Unverified50% confidenceBenchmarkExact time
When 32 requests share the same 4k-token system prompt, Superfluid's prefix cache hit rate reaches 97%, on par with Ollama, while mlx_lm.server is 0%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Entities
Related Claims
UnverifiedReal-world testing with the 284-billion-parameter DeepSeek V4 Flash shows token generation rising from 37 to 53 tokens/s (approximately 1.5x), and Prompt Processing jumping from 483 to 1,485 tokens/s (approximately three times).67% similarUnverifiedIn a text generation test using an 8-billion-parameter model, the M5 Ultra achieved 188 tokens per second versus 129 tokens per second for the M3 Ultra, representing only 1.45x.64% similarUnverifiedOn a Korean NPU server (8 cards, 32 chips in total), VKORE raises the single-user response of the Gemma 4 26B model from 14.1 tokens/s to 53.9 tokens/s, about 3.8 times.64% similarUnverifiedDeveloper Primitive reported: 91.2 tokens/s with speculative decoding off, reaching 142.6 tokens/s after enabling 3 speculative tokens, approximately a 56% speedup under the same configuration.63% similarUnverifiedWhen running 4-bit DeepSeek R1 on the M3 Ultra, a 1,000-token context generates approximately 19 tokens per second using 392GB, dropping to approximately 15 tokens per second using 410GB at a 16,000-token context.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1013613API
curl https://kongchang.com/api/v1/knowledge/claims/1013613MCP
get_claim(id=1013613)