OpenAI · GPT · AMD · Broadcom · OpenAI
Progress in AI compounds fastest when the entire system improves together
Compiled by KHAO Editorial — aggregated from 1 source. See llms.txt for citation guidance.
★ Tier-1 Source
Better software makes hardware more productive.
Key facts
- On the Artificial Analysis Coding Agent Index, GPT‑5.6 Sol with max reasoning reached a new high while using 54% fewer output tokens than another leading model
- Today, their portfolio also includes AWS, AMD, Broadcom, Cerebras, CoreWeave, Oracle, SB Energy and SoftBank
- Progress in AI compounds fastest when the entire system improves together
- On InferenceX, a public benchmark using GPT‑OSS 120B, Jalapeño delivered more peak throughput per kilowatt and lower token latency than the commercial systems in the comparison
Summary
Today, they shared the first measured performance results from Jalapeño, OpenAI’s first custom inference chip. Jalapeño gives them greater control over how their models run and over the economics of serving them. Their goal is to stay on the Pareto frontier: continually seeking the strongest mix of capability, speed, reliability, efficiency, and cost for each workload. Their portfolio gives them the range to meet those needs.