The AI company says its models improved multiple key performance metrics of AMD's flagship GPU by up to 14%.
OpenAI released its latest test results on Wednesday, demonstrating the growing capabilities of its AI models in the chip design field; the report showed that the designs of OpenAI's self-developed Jalapeño chip and Advanced Micro Devices (AMD)'s flagship graphics processor both achieved significant optimization.
In an official blog post, OpenAI stated that with the help of AI models, Advanced Micro Devices (AMD)'s flagship GPU achieved improvements of up to 14% across three core chip metrics: power consumption, performance, and chip area.
OpenAI did not disclose the baseline data used for this optimization, but described the result as "completing about a full year of R&D progress in less than 5 months."
The blog stated: "Advanced Micro Devices (AMD) has already applied some of the optimization changes to chip designs recently sent for tape-out, while the remaining optimizations have been written into production design files for subsequent iterations."
OpenAI also used AI models to optimize the design of its self-developed Jalapeño chip, focusing on re-laying out certain modules of the chip to reduce the chip's footprint without lowering processing capability.
This result builds on OpenAI's previously released preliminary test data for the Jalapeño chip.
Comments