Nvidia still owns roughly 80% of the AI accelerator market and posted $60.4B in data center compute revenue in a single quarter. That's the short answer. The longer answer is that AMD and Google are both taking real share for the first time in years, just from very different starting points.
For most of the last three years, "AI hardware" has been shorthand for "Nvidia GPUs." That's still directionally true, but it's no longer the whole story. AMD's Instinct MI300X and MI350 lines are winning real inference workloads on price, Google's TPU v6 Trillium just landed a landmark deal with Anthropic for up to 1 million chips, and hyperscalers are quietly building their own silicon to reduce Nvidia dependence. This is a look at where the three real contenders — Nvidia H200, AMD MI300, and Google TPU — actually stand on revenue, share, performance, and price in mid-2026.
Who's winning the AI hardware wars: Nvidia H200 vs AMD MI300 vs Google TPU v5/v6?
Nvidia is winning on revenue and share by a wide margin — roughly 80% of the AI accelerator market and $60.4B in quarterly data center compute revenue — while AMD is winning select inference deals on price and Google is winning internal and Anthropic workloads through TPU v6 Trillium, which isn't sold externally at all.
The three companies aren't really competing head-to-head for the same customer in the same way. Nvidia sells to everyone. AMD sells to hyperscalers and neoclouds looking for a cheaper second source. Google only serves its own cloud and a small number of anchor tenants like Anthropic. That structural difference is why "market share" numbers need context, not just a single percentage.
Nvidia H200 vs AMD MI300X: Specs, Performance, and Price
| Metric | Nvidia H200 | AMD MI300X |
|---|---|---|
| Architecture | Hopper | CDNA 3 |
| Memory | 141GB HBM3e | 192GB HBM3 |
| Memory Bandwidth | 4.8 TB/s | 5.3 TB/s |
| Theoretical FP16 Compute | ~989 TFLOPS | ~1,300 TFLOPS |
| Real-World Inference Throughput | ~25,300 tokens/sec (single GPU) | ~18,752 tokens/sec (74% of H200) |
| Cloud Rental (per hour, range) | $3.50-$8.00+ | $1.85-$7.86 |
| Est. Unit Price | $25,000-$35,000 | $10,000-$15,000 |
Sources: Silicon Analysts 2026 GPU market share and TCO comparisons, SemiAnalysis MI300X vs H100/H200 benchmark series, aimultiple GPU benchmark reports, Uvation and Spheron GPU comparison data. Real-world throughput derived from single-GPU inference benchmarks; production throughput varies materially by model and batch size.
Nvidia's revenue advantage: $60.4B in one quarter
Nvidia's Q1 fiscal 2027 results (reported for the quarter ended April 26, 2026) show data center revenue of $75.2B total, up 92% year-over-year, split between $60.4B in compute revenue (up 77% YoY, up 18% sequentially) and $14.8B in networking revenue (up 199% YoY). That single quarter's compute revenue alone is more than 10x AMD's entire data center segment for the same period.
AMD's Q1 2026 data center segment revenue was $5.8B, up from $3.7B a year earlier, driven by 5th-gen EPYC CPUs alongside Instinct MI350 GPUs, part of overall AMD Q1 2026 revenue of $10.3B (up 38% YoY). Some analysts project AMD's Instinct AI GPU line could bring in up to $12B in standalone revenue for full-year 2026 — a real number, but still a fraction of Nvidia's quarterly run rate. Our Nvidia revenue breakdown by segment has the full quarterly trend if you want to track how compute vs. networking revenue is shifting.
Where Google's TPU v6 Trillium actually fits
Google's TPU v6 (Trillium), announced in 2024 and now in broad use alongside a TPU v7 preview, claims a 4.7x price-performance improvement over TPU v5e and 67% lower inference power consumption versus high-end Nvidia GPUs, built on the same N5 process node as TPU v5p but with roughly 2x the peak theoretical FLOPs. The catch: TPUs are not sold as hardware. Access is GCP-only, which makes direct unit-for-unit market share comparisons to Nvidia and AMD structurally different.
The number that matters most for 2026 is Anthropic's October 2025 expansion, which secured access to up to 1 million TPU chips and more than a gigawatt of capacity coming online through 2026, a deal reportedly worth tens of billions of dollars. Google and Meta are also reportedly in advanced discussions for Meta to lease TPUs starting in 2026 and potentially purchase systems outright from 2027. If either scales as described, Google moves from "internal workload chip" to a real third leg of frontier-model compute supply, not just a rounding error next to Nvidia and AMD.
