tech
OpenAI says its Jalapeño chip can power faster AI responses than the competition
Posts from this topic will be added to your daily email digest and your homepage feed.
TL;DR
- OpenAI's new AI chip, Jalapeño, outperforms Nvidia's superchips on AI inference benchmark tests.
- Jalapeño offers lower latency and higher throughput, providing a balance typically not achieved in AI systems.
- The chip, an ASIC made with Broadcom, is designed for AI inference, running trained AI models.
- Tests show Jalapeño delivers 1.5 to 1.9 times more AI work per watt and 1.7 to 3.6 times lower latency than comparison systems across several models.
- OpenAI plans to deploy Jalapeño in small volumes by the end of the year, with increased volume in 2027.
- OpenAI will continue to partner with companies like Nvidia and develop future generations of Jalapeño.