Sign Up
Stories
AMD Surpasses Nvidia in GPU Benchmarks
Share
AI Performance Race: Nvidia Leads
GPU Market's Imminent Surge
Intel and NVIDIA's AI Market Showdown
AI Chip Battle at CES
AMD Dominates Intel in AI
AMD's MI300X Outshines Nvidia's H100
Overview
API
AMD's MI300X outperforms Nvidia's H100 in benchmarks, showcasing a 30% performance boost. The performance comparison emphasizes AMD's use of FP16 and vLLM, refuting Nvidia's use of a limited set of inferencing workloads and vLLM instead of TensorRT-LLM. The article also highlights AMD's improved performance and reduced latency in comparison to H100 with TensorRT-LLM.
Ask a question
How do these benchmarks reflect the ongoing competition between AMD and Nvidia in the GPU market?
How might the performance difference between AMD's MI300X and Nvidia's H100 impact the choice of GPUs for data centers?
What implications does the shift towards FP8 from FP16 with TensorRT-LLM's closed system have for the broader AI and machine learning industry?
Article Frequency
0.2
0.4
0.6
0.8
1.0
Sep 2023
Oct 2023
Nov 2023
Coverage