Monday, May 19, 2025
LBNN
  • Business
  • Markets
  • Politics
  • Crypto
  • Finance
  • Energy
  • Technology
  • Taxes
  • Creator Economy
  • Wealth Management
  • Documentaries
No Result
View All Result
LBNN

AI Training:Newest Google and Nvidia Chips Speed AI Training

Simon Osuji by Simon Osuji
November 13, 2024
in Artificial Intelligence
0
AI Training:Newest Google and Nvidia Chips Speed AI Training
0
SHARES
1
VIEWS
Share on FacebookShare on Twitter


Nvidia, Oracle, Google, Dell and 13 other companies reported how long it takes their computers to train the key neural networks in use today. Among those results were the first glimpse of Nvidia’s next generation GPU, the B200, and Google’s upcoming accelerator, called Trillium. The B200 posted a doubling of performance on some tests versus today’s workhorse Nvidia chip, the H100. And Trillium delivered nearly a four-fold boost over the chip Google tested in 2023.

The benchmark tests, called MLPerf v4.1, consist of six tasks: recommendation, the pre-training of the large language models (LLM) GPT-3 and BERT-large, the fine tuning of the Llama 2 70B large language model, object detection, graph node classification, and image generation.

Training GPT-3 is such a mammoth task that it’d be impractical to do the whole thing just to deliver a benchmark. Instead, the test is to train it to a point that experts have determined means it is likely to reach the goal if you kept going. For Llama 2 70B, the goal is not to train the LLM from scratch, but to take an already trained model and fine-tune it so it’s specialized in a particular expertise—in this case,government documents. Graph node classification is a type of machine learning used in fraud detection and drug discovery.

As what’s important in AI has evolved, mostly toward using generative AI, the set of tests has changed. This latest version of MLPerf marks a complete changeover in what’s being tested since the benchmark effort began. “At this point all of the original benchmarks have been phased out,” says David Kanter, who leads the benchmark effort at MLCommons. In the previous round it was taking mere seconds to perform some of the benchmarks.

A line graph with one diagonal blue line and many colored and dashed branches rising up from that line.Performance of the best machine learning systems on various benchmarks has outpaced what would be expected if gains were solely from Moore’s Law [blue line]. Solid line represent current benchmarks. Dashed lines represent benchmarks that have now been retired, because they are no longer industrially relevant.MLCommons

According to MLPerf’s calculations, AI training on the new suite of benchmarks is improving at about twice the rate one would expect from Moore’s Law. As the years have gone on, results have plateaued more quickly than they did at the start of MLPerf’s reign. Kanter attributes this mostly to the fact that companies have figured out how to do the benchmark tests on very large systems. Over time, Nvidia, Google, and others have developed software and network technology that allows for near linear scaling—doubling the processors cuts training time roughly in half.

First Nvidia Blackwell training results

This round marked the first training tests for Nvidia’s next GPU architecture, called Blackwell. For the GPT-3 training and LLM fine-tuning, the Blackwell (B200) roughly doubled the performance of the H100 on a per-GPU basis. The gains were a little less robust but still substantial for recommender systems and image generation—64 percent and 62 percent, respectively.

The Blackwell architecture, embodied in the Nvidia B200 GPU, continues an ongoing trend toward using less and less precise numbers to speed up AI. For certain parts of transformer neural networks such as ChatGPT, Llama2, and Stable Diffusion, the Nvidia H100 and H200 use 8-bit floating point numbers. The B200 brings that down to just 4 bits.

Google debuts 6th gen hardware

Google showed the first results for its 6th generation of TPU, called Trillium—which it unveiled only last month—and a second round of results for its 5th generation variant, the Cloud TPU v5p. In the 2023 edition, the search giant entered a different variant of the 5th generation TPU, v5e, designed more for efficiency than performance. Versus the latter, Trillium delivers as much as a 3.8-fold performance boost on the GPT-3 training task.

But versus everyone’s arch-rival Nvidia, things weren’t as rosy. A system made up of 6,144 TPU v5ps reached the GPT-3 training checkpoint in 11.77 minutes, placing a distant second to an 11,616-Nvidia H100 system, which accomplished the task in about 3.44 minutes. That top TPU system was only about 25 seconds faster than an H100 computer half its size.

