Unverified50% confidenceFactExact time
Community quantized versions (AMD, Intel, Unsloth) have reached about 2-bit, and are reportedly runnable with llama.cpp on a 256GB machine.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/16/2026
First Seen
Valid until: 10/30/2026
Sources
Related Entities
Related Claims
Unverified该1-bit量化的27B模型能够在仅8GB内存的设备上运行68% similarUnverifiedAMD's Ryzen AI Max chip offers up to 128 GB of unified memory in Windows mini PCs and laptops.67% similarUnverifiedThe AMD Radeon AI Pro R9700 and Intel Arc Pro B70 both feature 32GB of VRAM and a 256-bit memory bus width, with memory bandwidth differing by less than 5%.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1042239API
curl https://kongchang.com/api/v1/knowledge/claims/1042239MCP
get_claim(id=1042239)