Nvidia H200 vs AMD MI300 vs Google TPU v5: the software gap is the real story
The Nvidia H200 vs AMD MI300 comparison keeps coming back to software, not silicon. AMD's MI300X wins on paper spec sheets — more memory, higher theoretical FLOPS, lower price — but Nvidia's CUDA ecosystem extracts a much higher percentage of that theoretical performance in production. That's why MI300X's real-world single-GPU inference throughput lands around 74% of H200's, even though its spec sheet compute number is higher. Google's ROCm-equivalent problem doesn't really exist, because Google controls its own stack top to bottom for internal and Anthropic workloads, but that same vertical control is what prevents TPUs from being purchased and deployed the way GPUs are.
For investors and operators tracking AI infrastructure spend, the practical read is: price Nvidia at a premium that's mostly justified by software maturity and ecosystem lock-in, price AMD as a credible but still-smaller-scale alternative worth watching for inference-heavy workloads where cost per token matters more than peak throughput, and treat Google TPU access (via GCP rental or an anchor-tenant deal like Anthropic's) as a compute-diversification play rather than a like-for-like GPU substitute. Our AI valuations dashboard tracks how compute-supply diversification is showing up in AI-lab funding terms.
Total cost of ownership: why AMD's cheaper price tag doesn't always win
A pure per-hour price comparison makes the AMD MI300X look like an easy win: roughly $1.85-$7.86/hour versus $3.50-$8.00+/hour for the H200, and a unit price of $10,000-$15,000 versus $25,000-$35,000. But total cost of ownership for a large training or inference cluster depends on more than sticker price. It depends on how much of the chip's theoretical performance you can actually extract in production, how quickly your engineering team can debug distributed training failures, and how mature the surrounding tooling is for things like quantization, kernel fusion, and multi-node orchestration.
That's where Nvidia's roughly two-decade head start with CUDA still shows up in real deployment numbers. Engineering teams that have spent years optimizing on CUDA can often get a cluster of H200s to 95%+ of theoretical throughput on well-understood workloads, while ROCm-based MI300X deployments frequently land closer to 60-75% of theoretical throughput unless a team has invested heavily in AMD-specific tuning. For a hyperscaler renting out thousands of chips, that efficiency gap can matter more than a 30-50% discount on the hourly rate, because it changes how many effective GPU-hours you're actually buying per dollar. AMD has been closing this gap release over release, and MI350's improvements over MI300X are real, but the software maturity advantage Nvidia holds is still the single biggest reason the ~80% market share number hasn't moved faster despite years of AMD's hardware roadmap catching up on paper specs.
What the Nvidia vs AMD vs Google split means for AI infrastructure investors
For anyone underwriting AI infrastructure exposure in 2026 — whether that's a direct position in the public names, a venture bet on a neocloud, or LP exposure through a fund concentrated in AI compute — the practical read on Nvidia vs AMD vs Google TPU is that this is not a winner-take-all market anymore, but it's also not evenly split. Nvidia's $60.4B in single-quarter data center compute revenue and roughly 80% share still make it the default allocation, and its 92% year-over-year data center growth rate shows demand isn't slowing even as competition intensifies.
AMD is the more interesting asymmetric bet at this point in the cycle: a $5.8B data center segment growing off a small base, with analysts projecting Instinct-line revenue as high as $12B for full-year 2026, means AMD doesn't need to take Nvidia's crown to be a good investment — it just needs to keep converting a low single-digit percentage of Nvidia's addressable market every year. Google's TPU business is harder to underwrite directly since it isn't broken out as a standalone revenue line and isn't sold externally, but the Anthropic deal is the clearest evidence yet that TPUs are becoming a credible third leg of frontier compute supply rather than a Google-only curiosity. If Meta's rumored TPU leasing arrangement closes on top of that, expect more AI labs to treat multi-vendor compute sourcing as standard practice rather than a hedge, which should structurally cap how high any single vendor's market share can climb from here. Track how these shifts flow into fund-level performance on our VC performance dashboard.
Bottom line
Nvidia is still winning the AI hardware wars by revenue and share — $60.4B in quarterly data center compute revenue against roughly 80% accelerator market share — but the margin is narrowing for the first time in years. AMD's MI300X is real competition on price for inference workloads, and Google's TPU v6 just landed a 1-million-chip commitment from Anthropic that makes it a genuine third option, not a footnote. If you're modeling AI infrastructure spend for 2026-2027, plan for a market with a dominant leader and two credible, fast-growing alternatives, not a monopoly.
Get VC data most people never see — free.
Weekly benchmarks, valuations, and fund data. No spam, unsubscribe anytime.