GPU Performance Hierarchy: Video Cards Ranked from Fastest to Slowest
When you want to run games or do high-end creative work like 4K video editing, your graphics card plays the biggest role in determining performance. Get a good GPU, and you can play recent games at smooth frame rates; get a great one and you can enjoy those games at high resolution, with the special effects turned up.
To help you decide which graphics card you need, we've developed the GPU hierarchy below, which ranks all the current chips from fastest to slowest. For comparison purposes, we've assigned each a score where the fastest card gets 100 and all others are graded relative to it. These numbers are based on the geometric mean fps from our Far Cry 5, Forza Motorsport 7, and Ashes of the Singularity: Escalation benchmarks, giving us a good mix of game genres and APIs.
| Score | GPU | Base/Boost | Memory | Power | Buy | |
| Nvidia Titan Xp | 100 | GP102 | 1405/1480 MHz | 12GB GDDR5X | 250W | $1200Nvidia |
| Nvidia GeForce GTX 1080 Ti | 99.9 | GP102 | 1480/1582 MHz | 11GB GDDR5X | 250W | $694.99Newegg |
| Nvidia GeForce GTX 1080 | 88.8 | GP104 | 1607/1733 MHz | 8GB GDDR5X | 180W | $701Newegg |
| AMD Radeon RX Vega 64 | 88.6 | Vega 10 | 1274/1546 MHz | 8GB HBM2 | 295W | $679Newegg |
| Nvidia GeForce GTX 1070 Ti | 82.6 | GP104 | 1607/1683 MHz | 8GB GDDR5 | 180W | $479Newegg |
| AMD Radeon RX Vega 56 | 79.8 | Vega 10 | 1156/1471 MHz | 8GB HBM2 | 210W | $642Newegg |
| Nvidia GeForce GTX 1070 | 73.2 | GP104 | 1506/1683 MHz | 8GB GDDR5 | 150W | $459Newegg |
| AMD Radeon RX 580 8GB | 61.2 | Polaris 10 | 1257/1340 MHz | 8GB GDDR5 | 185W | $495Newegg |
| Nvidia GeForce GTX 1060 6GB | 56.4 | GP106 | 1506/1708 MHz | 6GB GDDR5 | 120W | $314Newegg |
| Nvidia GeForce GTX 1060 3GB | 45.3 | GP106 | 1506/1708 MHz | 3GB GDDR5 | 120W | $229Newegg |
| AMD Radeon RX 570 | 51.5 | Polaris 10 | 1168/1244 MHz | 4GB GDDR5 | 150W | $289Newegg |
| Nvidia GeForce GTX 1050 Ti | 33.6 | GP107 | 1290/1392 MHz | 4GB GDDR5 | 75W | $189Newegg |
| AMD Radeon RX 560 | 30.1 | Polaris 11 | 1175/1275 MHz | 4GB GDDR5 | 80W | $149Newegg |
| Nvidia GeForce GTX 1050 | 26.1 | GP107 | 1354/1455 MHz | 2GB GDDR5 | 75W | $139Newegg |
| AMD Radeon RX 550 | 18.7 | Polaris 12 | 1100/1183 MHz | 4GB GDDR5 | 50W | $119Newegg |
| Nvidia GeForce GT 1030 | 11.2 | GP108 | 1228/1468 MHz | 2GB GDDR5 | 30W | $92Newegg |
Nvidia’s GeForce GTX 1080 Ti and Titan Xp are the logical choices for 4K gaming, even if some titles require dialed-back detail settings for fluid gameplay.
AMD’s Radeon RX Vega 64 and Nvidia’s GeForce GTX 1080 occupy a lower performance tier. They’re ideal for running at 2560x1440 using the highest quality presets. And as you can see, they score similarly in our index.
The GeForce GTX 1070 Ti, Radeon RX Vega 56, and GeForce GTX 1070 don’t land far from each other, either. All three cards are capable of great frame rates at 2560x1440, and they’re all good options for driving a VR head-mounted display.
