By Unexpected! There was a huge mashup between Nvidia and Intel when we woke up this morning. In return for Team Green’s impressive $5 billion investment in Intel, the two businesses will collaborate to develop x86 processors for consumers and data centers that are integrated with Nvidia’s RTX graphics. Dogs and cats living together, human sacrifice… HYSTERIA MASS!
With the world’s leading graphics pioneer teaming hands with the business formerly known as Chipzilla, it’s a startling upheaval of the PC chip trifecta (AMD must be furious), a much-needed lifeline for struggling Intel, and a recipe for a potentially exciting future. Just think about what could be!
The brief history of Intel Arc indicates potential.
Arc is a relatively new technology. In the 2000s, Intel notably discontinued its early “Larrabee” graphics design, which proved to be a liability as the emergence of Bitcoin and artificial intelligence demonstrated the formidable capabilities of GPUs. After realizing it had missed the mark, Intel hurried—sometimes slowly—to coordinate the Arc brand and the Xe graphics architecture that supported it.
Despite a slew of irksome faults, the first Arc graphics cards were released in October 2022 and offered excellent value for the money. After Intel painstakingly resolved those issues over time, we dubbed the second-generation Arc B580 “the first worthy budget GPU of the decade” when it was released in late 2024.
…but possibly weak foundations
When trying to enter a new market where Nvidia and AMD have a decades-long advantage, issues other than software defects can arise. The size of the discrete Arc GPU die sizes indicates that Intel’s GPU hardware capabilities are still not on par with those of its competitors. The cost of making larger dies is significantly higher. The die size of the $250 Arc B580 is a rather large 292 mm. The RTX 4060 from Nvidia, on the other hand, measured about 150 mm, and the RTX 5060 measures about 181 mm.
That is important. Last year, Tom Peterson of Intel, who frequently appears on our Full Nerd podcast, acknowledged that the Arc B580 is a “loss leader”—a product that costs the company money to sell in an effort to draw in customers. In order to prepare for a more powerful GPU in the future, Intel decided it was worth taking that loss.
Despite Intel’s claim that no significant changes are currently planned, the company’s cooperation with Nvidia calls into question that. An Intel representative informed me that while “we’re not talking about specific roadmaps at this time, the collaboration is complementary to Intel’s roadmap and Intel will continue to have GPU product offerings.”
Since AI is where the real money lies, Intel needs powerful GPUs to compete with Nvidia in the data center. To that end, the consumer Arc cards serve as stepping stones. In order to incorporate RTX graphics into at least some Intel consumer CPUs and develop data center solutions that combine Nvidia’s industry-leading graphics with Intel’s x86 chips, Nvidia is now investing $5 billion in Intel, or about 5% of the company, if the recent government investment is any indication.
In the data center, where AMD has been stealing Intel’s lunch, a significant investor may provide Intel a lifeline by introducing GPU technology to its CPUs, extending from consumer to corporate. Is it still worthwhile to invest in Arc separately?
