Slicast
Thursday, October 1, 2026
DarkSubscribe
Slicast
AI Infrastructure · News & Analysis
Home › Chips & Hardware › Report
Chips & Hardware · Report

OpenAI is launching Ultrafast mode for GPT-5.6 Sol inference, achieving real-time speed of 750 output tokens per second powered by Cerebras hardware, with pricing and availability TBA.

Cerebras' custom AI silicon is proving productizable at scale; inference pricing models and hardware availability timelines will reshape software economics within six months.
Trade pressSlicast · August 14, 2026 at 15:54 UTC · US · Source: Tech Times
importance 65
Read the original
Companies in this story
OpenAI
Cerebras
In-depth analysis
CommentaryOpenAI Agent Security Breach, September 2026CommentaryWhy OpenAI's $122 Billion Capital Raise Marks Stargate's Shift to Compute Utility, September 2026
Related reports
Chips & HardwareBroadcom secures multi-year AI chip deals with OpenAI, Google, and Meta—locks networking/custom silicon supply across all major hyperscalers.2026-08-14Chips & HardwareCerebras fast inference cloud business nearly quadruples Q2; positioned to own fast inference market segment.2026-08-13Chips & HardwareOpenAI deploys a new ultrafast inference mode utilizing Cerebras WSE-3 chips instead of traditional Nvidia GPUs.2026-08-17Chips & HardwareNVIDIA is partnering with SB Energy and OpenAI to secure land, power, and shell infrastructure at the PORTS-Pike campus 2026-09-07Chips & HardwareNvidia is building a 1-trillion-parameter Nemotron 4 model to compete with large proprietary AI models from OpenAI and other labs.2026-08-12
OpenAI is launching Ultrafast mode for GPT-5.6… · Slicast