Advanced Micro Devices Inc. (NASDAQ:AMD) CEO Lisa Su highlighted the company’s AI hardware’s superior performance over NVIDIA Corp (NASDAQ:NVDA) and its potential to cut customer costs during the first quarter earnings call this year.

What Happened: Su claimed that AMD’s MI300X GPUs offer better inferencing performance and significant total cost of ownership (TCO) advantages compared to Nvidia’s H100 GPUs.

She also noted that the MI300X GPUs could deliver the same amount of compute with 45% fewer servers than the competition, potentially reducing initial capex by up to half and lowering annual opex by more than 40%.

“Our partners are seeing very strong performance in their AI workloads. As we jointly optimize for their models, MI300X GPUs are delivering leadership, inferencing performance and substantial TCO advantages compared to H100,” Su said.

See Also: Supreme Court: Drivers Hauling Baked Goods Are in Transportation, Not Baking

She added, “For instance, several of our partners are seeing significant increases in tokens per second when running their flagship LLMs on MI300X compared to H100.”

Why It Matters: AMD’s first quarter earnings call comes on the heels of a modest beat and lukewarm guidance, which caused a pullback in the stock. Despite this, Su’s remarks about the MI300X GPUs’ performance and cost advantages could help bolster investor confidence in the company’s AI hardware.

Earlier in April, analyst predictions about the MI300 chip’s strong growth and success in the server market had already set a positive tone for AMD’s AI hardware. Su’s recent comments further reinforce the potential of AMD’s AI hardware in the market.

Price Action: AMD stock traded lower in after-hours trading, with a decrease of 6.91% to $147.43 on Tuesday. Nonetheless, the stock has seen a 14.29% increase year-to-date, according to the data from Benzinga Pro.

Read Next: Goldman Sachs Tempers Investor Excitement On Tesla Self-Driving Tech In China: ‘Not Yet An Eyes-Off, Unsupervised Product’

Image Via Shutterstock

This content was partially produced with the help of Benzinga Neuro and was reviewed and published by Benzinga editors.