Nvidia’s new Blackwell chips are twice as fast as their predecessors in training massive AI models, cutting chip use and time dramatically, new benchmarks reveal.
Nvidia's (NVDA.O), opens new tab newest chips have made gains in training large artificial intelligence systems, new data released on Wednesday showed, with the number of chips required to train large language models dropping dramatically.
MLCommons, a nonprofit group that publishes benchmark performance results for AI systems, released new data about chips from Nvidia and Advanced Micro Devices (AMD.O), opens new tab, among others, for training, in which AI systems are fed large amounts of data to learn from. While much of the stock market's attention has shifted to a larger market for AI inference, in which AI systems handle questions from users, the number of chips needed to train the systems is still a key competitive concern. China's DeepSeek claims to create a competitive chatbot using far fewer chips than U.S. rivals.
The results were the first that MLCommons has released about how chips fared at training AI systems such as Llama 3.1 405B, an open-source AI model released by Meta Platforms (META.O), opens new tab that has a large enough number of what are known as "parameters" to give an indication of how the chips would perform at some of the most complex training tasks in the world, which can involve trillions of parameters.
Nvidia and its partners were the only entrants that submitted data about training that large model, and the data showed that Nvidia's new Blackwell chips are, on a per-chip basis, more than twice as fast as the previous generation of Hopper chips.
In the fastest results for Nvidia's new chips, 2,496 Blackwell chips completed the training test in 27 minutes. It took more than three times that many of Nvidia's previous generation of chips to get a faster time, according to the data.
In a press conference, Chetan Kapoor, chief product officer for CoreWeave, which collaborated with Nvidia to produce some of the results, said there has been a trend in the AI industry toward stringing together smaller groups of chips into subsystems for separate AI training tasks, rather than creating homogenous groups of 100,000 chips or more.
"Using a methodology like that, they're able to continue to accelerate or reduce the time to train some of these crazy, multi-trillion parameter model sizes," Kapoor said.
Read next
14:06
Ahead of the IAEA meeting in Vienna, Iran’s foreign minister reaffirmed the country’s refusal to halt uranium enrichment and warned Europe against seeking renewed UN sanctions, calling such efforts a “major strategic mistake.”
14:00
world news
The European Union and Ukraine are negotiating a new agreement on agricultural import quotas, aiming to find a balance between pre-war trade limits and the temporary exemptions granted after 2022.
13:49
Eid al-Adha
Eid al-Adha, known as the "Festival of Sacrifice," stands as one of Islam's most significant celebrations, commemorating Prophet Ibrahim's willingness to sacrifice his son in obedience to Allah's command.
13:40
A U.S. federal judge has temporarily blocked President Donald Trump’s attempt to bar foreign students from attending Harvard University, following a legal challenge filed by the institution.
12:40
Exams in Congo
42 thousands of secondary school students sat for state exams in rebel-held eastern Congo this week, a complicated logistical feat requiring rare cooperation between the government and M23 rebels.
What is your opinion on this topic?
Leave the first comment