Key Points
-
Nvidia plans to launch a dedicated AI inference chip at its GPU Technology Conference (GTC) on March 16, 2023.
-
Competitors such as Amazon and Alphabet are developing more energy-efficient processors, raising concerns over Nvidia’s market share.
-
Nvidia currently holds approximately 92% of the data center GPU market.
Nvidia (NASDAQ: NVDA) is making a strategic move to maintain its leading position in AI technology by introducing a specialized chip designed for AI inference at its GTC later this month. This initiative comes as the company faces increased competition from rivals offering more energy-efficient alternatives, such as Amazon’s Inferentia 2 chips, which boast a claimed 30% to 40% improvement in energy efficiency compared to Nvidia’s GPUs.
The new chip, reportedly utilizing technology from AI start-up Groq, marks a significant pivot in Nvidia’s approach to the AI market, particularly as it looks to enhance its alignment with evolving developer needs focused on inference. OpenAI is expected to be one of the first adopters, committing to 3GW from Nvidia, which could solidify Nvidia’s standing in AI development and deployment.