There’s a larger group of Radeon RX 580, GeForce GTX 1060 6GB, GeForce GTX 1060 3GB, and Radeon RX 570 boards that excel at 1920x1080. They can even be made serviceable at 2560x1440 if you dial down your detail settings far enough. But beware: insufficient GDDR5 memory on cards like the 3GB GeForce GTX 1060 may cause severe performance issues at higher resolutions. That model is usually faster than Radeon RX 570. However, because we benchmark Forza at 2560x1440 using High settings, its 3GB just aren’t enough, skewing the score lower.
The same thing happens in the next category, where GeForce GTX 1050 appears much slower than Radeon RX 560 due to the former’s 2GB of memory, hurting it in our Forza Motorsports 7 benchmark. Gaming at 1920x1080 or 1280x720 is far more realistic with these cards.
Finally, Radeon RX 550 and GeForce GT 1030 close out this generation’s hierarchy with performance well-suited to popular eSports titles.
Test System & Configuration
| Test System | |
| Processor | Intel Core i7-7700K |
| Motherboard | MSI Z270 Gaming Pro Carbon |
| Memory | 32GB G.Skill TridentZ DDR4-3200 (4x 8GB) |
| Storage | 500GB Crucial MX200 SSD |
| Operating System | Windows 10 Pro v.1803 |
| Games | |
| Far Cry 5 | v.1.06, Fullscreen, 1920x1080, High Graphics Quality, High Texture Filtering, High Shadows, High Geometry & Vegetation, High Environment, High Water, High Terrain, High Volumetric Fog, TAA Anti-Aliasing, Motion Blur On |
| Ashes of the Singularity: Escalation | v.2.7532453, High Quality Profile, 2560x1440, Fullscreen, DirectX 12 |
| Forza Motorsport 7 | v.1.126.9433.2, Ultra Dynamic Render Quality, 2560x1440, Custom Dynamic Optimization, Fullscreen, Unlocked Performance Target, 100% Resolution Scale, 16x Anisotropic Filtering, 2x MSAA Quality, Car Headlights In Rain Off, and all other options set to High |
Legacy GPU Hierarchy
Tom's Hardware readers accustomed to referencing our legacy desktop GPU hierarchy with historical comparisons dating back to the 1990s will find them below Here we group cards into performance tiers, pairing disparate generations where overlap occurs.
| Nvidia GeForce | AMD Radeon |
|
Titan Xp ($1,200 On Nvidia)
Titan X (Pascal) ($1,649.99 On Amazon) GTX 1080 Ti ($699 On Amazon) GTX 1080 ($559.89 On Amazon) | RX Vega 64 ($649.99 On Best Buy) |
| GTX Titan X (Maxwell) GTX 1070 Ti ($469.99 On -) GTX 1070 ($459.89 On Amazon) 980 Ti ($537.83 On Amazon) | R9 295X2 RX Vega 56 ($469.99 On Newegg) R9 Fury X |
| GTX Titan Black GTX 980 ($539.99 On Amazon) GTX 690 | R9 Fury R9 Fury Nano |
| GTX Titan GTX 1060 ($249.99 On Newegg) GTX 970 GTX 780 Ti GTX 780 |
RX 580 ($296.89 On Amazon)
RX 570 ($229.89 On Amazon) RX 480 ($194.97 On Amazon) RX 470 ($204.00 On Amazon) R9 390X ($429.99 On Amazon) R9 390 ($319.99 On Newegg) R9 290X R9 290 HD 7990 |
| GTX 770 GTX 680 GTX 590 | R9 380X R9 380 ($229.99 On Amazon) R9 280X HD 7970 GHz Edition HD 6990 |
|
GTX 1050 Ti ($157.99 On Amazon)
GTX 960 ($199.99 On Amazon) GTX 670 GTX 580 | R9 285 R9 280 HD 7950 HD 7870 LE (XT) HD 5970 |
|
GTX 1050 ($119.99 On Amazon)
GTX 950 GTX 760 GTX 660 Ti |
RX 560 ($129.99 On Amazon)
RX 460 R7 370 R9 270X R9 270 HD 7870 |
