Unverified50% confidenceBenchmarkExact time
TensorFold's build runs a hybrid 2/3-bit model on an 80-core M5 Ultra chip at 70-105 tokens/s written; the MLX route achieves 26-29 tokens/s written on 64 cores.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/31/2026
First Seen
Valid until: 1/29/2027
Sources
Related Entities
Related Claims
Unverified在 vLLM 或 llama.cpp 等引擎的 Tensor Parallel (TP=2) 模式下,PCIe 4.0 x8/x8 不会显著卡住单流 token 生成。66% similarUnverifiedRTX 5080搭载GB203芯片,提供更高密度的FP8 Tensor Core,并采用GDDR7显存66% similarUnverifiedThe same llama.cpp build shows 'has tensor = false' on the M3 Ultra and 'has tensor = true' on the M5 Ultra, indicating that llama.cpp has added a dedicated code path for the M5 generation.65% similarVerifiedNVIDIA designed a dedicated Tensor Core instruction mma.kind::mxf4 for NVFP4 in the Blackwell architecture (SM 100/103, such as the B100 and B200 series), and adopted an FP8 scaling factor mechanism with one scale per group of 16 elements.64% similarUnverifiedJohn Rocky's Apple Silicon LLM benchmark recorded NVIDIA Nemotron 3 Nano at 159.7 tokens/s on MLX and 86.2 tokens/s on LLAMA.cpp, but the quantization of the two runs was not consistent.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1121821API
curl https://kongchang.com/api/v1/knowledge/claims/1121821MCP
get_claim(id=1121821)