Gimlet and Cerebras plan 100 MW capacity, targeting 3,000 tokens/sec inference
Sethwinterroth · x · 2026-09-29
Gimlet Labs is deepening its partnership with Cerebras: the two already serve inference traffic together and are now planning 100 MW of capacity, with the first Cerebras-powered datacenter expected this year and speeds up to 3,000 tokens/sec. Gimlet's pitch is inference disaggregation—mapping each phase of inference to the silicon best suited to it—expanding the throughput and interactivity frontier through a single API, enabling products built around agents finishing in seconds.
Related event: Gimlet Partners with Cerebras on 100MW Inference Cloud(2 posts)→
More from Infra
- India's Sarvam launches Token Factory, hosting top open models on local infra — HarveenChadha · 2026-11-16
- Gemini 3.5 Flash now runs on air-gapped Google Distributed Cloud, offline up to 12 months — shashib · 2026-11-16
- Gemma.c: full Gemma 1/2/3 multimodal inference in a single C file, ~72 tok/s on CPU at INT8 — This-Purchase-3325 · 2026-11-16
- BIS Issues Interim Final Rule Addressing Remote Access to Advanced AI Chips — apples_jimmy · 2026-11-16
- Jailbroken PS4 runs a 9B open-source model at up to 43 tok/s by talking directly to its 2013-era GPU — AIFlow_ML · 2026-11-16
- Mobile AI Factory: agents auto-build on-device SLMs with 19x less GPU time — pmttyji · 2026-11-16