Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Prefill Latency Measurement on LLaMA 8B Instruct 16K context length 3.1
Loading...
1,224
Attention Prefill Latency
Dense
1,164.94
1,563.595
1,962.25
2,360.905
May 16, 2026
Attention Prefill Latency
End-to-End Prefill Latency
Updated 2mo ago
Evaluation Results
Method
Method
Links
Attention Prefill Latency
End-to-End Prefill Latency
Dense
GPU=RTX PRO 6000, TP=2...
2026.05
1,224
4,858
FlashPrefill
GPU=RTX PRO 6000, TP=2...
2026.05
1,356.2
4,978.7
CompactAttention-FP
GPU=RTX PRO 6000, TP=2...
2026.05
1,393.6
4,953.2
QUOKA
GPU=RTX PRO 6000, TP=2...
2026.05
1,599.5
5,156.3
SeerAttention
GPU=RTX PRO 6000, TP=2...
2026.05
2,548.7
6,286.4
CompactAttention-SA
GPU=RTX PRO 6000, TP=2...
2026.05
2,675.7
6,382.9
XAttention
GPU=RTX PRO 6000, TP=2...
2026.05
2,700.5
6,243.9
Feedback
Search any
task
Search any
task