On the edge of Apple Silicon memory speeds
I have developed open source CLI-tool for Apple Silicon macOS. It measures memory speeds in different ways and also latency. It can achieve up to 96-97% efficiency on read speed on M4 base what is advertised as 120GB/s. All memory operations are in assembly. I would really appreciate for results on different CPU's how benchmark works on those. I have been able to test this on M1 and M4. command : 'memory_benchmark -non-cacheable -count 5 -output results.JSON' (close all applications before running) This will generate JSON file where you find sections copy_gb_s, read_gb_s and write_gb_s…
What it does
In the maker’s words, at launch
I have developed open source CLI-tool for Apple Silicon macOS. It measures memory speeds in different ways and also latency. It can achieve up to 96-97% efficiency on read speed on M4 base what is advertised as 120GB/s. All memory operations are in assembly. I would really appreciate for results on different CPU's how benchmark works on those. I have been able to test this on M1 and M4. command : 'memory_benchmark -non-cacheable -count 5 -output results.JSON' (close all applications before running) This will generate JSON file where you find sections copy_gb_s, read_gb_s and write_gb_s statics. Example M4 with 10 loops: "copy_gb_s": { "statistics": { "average": 106.65421233311835, "max": 106.70240696071005, "median": 106.65069297260811, "min": 106.6336774994254, "p90": 106.66606919223108, "p95": 106.68423807647056, "p99": 106.69877318386216, "stddev": 0.01930653530818627 }, "values": [ 106.70240696071005, 106.66203166240008, 106.64410802226159, 106.65831409449595, 106.64148106986977, 106.6482935780762, 106.63974821679058, 106.65896986001393, 106.6336774994254, 106.65309236714002 ] }, "read_gb_s": { "statistics": { "average": 115.83111228356601, "max": 116.11098114619033, "median": 115.84480882265643, "min": 115.56959026587722, "p90": 115.99667266786554, "p95": 116.05382690702793, "p99": 116.09955029835784, "stddev": 0.1768243167963439 }, "values": [ 115.79154681380165, 115.56959026587722, 115.60574235736468, 115.72112860271632, 115.72147129262802, 115.89807083151123, 115.95527337086908, 115.95334642887214, 115.98397172582945, 116.11098114619033 ] }, "write_gb_s": { "statistics": { "average": 65.55966046805113, "max": 65.59040040480241, "median": 65.55933583741347, "min": 65.50911885624045, "p90": 65.5840272860955, "p95": 65.58721384544896, "p99": 65.58976309293172, "stddev": 0.02388146120866979 }, Patterns benchmark also shows bit more of memory speeds. command: 'memory_benchmark -patterns -non-cacheable -count 5 -output patterns.JSON' Example M4 from 100 loops: "sequential_forward": { "bandwidth": { "read_gb_s": { "statistics": { "average": 116.38363691482549, "max": 116.61212708384109, "median": 116.41264548721367, "min": 115.449510036971, "p90": 116.54143114134801, "p95": 116.57314206456576, "p99": 116.60095068065866, "stddev": 0.17026641589059727 } } } } "strided_4096": { "bandwidth": { "read_gb_s": { "statistics": { "average": 26.460392735220456, "max": 27.7722419653915, "median": 26.457051473208285, "min": 25.519925729459107, "p90": 27.105171215736604, "p95": 27.190715938337473, "p99": 27.360449534513144, "stddev": 0.4730857335572576 } } } } "random": { "bandwidth": { "read_gb_s": { "statistics": { "average": 26.71367836895143, "max": 26.966820487564327, "median": 26.69907406197067, "min": 26.49374804466308, "p90": 26.845236287807374, "p95": 26.882004355057887, "p99": 26.95742242818151, "stddev": 0.09600564296001704 } } } } Thank you for reading :)
Does the same job
all alternatives →- OSOpen-source engine running Gemma 4 26B in 2 GB RAM on any M-series MacJul 2026 · github.com · ▲919
Hi HN, I built a specialized inference engine for running 4-bit Gemma 4 26B-A4B-IT on any M-series Mac using about 2 GB of RAM. It is called TurboFieldfare and is written in Swift and Metal. I have always adored on-device AI. It feels like magic that you can run a powerful NN on your Mac or iPhone. So I wanted to push the limits a bit and run a model whose weights don’t fit in memory. The model’s 4-bit quantized weights occupy roughly 14 GB, which makes running it with conventional inference tools almost impossible on an 8 GB or even 16 GB Mac once the OS, applications, and KV cache are…
- G4Gemma 4 Multimodal Fine-Tuner for Apple SiliconApr 2026 · github.com · ▲235
About six months ago, I started working on a project to fine-tune Whisper locally on my M2 Ultra Mac Studio with a limited compute budget. I got into it. The problem I had at the time was I had 15,000 hours of audio data in Google Cloud Storage, and there was no way I could fit all the audio onto my local machine, so I built a system to stream data from my GCS to my machine during training. Gemma 3n came out, so I added that. Kinda went nuts, tbh. Then I put it on the shelf. When Gemma 4 came out a few days ago, I dusted it off, cleaned it up, broke out the Gemma part from the Whisper…
More dev tools this month
the category →
Open-source GTM skills for technical founders
Dev tools · 29d ago · gtmcofounder.com
OpenTrailPaper is open-source bike computer firmware for the LilyGO T5S3 4.7" E-Paper PRO. It supports offline maps, GPX routes, FIT recording and Bluetooth sensors.
Dev tools · 2d ago · opentrailpaper.com
Launched alongside, January 2026
the whole month →- IN
Hey HN! I wanted to share something I built over the last few weeks: isometric.nyc is a massive isometric pixel art map of NYC, built with nano banana and coding agents. I didn't write a single line of code. Of course no-code doesn't mean no-engineering. This project took a lot more manual labor than I'd hoped! I wrote a deep dive on the workflow and some thoughts about the future of AI coding and creativity: http://cannoneyed.com/projects/isometric-nyc
AI · Jan 2026 · cannoneyed.com
Automatic AI-powered code reviews the moment you open a PR
Dev tools · Jan 2026 · kilo.ai
