Who Makes GPUs? NVIDIA, AMD, and Intel Compared
By GPU Alpha
Explore the GPU market landscape with a comparison of NVIDIA, AMD, and Intel. Understand their strengths, weaknesses, and market positions.
Who Makes GPUs? NVIDIA, AMD, and Intel Compared
The market for graphics processing units (GPUs) has never been more important, or more concentrated. Whether you are building a gaming PC, setting up a machine learning workstation, or exploring local AI inference, the GPU you choose will largely come from one of three companies: NVIDIA, AMD, or Intel. Understanding how these manufacturers differ, where they compete, and how they influence pricing can save you real money and help you avoid costly mistakes.
The Three Players and Where They Stand
The GPU market is one of the most lopsided competitive landscapes in consumer technology. According to data reported by TechSpot, NVIDIA held a 94% share of the discrete GPU market as of Q2 2025. AMD accounted for the remaining 6%, and Intel's share was negligible. A discrete GPU, for context, is a standalone graphics card rather than graphics processing built into a CPU chip.
By the end of 2025, AMD's position had weakened further. Tom's Hardware reported that AMD's share had slipped to approximately 5%, a historical low for the company. Intel, despite launching a new generation of graphics cards, had not yet gained meaningful traction in the market.
These numbers tell a clear story. NVIDIA is the dominant force by a wide margin, AMD is fighting to hold its ground, and Intel is still working to establish itself as a credible option.
NVIDIA: The Market Leader
NVIDIA's current flagship lineup is the GeForce RTX 50-series, which includes the RTX 5080 and the RTX 5090. These cards represent the top of the consumer GPU market in terms of raw performance.
What sets NVIDIA apart technically is its investment in two areas. The first is ray tracing, a rendering technique that simulates how light behaves in a physical environment to produce more realistic visuals. The second is DLSS 4, which stands for Deep Learning Super Sampling. DLSS uses AI processing to reconstruct a high-resolution image from a lower-resolution input, effectively boosting frame rates without a proportional loss in visual quality.
These features have made NVIDIA the preferred choice for gamers who want the best performance available, and for developers building AI applications who benefit from NVIDIA's mature software ecosystem called CUDA. CUDA (Compute Unified Device Architecture) is a programming platform that allows developers to write software that runs directly on NVIDIA hardware, and it has become deeply embedded in the AI and machine learning industry.
AMD: The Value-Oriented Challenger
AMD's current flagship consumer GPU is the Radeon RX 9070 XT, part of the RX 9000-series. AMD's approach to the market has historically emphasised raw rasterization performance at competitive price points. Rasterization is the traditional method of rendering 3D graphics, and AMD's cards have generally performed well in this area relative to their cost.
AMD also supports FSR, or FidelityFX Super Resolution, its alternative to NVIDIA's DLSS. FSR is an open standard, meaning it works across a wider range of hardware including some NVIDIA cards, which some analysts argue gives it a long-term compatibility advantage.
Despite its weakening market share figures, AMD has seen some individual product success. According to TechRadar, the Radeon RX 9070 XT held a 1.33% share in the Steam Hardware Survey as of May 2026, which placed it ahead of NVIDIA's GeForce RTX 4070 Ti in that particular survey. Steam is a major PC gaming platform, and its hardware survey provides a useful snapshot of what cards gamers are actually using. This suggests that AMD's mid-range and upper-mid-range cards are finding buyers even as the company's overall share declines.
Intel: The New Entrant
Intel entered the discrete GPU market with its Arc series and has continued with the Arc Battlemage generation. Intel's cards are positioned at the entry-level and budget segments of the market, with pricing intended to attract buyers who want a functional GPU without spending heavily.
The challenge Intel faces is one of trust and track record. Early Arc cards encountered compatibility issues in some games, which created hesitation among buyers. Building a reputation in a market where NVIDIA has decades of driver and software development behind it is a slow process. Intel's Arc GPUs have not yet gained significant traction, but they remain a relevant option for buyers with tight budgets who are willing to accept some trade-offs.
Pricing and Availability in 2025 and 2026
The pricing environment for GPUs has been complicated by supply chain pressures. As of June 2026, PC Gamer reported that high-end cards such as the RTX 5090 are scarce and trading above their recommended retail prices due to a global memory shortage affecting production.
This situation is not unusual for the GPU market. Supply constraints, whether from component shortages, manufacturing bottlenecks, or sudden spikes in demand, have periodically pushed prices well above official launch prices. For buyers who need a GPU now rather than waiting for supply to normalise, mid-range cards from both NVIDIA and AMD currently offer better value and more reliable availability than flagship models.
Different Perspectives on the Competition
The market share numbers favour NVIDIA decisively, but that does not mean AMD is irrelevant to your purchasing decision. If your primary use case is gaming at 1080p or 1440p resolution, and you are working within a defined budget, AMD's mid-range cards can deliver strong performance per dollar compared to NVIDIA equivalents.
On the technology side, NVIDIA's DLSS is widely regarded as technically superior to AMD's FSR in terms of image quality. However, FSR's open-source nature means it is available on more hardware and is not tied to a single manufacturer's ecosystem. For buyers who value flexibility or who may switch GPU brands in future, FSR's broader compatibility is a practical consideration.
What This Means for Buyers
The GPU market in 2025 and 2026 is effectively a market with one dominant supplier and two challengers at very different stages of maturity. NVIDIA sets the pace on performance and software features, AMD competes on value particularly in the mid-range, and Intel is building its presence from a low base.
For most buyers, the practical decision comes down to budget and use case. High-end AI workloads and professional applications lean heavily toward NVIDIA due to its software ecosystem. Gaming on a budget points toward AMD's mid-range options. Entry-level buyers with modest requirements might find Intel's Arc cards worth considering, provided they verify game compatibility before purchasing.
Tracking current prices across all three manufacturers is the most reliable way to find value at any given moment, since the gap between official pricing and actual market prices can shift significantly depending on supply conditions.