GPU Performance Per Dollar
Every GPU in our database ranked by benchmark score divided by price. Filter by brand and resolution — the chart updates instantly. Prices updated at least once every 24 hours.
How the Performance-Per-Dollar Score Is Calculated
The score is straightforward: normalized benchmark score ÷ current retail price × 100. A GPU scoring 25.00 delivers twice the measured gaming performance per dollar as one scoring 12.50. Higher is always better.
Benchmark scores are normalized to the RTX 5090 = 100 — the fastest GPU in the database sets the ceiling. Every other card's score shows its performance relative to that reference point. Scores are recorded at three resolutions (1080p, 1440p, and 4K) because the value ranking shifts at each one.
Scores are aggregated from published reviews across Gamers Nexus, Hardware Unboxed, Linus Tech Tips, and Tom's Hardware, tested with a fast CPU to eliminate CPU bottlenecking and isolate GPU performance differences.
Prices pull from live retailer data and update regularly. A GPU that's value-competitive one month may slip after a competitor price cut — use this chart to check current rankings before buying, not six months after.
Why Value Rankings Change by Resolution
At 1080p, gaming is often CPU-limited rather than GPU-limited, which means mid-range GPUs score well above the performance you actually need — and their lower price makes them look exceptional on a value chart. An expensive card may benchmark 30% higher but deliver no meaningful improvement in practice at that resolution.
At 1440p and 4K, the GPU becomes the true bottleneck. Raw benchmark scores matter more because you need the headroom. A high-end GPU that costs 60% more but benchmarks 80% higher actually improves its value position at higher resolutions compared to 1080p. Switch the resolution toggle to see this shift in real time.
This is why matching a GPU to your target resolution — not just picking the top of any value chart — is the correct approach. The GPU buying guide covers how to match GPU tier to resolution and target benchmark.
Nvidia vs AMD vs Intel Arc — Where Each Brand Wins on Value
Value leadership shifts by price tier, not by brand loyalty. Intel Arc cards tend to dominate the budget value chart at 1080p — the Arc B580 in particular offers more VRAM per dollar than any competitor in its price range. AMD typically leads the mid-range value chart at 1440p, where its rasterization performance-per-dollar advantage is most pronounced. Nvidia wins on raw benchmark ceilings at the high end and on features like DLSS and ray tracing quality, but premium pricing often costs it the value ranking.
Use the brand toggles in the chart to isolate each manufacturer and compare within-brand tiers, or show all three to find the cross-brand value leaders at your target resolution.
If power draw matters — SFF build, high electricity costs, or a case with a TDP cap — the GPU Performance Per Watt chart ranks the same GPUs by efficiency rather than price.
When the Best-Value GPU Isn't the Right Buy
Performance per dollar is the right metric when you're choosing between GPUs that all score high enough for your use case. If every card on your shortlist posts a strong benchmark at your target resolution, pick the cheapest one — the value chart does that automatically.
Where value ranking breaks down: when the top-value GPU has a raw benchmark score that's too low for your resolution target. A card scoring 22.00 on value but only 45/100 on the raw benchmark at 1440p may not suit a high-refresh monitor, regardless of its value position. Check the raw benchmark column in the table alongside the perf/$ column — both matter.
VRAM is the other factor not captured in this chart. A GPU with a strong benchmark score today may throttle in future titles if it has 8GB. The VRAM guide covers how much you actually need by resolution and build lifespan.
Pairing with the right CPU matters too — an underpowered CPU creates a bottleneck at 1080p regardless of GPU quality. The CPU Performance Per Dollar chart applies the same methodology to processors.
If you want the selection handled automatically — matching GPU to resolution, benchmark requirement, and budget simultaneously — the PC Builder handles it end-to-end.
Frequently Asked Questions
What does GPU performance per dollar mean?
Performance per dollar measures how much gaming benchmark score you get for each dollar you spend. The score is calculated as: benchmark score ÷ price × 100. A GPU scoring 20.00 delivers twice the value of one scoring 10.00 at the same price. Higher is better.
Which GPU has the best performance per dollar?
It depends on resolution. Budget mid-range cards typically lead at 1080p because raw performance is more than enough at that resolution, making price the main differentiator. At 1440p and 4K, higher-tier GPUs close the gap because the benchmark score grows faster than the price premium. Use the resolution toggle in the chart above to see rankings for your target resolution.
Should I always buy the GPU with the best performance per dollar?
Not always. The value leader is the right pick when its raw benchmark score is high enough for your use case. But if you need a higher benchmark score than the value leader provides — for example, at 1440p or 4K where the GPU is the main bottleneck — a card with a slightly lower value score but a much higher raw benchmark may be the better choice. Check both columns in the table.
Does resolution change which GPU wins on value?
Yes, significantly. At 1080p, gaming is often CPU-limited, so mid-range GPUs punch above their weight — their lower raw scores still deliver smooth frame rates, making them look exceptional on a value chart. At 1440p and 4K, the GPU becomes the bottleneck, and higher-end cards close the price gap because the benchmark improvement scales faster than the cost.
How are benchmark scores in this chart calculated?
All scores are normalized to the RTX 5090 = 100, so every other GPU's score shows its performance relative to the fastest card in the database. Performance per dollar is then computed as: normalized benchmark score ÷ current retail price × 100. Prices update regularly from live retailer data.