2020-05-19-bert-record.md

---
layout: single
title: "DeepSpeed optimizes transformer kernels to achieve the world's fastest and most efficient BERT training record: 44 minutes on 1024 NVIDIA V100 GPUs"
excerpt: ""
categories: news
new_post: true
date: 2020-05-19 00:00:00
---

We introduce new technology to accelerate single GPU performance via
kernel optimizations. These optimizations not only create a strong
foundation for scaling out large models, but also improve the single GPU
performance of highly tuned and moderately sized models like BERT by more
than 30%, reaching a staggering performance of 66 teraflops per V100 GPU,
which is 52% of the hardware peak. **Using these optimizations as the building
block, DeepSpeed achieves the fastest BERT training record: 44 minutes on
1,024 NVIDIA V100 GPUs**, compared with the best published result
of 67 minutes on the same number and generation of GPUs.

**Code and tutorials are coming soon!**

For a technical overview, see our [blog post](linklink).