First Benchmarks Revealed for Jalapeño, OpenAI’s Clean-Sheet General Purpose AI Accelerator ASIC
By Nitin Dahad, EE Times | August 27, 2026

At a pre-Hot Chips media briefing this week, OpenAI’s VP of hardware, Richard Ho, revealed results of first benchmarks for its clean‑sheet ASIC designed specifically for large language models (LLMs). The company’s in‑house “Jalapeño” accelerator, developed with Broadcom and Celestica, targets both high throughput and low latency in the same architecture—while addressing what Ho framed as the real enemy in modern AI systems: data movement.
OpenAI announced the chip in June, stating it was built from the ground up for current and future LLMs across the industry, developed from design to production in nine months (which itself was accelerated by OpenAI’s models).
This week, the company revealed that the tests of the chip and the systems around it showed significant performance advances, with Jalapeño serving more AI work per unit of power, while also returning responses more quickly. “Jalapeño delivers both higher throughput and lower latency with one architecture, where existing hardware systems often have to make a tradeoff between the two,” OpenAI said in an official statement.
To read the full article, click here
Related Semiconductor IP
- GPU
- B-Series GPU IP
- Arm’s flagship GPU providing ultimate mobile gaming experiences
- Real-Time GPU IP for Path Tracing
- A-Series GPU IP
Related News
- SEMIFIVE Secures USD 52 Million AI Accelerator Contract with U.S. Fabless Company, Validating End-to-End ASIC Model in North America
- Aion Silicon Joins Intel Foundry Accelerator Value Chain Alliance to Design and Deliver Best-in-Class ASIC and SOC Solutions
- SEMIFIVE Commences Mass Production of HyperAccel’s LLM AI Inference Accelerator ‘Bertha’ on Samsung 4nm, Spurring Growth Momentum
- Processing Neuromorphic AI with INTERA’s EAGLE Cognit Accelerator
Latest News
- Cyient Semiconductors Secures Strategic Investment from GlobalFoundries
- JEDEC® Previews Next-Generation DDR6 DRAM: Higher Performance, Greater Capacity, Improved Power Efficiency and Data-Center-Class Reliability
- INTELPRO Adopts Andes AX46MPV Flagship Multicore Vector Processor and AndesAIRE™ AnDLA™ I370 to Power Next-Generation Heterogeneous AI and Robotics Solutions
- VeriSilicon Introduces Low-Power, High-Performance Wearable Subsystem Solution for Next-Generation AI/AR Glasses
- IC Manage Announces MCP Server for GDP-AI Design and IP Management System