Channel3 cut AI cost per product by 10× by moving latency-tolerant stages of its product graph pipeline to asynchronous inference on Sail, the company and Sail reported. Most jobs run on Gemma 4 31B t…
In a blog post, the team at an AI infrastructure company detailed their optimization of Gemma 4 31B on Google's TPU v6e, boosting prefill efficiency from ~32% to ~63% MFU. They highlighted the chip's …
Intel Gaudi 3's vLLM-Gaudi PagedAttention implementation underutilizes hardware for sliding window attention, but optimizations for serving Gemma 4 31B expanded usable KV cache by 3.5×, raised decode …
Sail Research has open-sourced HTDYM (How To Deploy Your Model), an internal performance modeling infrastructure that helps determine the optimal hardware and sharding configuration for serving large …
Sailboxes, a cloud-environment designed for long-horizon AI agents, is now generally available after a beta period, offering pricing over 70% lower than competitors with no runtime limits and full Lin…
Sail, an infrastructure startup for long-horizon AI agents, announced an $80M seed and Series A funding round led by Sequoia and Kleiner Perkins. The company builds inference servers and sandboxes opt…
Sail Research launched Voyages, an agent telemetry SDK that provides a system of record for long-horizon agentic workflows, recording traces of model calls, sandbox commands, and errors. The tool inte…