Unverified50% confidenceFactExact time
The author released an experimental version: running a small language model with 260,000 parameters as a C program on QNPUTER, with the weights loaded into CPU memory as data, takes about 41 seconds to generate each token.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/31/2026
First Seen
Valid until: 11/14/2026
Sources
Related Entities
Related Claims
UnverifiedWith the QNPUTER approach, on an NVIDIA L4 GPU a single calculator call takes half a second to one second using a dedicated acceleration program; if only ordinary model code is used, it is more than ten times slower.74% similarUnverifiedAfter writing the KV cache in version V3, the time to generate 100 tokens from a 512-token prompt dropped from 278 seconds to 28 seconds, achieving nearly a 10x inference speedup.73% similarUnverified通过NumPy向量化计算和Python的multiprocessing并行化可将5万次模拟运行时间从数分钟压缩至数秒71% similarUnverifiedv0.0.90 only writes entities that have changed and merges writes in batches, packaging them into a single transaction approximately every 300 milliseconds.70% similarUnverifiedMiniMax can read an entire project and output a complete project organization logic and paper work file in approximately 1 minute and 26 seconds.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1121157API
curl https://kongchang.com/api/v1/knowledge/claims/1121157MCP
get_claim(id=1121157)