Unverified50% confidenceBenchmarkExact time
Overall P90 tail latency decreased by 22%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Entities
Related Claims
VerifiedQuantization reduces FP32 (32-bit) weights to INT4 (4-bit), achieving a theoretical compression ratio of 8×, but introduces precision loss.58% similarUnverified从BF16到FP8内存占用减半57% similarUnverifiedOn GPU fp16 (RTX 5090), Largha-150M-v1 reports a p50 latency of 21.0 ms per decision.57% similarUnverifiedA 140 GB FP16 70B model can be compressed to approximately 35–40 GB after 4-bit quantization, reducing communication overhead by about 75%; 2-bit quantization can compress it to approximately 18 GB but introduces accuracy loss.56% similarUnverified13B–14B class models (4-bit quantization) see speeds drop to 2–5 tokens per second under this configuration56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1020720API
curl https://kongchang.com/api/v1/knowledge/claims/1020720MCP
get_claim(id=1020720)