Daily AI news, sourcedWhere it stands
As of Oct 20, Shishi Technology launched TokenFleet, a distributed scheduling framework for GPU inference clusters, at the 2026 CCF national high-performance computing conference, QbitAI reports.
- Reported by QbitAI
Shishi Technology launches TokenFleet, an inference scheduler it says cuts P95 time to first token by 32.4%, per QbitAI
Not confirmed by the company — reported on Oct 20, 2026 (source: qbitai.com).
Shishi Technology launched TokenFleet, a distributed scheduling framework for GPU inference clusters, at the 2026 CCF national high-performance computing conference, QbitAI reports. It has three parts: TokenPilot picks an instance by weighing KV cache reuse against queueing delay, FleetMind coordinates capacity across data centers, and SLOMax handles admission and queues for mixed workloads. In the company's own GLM5.3 tuning data on its test cluster, predicted-latency routing cut P95 time to first token by 32.4% and P50 by 71.3%. The results are self-reported.
Source: QbitAI
- Official
- An institution announces something that has already happened, through its own channel or a founder or executive account.
- Statement by …
- A named person's views, predictions, plans or comments, including remarks reported by media when the original is unavailable.
- Reported by …
- Only media or another third party is saying it. We name the outlet.
- Unconfirmed
- Circulating online (posts, comments, leaks). No company, named person or outlet has confirmed it.
Items older than 72 hours never go in Top stories and show their original date.