A Python package for simulating low precision arithmetic in scientific computing and machine learning
-
Updated
Sep 26, 2026 - Python
#
block-floating-point
Here are 2 public repositories matching this topic...
Fused and block-float ML kernels for Tenstorrent Tensix cores in TT-Lang and TT-Metalium C++, measured on Tenstorrent's simulators, with CI on simulated chips, containers, a Kubernetes JobSet sweep and Ansible provisioning.
docker kubernetes ansible simulator helm pytorch quantization github-actions gpt-2 kernel-fusion tenstorrent ttnn tt-metal tt-metalium block-floating-point tt-lang tensix jobset ttsim
-
Updated
Oct 8, 2026 - Python
Add this topic to your repo
To associate your repository with the block-floating-point topic, visit your repo's landing page and select "manage topics."