AI Inference
-
OpenAI Unveils Jalapeño: A Custom ASIC Designed to Redefine AI Inference Efficiency
OpenAI has debuted Jalapeño, a custom inference-focused ASIC developed with Broadcom to boost throughput and energy efficiency for large-scale AI workloads.
-
AMD and Cerebras Partner to Redefine AI Inference Architecture
AMD and Cerebras are launching a disaggregated AI inference solution, pairing AMD’s Helios systems with Cerebras’ wafer-scale engines to boost latency and throughput.
-
Nvidia’s New AI Inference Chip Integrates Groq LPU Technology
Nvidia will unveil a new AI inference chip at its GTC conference, featuring integrated LPU technology from Groq. This strategic move aims to boost AI processing and strengthen Nvidia’s market leadership.
-
OpenAI Boosts Compute with $10 Billion Cerebras Deal
OpenAI has signed a $10 billion deal with chipmaker Cerebras Systems to significantly enhance its AI inference capacity, aiming to accelerate ChatGPT’s response times and support its rapid revenue growth, which hit $20 billion in 2025.