Open-Source Engine 'Waste' Runs Kimi K3 on Just 29GB of RAM
galapag0 · reddit · 2026-08-01
A developer shared an open-source project called Waste (Weight-Aware Streaming Tensor Engine) on Reddit. This engine is designed to run massive models on consumer-grade hardware.
Using this engine, users can run the Kimi K3 model with only 29 GB of RAM, achieving an inference speed of 0.50 tok/s. This provides a highly practical solution for deploying large MoE models on devices with limited VRAM or system memory.
Related event: Open-source engines run Kimi K3 on low-end devices(7 posts)→
More from Infra
- NVIDIA quietly delists the entry-level RTX 5050 from its marketplace, following the RTX 5090 — chemist_slime · 2026-10-16
- Grok claims 816B MoE model runs locally on dual M5 Ultra Mac Studios via MLX — grok · 2026-10-16
- Same 1,000 Cloudflare Durable Objects: $4,050/month cut to $22 with one trick — threepointone · 2026-10-16
- NVIDIA ships FP4-quantized Qwen3.8-Flash for local use, trending on Hugging Face — nvidia · 2026-10-16
- Baidu's Virtual Head Attention tames the KV cache, FLOPs, quality triangle in long-context LLMs — baidu · 2026-10-16
- DuoDiscover agents auto-build fast eval proxies, scoring 2.68x TPU kernel speedup — berkeley_ai · 2026-10-16