This article first appeared on GuruFocus.
Intel (INTC, Financials) is making another push into the artificial intelligence chip market with its upcoming data center GPU, code-named Crescent Island, which customers will begin testing in the second half of 2026.
The chip represents Intel’s latest attempt to establish a foothold in AI acceleration after its earlier Gaudi processors struggled to compete against market leaders Nvidia (NVDA, Financials) and AMD. The company said the new GPU is designed to handle large-scale AI workloads used in training and inference for data center operations.
Intel’s roadmap is still behind those of its main competitors, which shows how hard it is for the business to catch up to Nvidia’s stronghold in the AI chip industry. Since OpenAI’s ChatGPT came out in late 2022, there has been a huge increase in the need for GPUs that can run generative AI models. Chipmakers are now rushing to increase their production capacity.
Analysts argue that Crescent Island will be a very important test of Intel’s long-term relevance in AI hardware. This is an area that cloud operators and enterprise clients are investing a lot of money on in data centers.
Intel hasn’t given out any precise specs or prices for Crescent Island yet, but they are anticipated to do so before the customer trials next year.