5 ms·
Show HN: We made our own inference engine for Apple Silicon
We wrote our inference engine on Rust, it is faster than llama cpp in all of the use cases. Your feedback is very welcomed. Written from scratch with idea that you can add support of any kernel and platform.
- sharifulin 1y agoWow! Sounds super interesting
- slavasmirnov 1y agothat’s exactly we are looking for not to waste on apis. Wonder how significant trade offs are
- TheMagicHorsey 1y agoAmazing! How was your experience using Rust on this project? I'm considering a project in an adjacent space and I'm trying to decide between Rust, C, and Zig. Rust seems a bit burdensome with its complexity compared to C and Zig. Reminds me of C++ in its complexity (although not as bad). I find it difficult to walk through and understand a complicated Rust repository. I don't have that problem with C and Zig for the most part. But I'm wondering if I just need to invest more time in Rust. How was your learning curve with the language?
- adastra22 1y agoYou are confusing familiarity with intrinsic complexity. I have 20 years experience with C/C++ before switching to rust a few years ago. After the initial hurdle, it is way easier and very simple to follow.
- TheMagicHorsey 1y agoAre you generally able to quickly understand what is going on in somebody else's codebase written in Rust? I find it quite difficult to understand other people's Rust code. Is this just a familiarity thing? I have not written anything particularly huge or complex in Rust, but I have written a few CLI utilities. With an equivalent level of Go exposure, I find it much easier to understand code written in Go, compared to code written in Rust. I'm quite proficient in C/C++ (started coding in C/C++ in 1997) but I still have a much harder time understanding a new C++ project compared to a C project.
- ednevsky 1y agonice
- ewuhic 1y ago>faster than llama cpp in all of the use cases What's your deliberate, well-thought roadmap for achieving adoption similar to llama cpp?
- pants2 1y agoProbably getting acquired by Apple :)
- khurs 1y agoOllama is the leader isn't it? Brew stats (downloads last 30 days) Ollama - 28,232 Lama.cpp - 7,826
- DiabloD3 1y agoOllama isn't an inference engine, its a GUI slapped onto a perpetually out-of-date vendored copy of Llama.cpp underneath. So, if you're trying to actually count LLama.cpp downloads, you'd combine those two. Also, I imagine most users on OSX aren't using Homebrew, they're getting it directly from the GH releases, so you'd also have to count those.
- imtringued 1y agoActually, ollama has stopped using llama.cpp and is using ggml directly nowadays.
- mintflow 1y agojust curios, will it be supported on iOS, it would be great to build local llm app with this project.
- AlekseiSavin 1y agoalready) https://github.com/trymirai/uzu-swift https://github.com/trymirai/uzu-swift
- cwlcwlcwlingg 1y agoWondering why use Rust other than C++
- adastra22 1y agoWhy use C++?
- khurs 1y agoSo C++ users don't need to learn something new.
- bee_rider 1y agoI wonder why they didn’t use Fortran.
- giancarlostoro 1y ago...or D? or Go? or Java? C#? Zig? etc they chose what they were most comfortable with. Rust is fine, it's not for everyone clearly, but those who use it produce high quality software, I would argue similar with Go, without all the unnecessary mental overhead of C or C++
- outworlder 1y agoWhy use C++ for greenfield projects?
- khurs 1y agoThe recommendation from the security agencies is to prefer Rust over C++ as less risk of exploits. Checked and Lama.cpp used C++ (obviously) and Llama uses Go.
- greggh 1y ago"trymirai", every time I hear the word Mirai I think of the large IOT DDoS botnet. Maybe it's just me though.
- fnord77 1y agoI think of the goofy Toyota fuel cell car. I think a grand total of about 6 have been sold (leased) in california
- rnxrx 1y agoI'm curious about why the performance gains mentioned were so substantial for Qwen vs Llama?
