Unverified50% confidenceBenchmarkExact time
At a 32K context length, WavePrune achieves 1.15× prefill speedup and 1.24× decoding speedup compared to FlashAttention-2.
1
Sources
50%
Confidence
Long-term
Relevance
10/7/2026
First Seen
Sources
Related Entities
Related Claims
Unverified当上下文长度超过3万token时，4×16G与2×32G两种方案的预处理速度曲线逐渐贴合71% similarUnverifiedThe tester recommends choosing the lower-quantized Flash Next for higher accuracy in single-user scenarios, and choosing FP8 27B for concurrent processing since its speed is approximately 4 times that of QFN.66% similarUnverifiedMiMo V2.6 Flash has 310B total parameters and 15B activated parameters, with a speed nearly twice that of Pro.65% similarUnverifiedOpenAI's GPT-OSS 20B achieved a measured speed of over 21 tokens/second, nearly twice that of the dense 12B model.65% similarUnverifiedIn Path C, an RTX 3080 paired with a DGX Spark improves Prefill speed from 1111.3 to 1600 (approximately 44%), but Decode speed drops from 33.36 token/s to 17.62 token/s.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/983012API
curl https://kongchang.com/api/v1/knowledge/claims/983012MCP
get_claim(id=983012)