MLCommons Unveiled New Benchmark Speed Test for Running AI Models

ODSC - Open Data Science
2 min readSep 14


On Monday, an AI benchmark group called MLCommons unveiled the results of some new tests that determine how quickly the best hardware can run AI Models. The report via Reuters shows that the top performer for these tests was a chip from Nvidia.

The MLCommons test was on a large language model and it had a chip from rival Intel coming in at second in the benchmark results. The MLCommons model benchmark in question is called the MLPerf benchmark. It is based on an LLM with 6 billion parameters that can summarize articles from CNN.

Nvidia got the top spot using a build of eight of its flagship H100 chips. However, its performance is of no surprise as the chipmaking giant has dominated the market for some time. Speaking of the results, Nvidia’s accelerated computing marketing director, Dave Salvator, said, “What you see is that we’re delivering leadership performance across the board, and again, delivering that leadership performance on all workloads.”

For second place. Intel was able to take the runner-up spot using its Gaudi2 chips. These chips are produced by one of their most recently acquired companies Habana. The acquisition came in 2019, and its chips were about 10% slower than Nvidia’s build.

Speaking of Gaudi2’s performance during the tests, Habana’s chief operating officer, Eitan Medina, said “We’re very proud of the results of inferencing, (as) we show the price performance advantage of Gaudi2.”

While speaking with Reuters, Intel commented on how its build is cheaper than Nvidia’s which from their claims is around the price of their rival’s last generation 100 systems. But it wasn’t just Intel and Nvidia looking to test their hardware.

During the tests, Google previewed its chip build. This was announced back in August but shows that the semiconductor industry, though dominated by a few names, has some emerging rivals. This would make sense as the market share for the industry is expected to see double-digit growth.

According to Fortune Business Insights, the semiconductor market share will likely see itself grow past a trillion dollars by 2029. Much of this is driven by demand for new electronic devices, AI-powered tools, the growth of the Internet of Things technology, and more.

Originally posted on

Read more data science articles on, including tutorials and guides from beginner to advanced levels! Subscribe to our weekly newsletter here and receive the latest news every Thursday. You can also get data science training on-demand wherever you are with our Ai+ Training platform. Interested in attending an ODSC event? Learn more about our upcoming events here.



ODSC - Open Data Science

Our passion is bringing thousands of the best and brightest data scientists together under one roof for an incredible learning and networking experience.