The AI chip race is heating up yet again, as Advanced Micro Devices (AMD) has revealed its latest designs, as well as a plan to develop more models to challenge industry leader Nvidia over the next couple of years.

The explosive proliferation of AI capabilities has galvanised an intense demand for advance chip technology, especially as energy costs dampen accessibility and excitement for emerging technology.

AMD’s CEO Lisa Su introduced a new chip, the MI325X accelerator, at the Computex technology trade show, saying it should be available by the end of 2024.

The California-based chip company has been vying for years to overtake Nvidia, the industry leader, which currently dominates the AI semiconductor market with 80% of its share.

AMD claims that their new chip will perform inference, the process of generative AI computation, 35 times better than its old models. The company also revealed that its new series, MI400 will arrive in 2026.

The company is matching Nvidia’s annual release cycle, attempting to keep up with demand for newer, more advanced chips to power AI computing needs.

“This annual cadence is something that is there because the market requires newer products and newer capabilities,” Su said at the event, as reported by Reuters.

Recommended reading

• Gartner: Global AI Chips Revenue to Total £56BN in 2024
• Is the UK Lagging Behind in Global AI Race?
• How Many Hours Will Generative AI Save UK Workers?

Investors are keen for semiconductor designers and manufacturers to keep up the pace – AMD saw their shares increase by more than double since the beginning of 2023. Nvidia, however, saw a seven-fold increase in their shares.

The AI race continues to heat up as Arm, the UK chip design company, announced that it expects 100 billion of its own chips to be ready for AI worldwide by the end of next year.

Gartner also recently predicted that worldwide AI chip revenue is expected to grow by a third (33%) just this year, set to total £56 billion.