here: <a href="https://github.com/punnerud/tailscale-mpe-rust" rel="nofollow">https://github.com/punnerud/tailscale-mpe-rust</a>
Examples include:<p>* replacements for the C64 PLA, CIA, and potentially the SID<p>* Glue logic for a bread board project where it would require a lot of 74 or 4k series logic<p>* Translating TTL to VGA
etc.<p>These are things people have used FPGAs for quite commonly
For FPGAs in general: When latency is important. Even a $1 MCU has a huge amount of compute power nowadays, and they can simulate many things, but if you need to deterministically react within 1 microsecond to what happens on a pin, and your MCU doesn't already have a hardware peripheral that does exactly what you want, you're probably going to need an FPGA.<p>Some recent advances, like the Raspberry Pi Pico's PIO units, have made it possible to do things that MCU's couldn't do before, like emulating a ROM chip. But there are limits to what the PIOs can do.<p>If it comes to pure computation and not latency, you need some very specific use case before an FPGA becomes better than a regular CPU. Most likely when it involves massive parallelism and custom caches and interconnects. Although nowadays NPUs are starting to fill that niche.<p>For tiny boards in particular: getting experience and having fun.
ICE40UP5K is large enough to synthesize a simple RISC-V CPU with some basic peripherals. It's not huge or fast, but it's capable enough.
Note, if you need a barebones CPU but want to save on gates, you can go with a serial (1-bit) CPU, e.g. SERV which only takes up 198 LUTs on ice40,[1] leaving plenty of space to do other things.<p>[1] <a href="https://github.com/olofk/serv" rel="nofollow">https://github.com/olofk/serv</a>
You mean it's large enough to synthesize several application specific RISC-V cores and let them form an embedded distributed system.
performing glitch attacks on embedded systems, consoles, phones etc.
small what boards? Do esp32 boards count?
Presumably small FPGA boards, like the $47 one in the linked article<p>Small boards have less I/O pins than larger boards typically, and small/cheap FPGA chips <i>on</i> small boards are also more limited than larger, more expensive parts.<p>Edit: precise price.
Jailbreaking consoles<p><a href="https://a.co/d/05f9gq9A" rel="nofollow">https://a.co/d/05f9gq9A</a>