Intel Targets Nvidia and AMD With New AI Chip

A Cheaper Bet on AI Inference

The plan, flagged in a widely shared market update, centers on a data-center graphics processing unit (GPU) called Crescent Island. Unlike the top-end accelerators from Nvidia and AMD that rely on expensive high-bandwidth memory, Intel’s chip is built around lower-cost LPDDR5X memory, supporting up to 480GB, and is designed to run in air-cooled server racks rather than demanding exotic liquid-cooling setups.

With its design, Intel is aiming at AI inference (the stage where trained models actually answer queries) rather than the most demanding training workloads where Nvidia dominates. By emphasizing “performance per dollar” and what executives call token economics, Intel hopes to undercut rivals on operating cost for the high- volume, always-on workloads that increasingly define commercial AI.

Customer sampling of Crescent Island is targeted for the second half of 2026, with an open, modular approach that lets buyers mix Intel GPUs with hardware from other vendors.

Intel is not entering the fight short of capital as the company has secured more than $18 billion in fresh funding, including $11.1 billion from the U.S. government and $5 billion from Nvidia itself. Not only that, but it also recently secured $2 billion from Japanese multinational investment holding giant Softbank.

Even then, the incumbents are formidable, given Nvidia’s accelerators remain the default for cutting-edge AI, and AMD has carved out a credible challenger position. Intel’s wager is that not every workload needs the fastest, priciest silicon, and that a meaningful slice of the market will trade peak performance for lower upfront and energy costs.

Crypto Needs to Be Paying Attention

For digital-asset readers, the chip race is not a sideshow since a lot of their equipment runs through the same companies. Bitcoin miners, squeezed by thin margins after the latest halving, have been repurposing their power-rich data centers to host AI compute, where revenue per megawatt can dwarf what mining returns.

A couple of months ago, Bitcoin.com News reported that AI data centers are now outpaying bitcoin mining, triggering a major industry shift as operators chase the higher-value workloads. Amidst this, the economics have been transformational for some firms as miners beat bitcoin by 70% in 2026, with Terawulf locking in $12.8 billion in AI contracts as it leaned into high-performance computing (HPC).

In fact, Terawulf has expanded its AI footprint thanks to its 1GW data campus and $3 billion in backing, something that’s part of a broader trend of mining companies reinventing themselves as AI infrastructure providers. Cheaper inference chips like Crescent Island could lower the cost of building out those facilities, potentially improving the returns miners-turned-hosts can earn and reshaping the capital math behind the buildout.

The Bigger Picture for Hardware Costs

A more competitive GPU market matters beyond any single company’s stock because if Intel can pressure prices on inference hardware, the cost of standing up AI capacity could fall across the board, benefiting the crypto-adjacent operators racing to fill data centers with rentable compute.

The energy angle is equally relevant here. Air-cooled, lower-power chips ease the strain on the electrical infrastructure that both miners and AI hosts compete for, a constraint that has become one of the defining bottlenecks of the sector. Power, not just silicon, is now the scarce resource, and hardware that does more with less directly affects who can scale.

Looking ahead for Intel, the near-term milestone seems to be customer sampling, followed by the benchmarks and design wins, since that will determine whether Crescent Island is a genuine threat to the status quo or just another niche alternative.