- AlekseiSavin 1y agoit looks like llama.cpp has some performance issues with bf16
- homarp 1y agoCan you explain the type of quantization you support? would https://docs.unsloth.ai/basics/kimi-k2-how-to-run-locally https://docs.unsloth.ai/basics/kimi-k2-how-to-run-locally be faster with mirai?
- AlekseiSavin 1y agoright now, we support AWQ but are currently working on various quantization methods in https://github.com/trymirai/lalamo https://github.com/trymirai/lalamo
- smpanaro 1y agoIn practice, how often do the models use the ANE? It sounds like you are optimizing for speed which in my experience always favors GPU.
- AlekseiSavin 1y agoYou're right, modern edge devices are powerful enough to run small models, so the real bottleneck for a forward pass is usually memory bandwidth, which defines the upper theoretical limit for inference speed. Right now, we've figured out how to run computations in a granular way on specific processing units, but we expect the real benefits to come later when we add support for VLMs and advanced speculative decoding, where you process more than one token at a time
- J_Shelby_J 1y agoVLMs = very large models?
- mmorse1217 1y agoProbably vision language models.
- skybrian 1y agoWhat are the units on the benchmark results? I’m guessing higher is better?
- AlekseiSavin 1y agoyeah, tokens per second
- dcreater 1y agoSomewhat faster on small models. Requires new format. Not sure what the goal is for this project? Not seeing how this presents adequate benefits to get adopted by the community
- koakuma-chan 1y agoWritten in Rust is a big one for me.
- worldsavior 1y agoIt's utilizing Apple ANE and probably other optimization tools provided by Apple's framework. Not sure if llama.cpp uses them, but if they're not then the benchmark on GitHub says it all.
- zdw 1y agoHow does this bench compared to MLX?
- jasonjmcghee 1y agoI use MLX in lmstudio and it doesn't have whatever issues llama cpp is showing here. Qwen3-0.6B at 5 t/s doesn't make any sense. Something is clearly wrong for that specific model.
- giancarlostoro 1y agoHoping the author can answer, I'm still learning about how this all works. My understanding is that inference is "using the model" so to speak. How is this faster than established inference engines specifically on Mac? Are models generic enough that if you build e.g. an inference engine focused on AMD GPUs or even Intel GPUs, would they achieve reasonable performance? I always assumed because Nvidia is king of AI that you had to suck it up, or is it just that most inference engines being used are married to Nvidia? I would love to understand how universal these models can become.
- darkolorin 1y agoBasically “faster” means better performance e.g. tokens/s without loosing quality (benchmarks scores for models). So when we say faster we provide more tokens per second than llama cpp. That means we effectively utilize hardware API available (for example we wrote our own kernels) to perform better.
- nodesocket 1y agoI just spun up a AWS EC2 g6.xlarge instance to do some llm work. The GPU is NVIDIA L4 24GB and costs $0.8048/per hour. Starting to think about switching to an Apple mac2-m2.metal instance for $0.878/ per hour. Big question is the Mac instance only has 24GB of unified memory.
- khurs 1y agoUnified memory doesn't compare to a Nvidia GPU, the latter is much better. Just depends on what performance level you need.
- floam 1y agoHow does this compare to https://github.com/Anemll/Anemll https://github.com/Anemll/Anemll?
- zackangelo 1y agoWe also wrote our inference engine in rust for mixlayer, happy to answer any questions from those trying to do the same. Looks like this uses ndarray and mpsgraph (which I did not know about!), we opted to use candle instead.
- khurs 1y agoHave you added it to HomeBrew and other package managers yet? Also any app deployed to PROD but developed on Mac need to be consistent i.e. work on Linux/in container.
- woadwarrior01 1y agoNeeds an "API key". https://github.com/trymirai/uzu-swift?tab=readme-ov-file#quick-start https://github.com/trymirai/uzu-swift?tab=readme-ov-file#qui...
- iglushenkov 1y agocooollll
- ethx64 1y ago[dead]