A Dell Technologies computer fine-tuned the Llama 2 70B large language model using about 75 cents worth of electricity.

In the closest head-to-head comparison between v5p and Trillium, with each system made up of 2048 TPUs, the upcoming Trillium shaved a solid 2 minutes off of the GPT-3 training time, nearly an 8 percent improvement on v5p’s 29.6 minutes. Another difference between the Trillium and v5p entries is that Trillium is paired with AMD Epyc CPUs instead of the v5p’s Intel Xeons.

Google also trained the image generator, Stable Diffusion, with the Cloud TPU v5p. At 2.6 billion parameters, Stable Diffusion is a light enough lift that MLPerf contestants are asked to train it to convergence instead of just to a checkpoint, as with GPT-3. A 1024 TPU system ranked second, finishing the job in 2 minutes 26 seconds, about a minute behind the same size system made up of Nvidia H100s.

Training power is still opaque

The steep energy cost of training neural networks has long been a source of concern. MLPerf is only beginning to measure this. Dell Technologies was the sole entrant in the energy category, with an eight-server system containing 64 Nvidia H100 GPUs and 16 Intel Xeon Platinum CPUs. The only measurement made was in the LLM fine-tuning task (Llama2 70B). The system consumed 16.4 megajoules during its 5-minute run, for an average power of 5.4 kilowatts. That means about 75 cents of electricity at the average cost in the United States.

While it doesn’t say much on its own, the result does potentially provide a ballpark for the power consumption of similar systems. Oracle, for example, reported a close performance result—4 minutes 45 seconds—using the same number and types of CPUs and GPUs.

From Your Site Articles

Related Articles Around the Web

Related posts

For Tech Whistleblowers, There’s Safety in Numbers

For Tech Whistleblowers, There’s Safety in Numbers

May 19, 2025
AI transforms Bitcoin price prediction and market strategy

AI transforms Bitcoin price prediction and market strategy

May 19, 2025



Source link

Previous Post

Starlink’s race for space – Developing Telecoms

Next Post

Top 10 African countries with the highest fertility rate in 2024

Next Post
Top 10 African countries with the highest fertility rate in 2024

Top 10 African countries with the highest fertility rate in 2024

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

RECOMMENDED NEWS

Beyon announces Q2 2024 and Half Year 2024 financial results with 7% YoY increase in revenues

Beyon announces Q2 2024 and Half Year 2024 financial results with 7% YoY increase in revenues

10 months ago
The key to successful African family businesses? Good governance and succession planning

The key to successful African family businesses? Good governance and succession planning

10 months ago
Trump’s approach to global trade: Market uncertainties rise

Trump’s approach to global trade: Market uncertainties rise

6 months ago
Analyst Predicts You Can Become a Millionaire with 500 Tokens

Analyst Predicts You Can Become a Millionaire with 500 Tokens

2 years ago

POPULAR NEWS

  • Ghana to build three oil refineries, five petrochemical plants in energy sector overhaul

    Ghana to build three oil refineries, five petrochemical plants in energy sector overhaul

    0 shares
    Share 0 Tweet 0
  • When Will SHIB Reach $1? Here’s What ChatGPT Says

    0 shares
    Share 0 Tweet 0
  • Matthew Slater, son of Jackson State great, happy to see HBCUs back at the forefront

    0 shares
    Share 0 Tweet 0
  • Dolly Varden Focuses on Adding Ounces the Remainder of 2023

    0 shares
    Share 0 Tweet 0
  • US Dollar Might Fall To 96-97 Range in March 2024

    0 shares
    Share 0 Tweet 0
  • Privacy Policy
  • Contact

© 2023 LBNN - All rights reserved.

No Result
View All Result
  • Home
  • Business
  • Politics
  • Markets
  • Crypto
  • Economics
    • Manufacturing
    • Real Estate
    • Infrastructure
  • Finance
  • Energy
  • Creator Economy
  • Wealth Management
  • Taxes
  • Telecoms
  • Military & Defense
  • Careers
  • Technology
  • Artificial Intelligence
  • Investigative journalism
  • Art & Culture
  • Documentaries
  • Quizzes
    • Enneagram quiz
  • Newsletters
    • LBNN Newsletter
    • Divergent Capitalist

© 2023 LBNN - All rights reserved.