NVIDIA is expected to generate $40 billion of revenue from AI GPU sales this year according to Bloomberg analysts, with competitors AMD and Intel in their dust.
AMD is estimated to generate around $3.5 billion in revenue from its new Instinct MI300X AI accelerator, while Intel will enjoy some loose change with its Gaudi AI accelerator, with analysts estimating just $500 million in sales. Both pale in comparison to $40 billion from NVIDIA, which is 80x more than Intel’s estimated $500 million.
These numbers were discussed by Bloomberg Technology during an interview with Intel CEO Pat Gelsinger, where he had no real answer to how Intel’s new Gaudi AI accelerator will battle NVIDIA. When NVIDIA is making 80x your revenue — and keep in mind, this is just AI GPU sales, not GeForce graphics cards or any other NVIDIA product — it’s hard to compete against the dominant force in AI.
AMD is in a position that’s not too far from Intel, with its Instinct MI300X AI accelerator expected to generate $3.5 billion in revenue, that’s a far cry from NVIDIA — 11x as far, in fact — but it shows that AMD is far out and ahead of Intel.
But… I think these numbers are way off. NVIDIA generated $18.4 billion in revenue from its data center business in Q4 FY24 alone, while over the entire financial year 2024 the company generated a whopping $47.4 billion. That’s with the current-gen Hopper H100, let alone the beefed-up Hopper H200 AI GPU, and the next-gen Blackwell AI GPUs including the new B200 AI GPU that was revealed last month.
$40 billion in revenue from AI GPUs is less than the last 12 months of revenue from its data center business for NVIDIA, so I’m sure we’re going to see far more than $50 billion, and more like $80-$100 billion in revenue over the next year from AI GPUs alone for the company. AMD can’t touch those numbers, and Intel certainly can’t come close… $500 million to let’s say $80 billion… is a huge 160x the revenue. One hundred and sixty times the revenue. Nuts.
2VIEW GALLERY – 2 IMAGES
Intel has no way out of this at all, with Gaudi 3 using older components like HBM2e memory while AMD and NVIDIA have been using the faster HBM3 memory standard for a while now, and the ultra-fast HBM3E memory will be used on both Instinct MI300X from AMD, and Hopper H200 and Blackwell B200 AI GPUs from NVIDIA.
$500 million to even $1 billion in revenue in 2025 could be hard for Intel at this stage, where they’re multiple generations behind at this point, let alone in 12-18 months time. The AI market is growing at speeds that are hard to keep up, and it seems Intel is not just dragging its feet, but they’re almost standing still.