General Compute signs multi-year Positron deal, claims up to 16.2x faster inference with GPU+ASIC split
testingcatalog · x · 2026-10-21
General Compute, a neocloud for alternative chips, announced a multi-year agreement to deploy Positron AI hardware, with first deployments live in Q1 2027.
- Atlas opens to General Compute customers in Q1 2027.
- Titan targets larger models and longer contexts using several terabytes of slower memory instead of high-bandwidth GPU memory, pending production of Positron's custom Asimov chip.
- The stack disaggregates inference: NVIDIA GPUs for prefill, ASICs for decode, claiming 2.8x throughput per rack and up to 16.2x faster end-to-end in open benchmarks.
- Also offers bare-metal Cerebras CS-3 systems for latency-sensitive agent and coding workloads.
Related event: General Compute partners with Positron AI for inference clusters(3 posts)→
More from Infra
- PyTorch details KTIR and Dataflow Scheduler behind IBM's Spyre accelerator backend — PyTorch · 2026-10-21
- Indie dev replaces $1k OpenAI bill with a YAML pipeline on spot GPUs and vLLM — cmm324 · 2026-10-21
- DeepSeek open-sources DeepInfer, the vLLM-based engine behind V4.1-Flash — The Decoder · 2026-10-21
- Robotics startup launches Deployments to serve fine-tuned HF checkpoints to robots in real time — arnie_hacker · 2026-10-21
- ByteDance, Alibaba, Tencent order 3M+ Nvidia RTX PRO 5500 cards worth $35B, Reuters reports — SumitGup · 2026-10-21
- TorchTPU demo: unified vLLM & SGLang backend for LLM serving on TPUs — Rob_Mulla · 2026-10-21