Unverified50% confidenceFactExact time
The authors explicitly state that GPU-side measurements (how much FP8 KV-cache can recover on AMD MI300X) have not yet been run, and there are no GPU numbers anywhere in the paper.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/19/2026
First Seen
Valid until: 11/2/2026
Sources
Related Entities
Related Claims
UnverifiedThe standard version of TensorFlow cannot directly recognize and utilize the M1's GPU.65% similarUnverifiedA machine without a dedicated GPU cannot ingest large prompts in a reasonable time, so a second host with a GPU is needed to solve the prefill problem.64% similarUnverifiedWhen only tensorflow-macos is installed, no GPU device appears in the device enumeration and computation can only run on the CPU; only after correctly installing tensorflow-metal will TensorFlow detect the Metal GPU device.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1067582API
curl https://kongchang.com/api/v1/knowledge/claims/1067582MCP
get_claim(id=1067582)