| GTX 660 GTX 570 GTX 480 GTX 295 | R7 265 HD 7850 HD 6970 HD 4870 X2 |
| GTX 750 Ti GTX 650 Ti Boost GTX 560 Ti (448 Core) GTX 560 Ti GTX 470 | R7 260X HD 6950 HD 5870 HD 4850 X2 |
| GTX 750 GTX 650 Ti GTX 560 | HD 7790 HD 6870 HD 5850 |
|
GT 1030 ( On -)
GTX 465 GTX 460 (256-bit) GTX 285 9800 GX2 |
RX 550 ($79.99 On Amazon)
R7 360 R7 260 HD 7770 HD 6850 |
| GT 740 GDDR5 GT 650 GTX 560 SE GTX 550 Ti GTX 460 SE GTX 460 (192-bit) GTX 280 GTX 275 GTX 260 | R7 250E R7 250 (GDDR5) HD 7750 (GDDR5) HD 6790 HD 6770 HD 5830 HD 5770 HD 4890 HD 4870 |
| GTS 450 GTS 250 9800 GTX+ 9800 GTX 8800 Ultra | R7 250 (DDR3) HD 7750 (DDR3) HD 6750 HD 5750 HD 4850 HD 3870 X2 |
| GT 730 (64-bit, GDDR5) GT 545 (GDDR5) 8800 GTS (512MB) 8800 GTX | HD 4770 |
| GT 740 DDR3 GT 640 (DDR3) GT 545 (DDR3) 9800 GT 8800 GT (512MB) | HD 7730 (GDDR5) HD 6670 (GDDR5) HD 5670 HD 4830 |
| GT 240 (GDDR5) 9600 GT 8800 GTS (640MB) | HD 6570 (GDDR5) HD 5570 (GDDR5) HD 3870 HD 2900 XT |
| GT 240 (DDR3) 9600 GSO 8800 GS | R7 240 HD 7730 (DDR3) HD 6670 (DDR3) HD 6570 (DDR3) HD 5570 (DDR3) HD 4670 HD 3850 (512MB) |
| GT 730 (128-bit, GDDR5) GT 630 (GDDR5) GT 440 (GDDR5) 8800 GTS (320MB) 8800 GT (256MB) | HD 5550 (GDDR5) HD 3850 (256MB) HD 2900 Pro |
| GT 730 (128-bit, DDR3) GT 630 (DDR3) GT 440 (DDR3) 7950 GX2 | HD 7660D (integrated) HD 5550 (DDR3) HD 4650 (DDR3) X1950 XTX |
| GT 530 GT 430 7900 GTX 7900 GTO 7800 GTX 512 | X1900 XTX X1950 XT X1900 XT |
| GT 220 (DDR3) 7950 G 7900 GT 7800 GTX | HD 7560D (integrated) HD 5550 (DDR2) HD 2900 GT X1950 Pro X1900 GT X1900 AIW X1800 XT |
| GT 220 (DDR2) 9500 GT (GDDR3) 8600 GTS 7900 GS 7800 GT | HD 7540D (integrated) HD 6550D (integrated) HD 6620G (integrated) R5 230 HD 6450 HD 4650 (DDR2) X1950 GT X1800 XL |
| 9500 GT (DDR2) 8600 GT (GDDR3) 8600 GS 7800 GS 7600 GT 6800 Ultra | 7480D (integrated) 6530D (integrated) 6520G (integrated) HD 3670 HD 3650 (DDR3) HD 2600 XT X1800 GTO X1650 XT X850 XT PE X800 XT PE X850 XT X800 XT |
| GT 520 8600 GT (DDR2) 6800 GS (PCIe) 6800 GT | 6480G (integrated) 6410D (integrated) HD 3650 (DDR2) HD 2600 Pro X800 GTO2/GTO16 X800 XL |
| 6800 GS (AGP) | 6380G (integrated) 6370D (integrated) X1650 GT X850 Pro X800 Pro X800 GTO (256MB) |
| 8600M GS 7600 GS 7300 GT (GDDR3) 6800 | X1650 Pro X1600 XT X800 GTO (128MB) X800 |
| 9400 GT 8500 GT 7300 GT (DDR2) 6800 XT 6800LE 6600 GT | HD 6320 (integrated) HD 6310 (integrated) HD 5450 HD 4550 HD 4350 HD 2400 XT X1600 Pro X1300 XT X800 SE X800 GT X700 Pro 9800 XT |
| 9400 (integrated) 9300 (integrated) 6600 (128-bit) FX 5950 Ultra FX 5900 Ultra FX 5900 | HD 6290 (integrated) HD 6250 (integrated) HD 4290 (integrated) HD 4250 (integrated) HD 4200 (integrated) HD 3300 (integrated) HD 3200 (integrated) HD 2400 Pro X1550 X1300 Pro X700 9800 Pro 9800 9700 Pro 9700 |
| FX 5900 XT FX 5800 Ultra | X1050 (128-bit) X600 XT 9800 Pro (128-bit) 9600 XT 9500 Pro |
| G 310 G 210 8400 G 8300 6200 FX 5700 Ultra 4 Ti 4800 4 Ti 4600 | Xpress 1250 (integrated) HD 2300 X600 Pro 9800 LE 9600 Pro |
| 9300M GS 9300M G 8400M GS 7300 GS FX 5700, 6600 (64-bit) FX 5600 Ultra 4 Ti4800 SE 4 Ti4400 4 Ti4200 | X1050 (64-bit) X300 9600 9550 9500 |
