But it’s worth clarifying that Neuchips and SiMa outperform Nvidia in terms of performance per watt compared to the Nvidia H100 and Jetson AGX Orin, respectively. The Qualcomm Cloud AI100 accelerator also performed well in terms of energy efficiency compared to the Nvidia H100 in some scenarios. In MLPerf Inference 3.0, the test suite has not changed, but a new script has been added – network.
Additionally, improved inference results were provided for Bert-Large, which is particularly interesting as it is closest in nature to major language models (LLMs) such as ChatGPT.
MLPerf Inference benchmarking is required to evaluate the performance of AI models in various scenarios. The final round showed benchmark participants’ achievements in terms of performance, energy efficiency and software optimization.
Source: Ferra
I am a professional journalist and content creator with extensive experience writing for news websites. I currently work as an author at Gadget Onus, where I specialize in covering hot news topics. My written pieces have been published on some of the biggest media outlets around the world, including The Guardian and BBC News.
