Tag
OpenAI announces the first performance results of Jalapeño, its custom inference chip, which delivers better throughput and latency than commercial systems, as part of an integrated compute strategy.