Flexible Machine Learning
Gennady emphasizes the necessity of standard benchmarks in hardware development while highlighting the unique needs of individual models. He discusses how his company leverages machine learning compilers to adapt to the growing complexity of AI, aiming for optimal performance without the extensive manual effort traditionally required. The conversation reveals the challenges vendors face in promoting their hardware and the importance of flexibility in an ever-evolving market.In this clip
From this podcast
The Hard Part with Evan McCann
Gennady Pekhimenko from CentML
Related Questions