User Profile

HugoAffaticati

Microsoft

Joined 3 years ago

16 Posts13 Likes

View All Badges

User Widgets

Recent Discussions

No content to show

Groups

Recent Blog Articles

Re: Optimizing Language Model Inference on Azure
Hello Dmonakhov, thank you for your interest in our benchmarking approach! The optimum values referenced in this post serve as examples to illustrate our optimization process. Variations in these v...
Nov 13, 2024 Place Azure High Performance Computing (HPC) Blog
0likes
1Comment
Optimizing Language Model Inference on Azure
Inefficient inference optimization can lead to skyrocketing costs for customers, making it crucial to establish clear performance benchmarking numbers. This blog sets the standard for expected perfor...
Nov 13, 2024 Place Azure High Performance Computing (HPC) Blog
1like
5Comments
A quick start guide to benchmarking AI models in Azure: Llama 2 from MLPerf Inference v4.0
Microsoft Azure has delivered industry-leading results for AI inference workloads amongst cloud service providers in the most recent MLPerf Inference results published publicly by MLCommons. The Azur...
Mar 27, 2024 Place Azure High Performance Computing (HPC) Blog
0likes
0Comments
A Quick Guide to Benchmarking AI Models on Azure: ResNet with MLPerf Training v3.0
Azure is pleased to showcase results from our MLPerf Training v3.0 submission. For this submission, we benchmarked ourND H100 v5 virtual machine (preview). In this blog post, you will learn how to r...
Nov 09, 2023 Place Azure High Performance Computing (HPC) Blog
0likes
0Comments
Re: Benchmarking the NC A100 v4, NCsv3, and NCas_T4_v3 series with NVIDIA Deep Learning Examples
Hellostkarlos, I believe that the problem is that the link is truncated on your command (presence of ...). You should try copying the full link with your wget command: "https://api.ngc.nvidia.com/v2...
Jul 05, 2023 Place Azure Compute Blog
0likes
0Comments
A quick start guide to benchmarking AI models in Azure: MLPerf Inference v2.1
Discover how to run MLPerf Inference v2.1 benchmarks on NC A100 v4, NDm A100 v4 and NVadsA10 v5 virtual machines.
Feb 13, 2023 Place Azure High Performance Computing (HPC) Blog
1like
0Comments
Accelerating AI applications using the JAX framework on Azure’s NDm A100 v4 Virtual Machines
The results highlight good scaling from 1 to 16 nodes on both the Large and XLarge T5 models running with JAX on Azure. The Large T5 model has a scaling efficiency of 84% at 16 nodes (128 GPUs...
Feb 07, 2023 Place Azure High Performance Computing (HPC) Blog
2likes
0Comments
A quick start guide to benchmarking LLM models in Azure: NVIDIA NeMo Megatron - Steps
Find here the steps one needs to take to run GPT-3 architecture models with NeMo Megatron on NDm A100 v4-series on Azure.
Jan 27, 2023 Place Azure High Performance Computing (HPC) Blog
0likes
0Comments
Tackling AI Inference workloads on Azure’s NC A100 v4 virtual machines with time to spare
The NC A100 v4-series offers great flexibility through MIG technology to handle different sizes of workload, from small to medium. While we compared the performance of a single MIG instance (1/7th of...
Jan 27, 2023 Place Azure High Performance Computing (HPC) Blog
1like
0Comments
A quick start guide to benchmarking AI models in Azure: MLPerf Inference v2.1 on Multi-Instance GPU
In this document, one will find the steps to run the MLPerf Inference v2.1 benchmarks for BERT, ResNet-50, RNN-T, and 3D-UNet on one of seven slices of NVIDIA-powered NC A100 v4-series Tensor Core GP...
Jan 27, 2023 Place Azure High Performance Computing (HPC) Blog
0likes
0Comments