¹ NV H100 comparison based on https://nvidia.github.io/TensorRT-LLM/performance.html#h100-gpus-fp8 , March 28, 2024. Reported numbers are per GPU. Vs Intel® Gaudi® 3 projections for LLAMA2-7B, LLAMA2-70B & Falcon 180B projections. Results may vary.
² NV H100 comparison based on https://nvidia.github.io/TensorRT-LLM/performance.html#h100-gpus-fp8 , March 28, 2024. Reported numbers are per GPU. Vs Intel® Gaudi® 3 projections for LLAMA2-7B, LLAMA2-70B & Falcon 180B. Power efficiency for both Nvidia and Gaudi 3 based on internal estimates. Results may vary.
³ NV H100 comparison based on: https://developer.nvidia.com/deep-learning-performance-training-inference/training, March 28, 2024. “Large Language Model” tab vs. Intel® Gaudi® 3 projections for LLAMA2-7B, LLAMA2-13B & GPT3-175B as of 3/28/2024. Results may vary.
⁴ Based on architectural projections as of Feb. 14, 2023, vs. prior generation platforms. Your results may vary.
⁵ Based on architectural projections as of Feb. 14, 2023, vs. prior generation platforms. Your results may vary. ​
⁶ Based on architectural projections as of Feb. 14, 2023, vs. prior generation platforms. Your results may vary.
⁷ See Vision 2024 section of intel.com/performanceindex for workloads and configurations. Results may vary.

© Intel Corporation. Intel, the Intel logo and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.

View source version on businesswire.com: https://www.businesswire.com/news/home/20240409077438/en/

Danielle Mann
1-973-997-1154
danielle.mann@intel.com