Karan Verma

Karan serves as a Member of Technical Staff at AMD, focusing on performance for LLMs and recommendation systems on AMD Instinct™ GPUs. On MLPerf, his performance-architecture work spans workload tuning, rigorous benchmarking, and performance forecasting. He also owns CI/CD workflows that give the team dependable, automated benchmarking pipelines.

Posts by Karan Verma

April 01, 2026

Reproducing the AMD MLPerf Inference v6.0 Submission Result

Provide instructions to potential customers and partners to verify our MLPerf Inference v6.0 submission result.

https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inf_v6.0-repro/README.html

April 01, 2026

AMD Instinct™ GPUs MLPerf Inference v6.0 Submission

In this blog, we share the technical details of how we accomplish the results in our MLPerf Inference v6.0 submission.

https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inference-v6.0/README.html

November 12, 2025

Technical Dive into AMD MLPerf Training v5.1 Submission

Learn about the technical details of how AMD achieved the results in the MLPerf Training v5.1 submission.

https://rocm.blogs.amd.com/artificial-intelligence/mlperf-training-v5.1/README.html

November 12, 2025

Reproducing AMD MLPerf Training v5.1 Submission Result

Learn how to reproduce AMD's MLPerf Training v5.1 submission result.

https://rocm.blogs.amd.com/artificial-intelligence/mlperf-training5.1-repro/README.html

September 09, 2025

Reproducing the AMD Instinct™ GPUs MLPerf Inference v5.1 Submission

In this blog, we will provide step by step instruction on how to reproduce AMD's MLPerf Inference v5.1 Submission

https://rocm.blogs.amd.com/artificial-intelligence/mlperf-inference5.1-repro/README.html

September 09, 2025

Slim Down Your Llama: Pruning & Fine-Tuning for Maximum Performance

This blog describes the technical details of how we prune and fine tune the Llama 3.1 405B model in our MLPerf Inference v5.1 submission.

https://rocm.blogs.amd.com/artificial-intelligence/mlperf-llama-pruning/README.html

June 04, 2025

AMD’s MLPerf Training Debut: Optimizing LLM Fine-Tuning with Instinct™ GPUs

Explore the techniques we used to improve the training performance on MI300X and MI325X in our MLPerf Training 5.0 submission.

https://rocm.blogs.amd.com/artificial-intelligence/mlperf-training-v5.0/README.html

June 04, 2025

Reproduce AMD's MLPerf Training v5.0 Submission Result with Instinct™ GPUs

Follow this step-by-step guide to reproduce AMDs MLPerf 5.0 Training Submission with Instinct GPUs using ROCm

https://rocm.blogs.amd.com/artificial-intelligence/reproduce-mlperf-training-v5.0/README.html

June 03, 2025

High-Throughput BERT-L Pre-Training on AMD Instinct™ GPUs: A Practical Guide

Learn how to optimize BERT-L training with mixed precision and Flash Attention v2 on AMD Instinct GPUs — follow our tested MLPerf-compliant step-by-step guide.

https://rocm.blogs.amd.com/artificial-intelligence/bert-training/README.html

April 02, 2025

Reproducing the AMD Instinct™ GPUs MLPerf Inference v5.0 Submission

A step-by-step guide to reproducing AMD’s MLPerf v5.0 results for Llama 2 70B & SDXL using ROCm on MI325X

https://rocm.blogs.amd.com/artificial-intelligence/reproducing-amd-mlperf-inference-submission/README.html