Katarzyna Petru
After months of silence and delays, Intel officially announced its return to competition in the AI chips market. At the Open Compute Summit, the company unveiled a new graphics processor for data centers – Crescent Island – set to launch in 2026. This will be the foundation of Intel's new approach to artificial intelligence – more open, modular, and focused on practical applications, rather than just numbers in benchmarks.
“This chip has been designed with inference in mind, for the real operation of AI models, not just their training. We want to deliver the best performance-to-price ratio – true performance per dollar,” said Sachin Katti, Intel's Chief Technology Officer.
Crescent Island – AI-optimized chip for efficiency
Crescent Island will be a GPU designed for data centers, created to handle AI-related tasks – from running language models to real-time analytics. The chip will offer 160 GB of memory, although it will not be the fast HBM (High Bandwidth Memory) known from AMD Instinct or Nvidia Hopper solutions. Instead, Intel has opted for its own proven solutions from consumer graphics cards to reduce costs and improve energy efficiency.
While this sounds like a compromise, in practice it could be a reasonable move – Intel is targeting customers who are looking for a cheaper, scalable alternative to expensive Nvidia platforms.
New strategy: open ecosystem and modular design
Katti emphasized that Intel is focusing on a open and modular model of collaboration with clients. This means that companies will be able to combine Intel chips with solutions from other manufacturers – including GPUs from Nvidia or AMD. "We do not want to lock clients into a single ecosystem. We aim for mix and match, because we know that this is what the AI market looks like today," said Katti.
This approach may prove to be crucial – especially in times when large cloud companies like Amazon or Google are creating their own AI chips and looking for ways to integrate with existing infrastructure.
Collaboration with Nvidia and the plan to return to annual launches
A surprise for the market was the news that Nvidia invested 5 billion dollars in Intel, acquiring about 4% of its shares. The two companies will collaborate on the development of new processors for PCs and data centers – which may be an attempt to strengthen collaboration in the face of increasing competition from AMD and cloud chip manufacturers.
Intel also plans a new launch rhythm: one AI chip per year, similar to Nvidia and AMD. This is intended to allow the company to respond more quickly to market changes and the development of generative AI, whose demand exploded after the release of ChatGPT.
Challenges: Delays, Memory, and Lost Momentum
However, not everything looks rosy. Crescent Island is emerging at a time when Intel is still chasing the competition, and its earlier projects, such as Gaudi and Falcon Shores, have been put on hold. The chip will also not receive top-tier HBM memory, which means that in pure performance, it may not match the flagship chips from the competition.
However, CEO Lip-Bu Tan assures that the company is getting back on track:
“We no longer just want to chase the market. We want to get back in the game – with our own identity and a real alternative to the current leaders.”
What's next?
Intel plans to expand its AI portfolio by gradually introducing chips with various performance levels – from low-cost inference cards to specialized processors for large language models. Crescent Island is to be the first step towards rebuilding Intel's reputation as a real player in the artificial intelligence sector.
