Shadowgraph
TrendingAbout
Sign in

okuvshynov/experiments

C++

2

492 commits

updated Aug 27, 2026

See the code

README

experiments

m1 cores

This is a test to illustrate the difference between the m1 cores (icestorm and firestorm), and how different ways to generate the code would be preferrable for each core.

Requires b63 and runs on Apple M1. R with ggplot is needed for plotting the summary.

rtdigest

T-Digest implementation which keeps track of min/max instead of mean, thus can return the interval for the quantile.

bswap

Experimenting with partial byte swaps

Contributors

okuvshynov

492 commits

Similar repos

CodeHalwell/rotquant

0

tzakharko/m4-sme-exploration

Exploring the scalable matrix extension of the Apple M4 processor

244

argonautlabsai/argodrive

Layout, balancer and instruments for running mixture-of-experts models from SSDs. Three models, two…

4

corsix/fast-crc32

Fast CRC32 implementations

141

apple/ml-stable-diffusion

Stable Diffusion with Core ML on Apple Silicon

17,967

powturbo/Turbo-Range-Coder

TurboRC - Fastest Range Coder + Arithmetic Coding / Fastest Asymmetric Numeral Systems

96

philipturner/metal-benchmarks

Apple GPU microarchitecture

632

JustVugg/colibri

Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from…

37,151

Languages

C++

34.1%

Python

28.6%

Swift

13.5%

JavaScript

7.8%

C

5.2%

HTML

3.6%

Go

1.6%

Shell

1.5%

okuvshynov/experiments

C++

2

492 commits

updated Aug 27, 2026

See the code

README

experiments

m1 cores

This is a test to illustrate the difference between the m1 cores (icestorm and firestorm), and how different ways to generate the code would be preferrable for each core.

Requires b63 and runs on Apple M1. R with ggplot is needed for plotting the summary.

rtdigest

T-Digest implementation which keeps track of min/max instead of mean, thus can return the interval for the quantile.

bswap

Experimenting with partial byte swaps

Similar repos

CodeHalwell/rotquant

0

tzakharko/m4-sme-exploration

Exploring the scalable matrix extension of the Apple M4 processor

244

argonautlabsai/argodrive

Layout, balancer and instruments for running mixture-of-experts models from SSDs. Three models, two…

4

corsix/fast-crc32

Fast CRC32 implementations

141

apple/ml-stable-diffusion

Stable Diffusion with Core ML on Apple Silicon

17,967

powturbo/Turbo-Range-Coder

TurboRC - Fastest Range Coder + Arithmetic Coding / Fastest Asymmetric Numeral Systems

96

philipturner/metal-benchmarks

Apple GPU microarchitecture

632

JustVugg/colibri

Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from…

37,151

Contributors

okuvshynov

492 commits

Languages

C++

34.1%

Python

28.6%

Swift

13.5%

JavaScript

7.8%

C

5.2%

HTML

3.6%

Go

1.6%

Shell

1.5%