When Nvidia achieves 94 percent market share in graphics cards, the typical free-market narrative usually runs: the best product won, consumers chose excellence, competition drove innovation. By most fiscal measures, this reads as a capitalist success story. Yet the latest data from Jon Peddie Research tells a more complex tale, one that should concern policymakers and industry observers alike. Record dominance masks market dysfunction.
The speed of AMD's collapse is remarkable. A year ago, AMD held roughly 17 percent of the add-in board graphics card market. By the end of 2025, that figure had cratered to just 5 percent. In the space of four quarters, AMD lost more than two-thirds of its market share. Intel now registers barely a rounding error at 1 percent. This is not a story of Nvidia simply making better chips. It is a story of a market tipping into monopoly-like conditions where consumer choice is evaporating.
The deeper concern is that gamers are buying fewer graphics cards altogether. According to JPR data, the percentage of desktop personal computers shipping with a discrete graphics card dropped to 55 percent in Q4 2025, down 12.3 percentage points from the previous quarter. That decline occurred during the traditionally robust holiday season. When shoppers do buy, they buy Nvidia; when they don't buy at all, AMD suffers most acutely.
AMD's strategic retreat from the high-end gaming market deserves scrutiny. The company signalled in 2024 that it would cede the premium segment to Nvidia, choosing instead to compete in the mainstream tier. This is rational accounting: why fight a battle you cannot win? Yet it abandons the innovation pipeline that filters down to mass-market products. When competitors stop competing at the top, the entire market feels the chill.
The supply side paints an equally troubling picture. Nvidia is preparing to slash gaming GPU production by 30 to 40 percent in the first half of 2026. The stated reason is memory shortage; the unstated reality is that data centre and artificial intelligence workloads are consuming production capacity and commanding premium margins. A GDDR7 memory crunch, driven partly by redirected manufacturing capacity toward AI infrastructure, is squeezing the consumer market. This creates the perverse outcome where Nvidia's supply cuts may further entrench its market dominance by starving competitors of components and raising prices across the board.
Jon Peddie Research projects that the graphics card market will contract by nearly 10 percent in 2026. This is not cyclical downturn; analysts attribute it to what they call "unstable conditions": rising memory costs, tariff uncertainty, and consumer hesitation to upgrade ageing hardware. Customers are holding off not out of satisfaction with their current systems, but out of rational caution in a volatile market.
The centre-right case for market dynamism rests on competition, innovation, and price discovery. When one firm commands 94 percent of a market, those mechanisms weaken. Nvidia's RTX 5070 claimed 9.4 percent of all surveyed systems in a single month, an extraordinary concentration for a single product. AMD's newest Radeon RX 9000 series, meanwhile, vanished from hardware surveys entirely despite optimistic industry commentary about price-to-performance ratios. The market is not choosing between genuine alternatives; it is consolidating around a single vendor.
Yet the supply-side story offers a different perspective. The memory shortage is real, driven by legitimate enterprise demand for AI infrastructure. Manufacturers cannot simply dial up production to serve both markets equally. Tariff uncertainty, partly caused by shifting trade policy, has created genuine barriers to forward planning. These are not Nvidia's failures; they are market conditions that affect all players. AMD's weakness is real, and partly of its own making: the company chose not to invest in high-end gaming GPU development with the same intensity.
The pragmatic resolution lies somewhere between market confidence and skepticism. Nvidia's dominance reflects genuine technical leadership and superior execution; that much is defensible. But dominance of this magnitude, combined with declining overall market sales and supply constraints, suggests the industry faces structural headwinds rather than healthy competition. Neither a completely laissez-faire approach nor heavy-handed intervention makes sense. What matters is transparency: regulators should monitor whether Nvidia's market power translates into anti-competitive conduct, whilst manufacturers should be enabled to expand capacity without artificial barriers.
For gamers, the message is uncomfortable. Graphics cards are becoming more expensive, harder to find, and concentrated in the hands of one vendor. That may be a temporary effect of AI boom dynamics, or it may signal a long-term shift in hardware economics. Either way, buying now may prove wiser than waiting for better conditions that may never arrive.