Tag

Mlperf

All articles tagged with #mlperf

technology2 years ago•3 min saved

NVIDIA Dominates MLPerf Benchmarks with Hopper H100, GH200 Superchips & L4 GPUs

NVIDIA has released its MLPerf Inference v3.1 performance benchmarks, showcasing the dominance of its AI GPUs such as Hopper H100, GH200, and L4. The benchmarks cover a wide range of AI use cases, including recommender systems, natural language processing, speech recognition, image classification, medical imaging, and object detection. NVIDIA's H100 outperformed competitors across all workloads, while the GH200 Superchip demonstrated a 17% improvement over the H100. The GH200's performance gain is attributed to higher VRAM capacities and bandwidth. Additionally, the Ada Lovelace-based L4 GPU ran all workloads efficiently and significantly outperformed x86 CPUs, while the Jetson Orion received an 84% performance boost through software updates.

via Wccftech|

#ai #gpus #mlperf

technology2 years ago•3 min saved

"Intel's Gaudi2 Chip Challenges NVIDIA GPUs in MLPerf Benchmarks"

Intel's Gaudi2 chip has emerged as the only alternative to NVIDIA GPUs for training large language models (LLMs) according to MLPerf benchmarks. Intel claims that Gaudi2 offers similar price/performance to NVIDIA's A100 in FP16 workloads and aims to surpass NVIDIA's H100 by September in FP8 workloads. The benchmarks show impressive time-to-train results for GPT-3 and other models, indicating the growing software maturity of Gaudi2. Additionally, Intel's Xeon Platinum CPUs also demonstrated strong performance in LLM training, achieving training times of less than 50 minutes for BERT and ResNet-50 models.

via Wccftech|

#gaudi2 #intel #llm-training