| 8300 (integrated) 8200 (integrated) 7300 LE 7200 GS 6600 LE 6200 TC FX 5700 LE FX 5600 FX 5200 Ultra 3 Ti500 | X1150 X300 SE 9600 LE 9100 8500 |
| FX 5500 FX 5200 (128-bit) 3 Ti200 3 | 9250 9200 9000 |
| FX 7050 (integrated) FX 7025 (integrated) FX 6150 (integrated) FX 6100 (integrated) FX 5200 (64-bit) | Xpress 1150 (integrated) Xpress 1000 (integrated) Xpress 200M (integrated) 9200 SE |
| 2 Ti 200 2 Ti 2 Ultra 4 MX 440 2 GTS | 7500 |
| 2 MX 400 4 MX 420 2 MX 200 256 | 7200 7000 DDR LE SDR |
| Nvidia TNT | Rage 128 |
MORE: Best Graphics
MORE: AMD Radeon RX 480 Roundup
MORE: Nvidia GeForce GTX 1060 Roundup
MORE: Nvidia GeForce GTX 1070 Roundup
MORE: Nvidia GeForce GTX 1080 Roundup
I like this scoring methodology, except...
1. It's not specified, but I feel that the scores used should be ones at some plausible preset and AA for a given game.
2. Separate scores (and columns) should be provided for 1920x1080, 2560x1440, and 3840x2160.
Otherwise, very nice.
Practically too, VEGA64 with a weaker cooler can throttle down quite a bit.
(I am rooting for AMD to come out with a more power efficient card even if it's not much more powerful than VEGA64... at some point AMD and NVidia should switch to multi-die GPU's which will be interesting)
I have seen little to nothing for AMD. Lots of secrecy with their next GPU. One thing for sure though is that if they do not get a handle on the crypto-mining inflation there is no way they will conquer the gaming market. Not with double the price of a GTX 1080 and a much (nearly double) higher TDP.
Short of the die hard AMD fans and crypto-miners no one would pay $1200 for a GPU when you can get the 1080 Ti for $850.
And thanks to that price inflation nVidia isn't going to drop prices on theirs anytime soon except when they launch their new top end GPUs.
The testing methodology for the new chart doesn't really seem up to par though. Basing numbers on performance in just 3 games is not going to provide particularly accurate results, especially when 2 of the 3 are running at resolutions that some of the cards aren't even designed to handle. If the choice of resolution is going to reposition cards in the chart, why not provide separate results for different resolutions? The vast majority of people are still gaming at resolutions of 1080p or below, so that should be the most relevant resolution to test. If someone is gaming at 1440p or higher with a higher-end card, then they are probably also more likely to read reviews, and less likely to need a chart like this.
I must say that I prefer the performance summary charts included in TechPowerUp's reviews, as they not only provide separate charts for 1080p, 1440p and 2160p, but they also base the results on the combined performance of the 20+ games in their test suite, preventing anomalous results in any one title from throwing off the average too much. Plus, they make the actual frame rate results for each game available in separate charts, so one can get an idea of whether a certain difference in performance might even be relevant at a given refresh rate, and in the games they actually play. Tom's of course provides much more detailed results for each game tested in their reviews, but that doesn't apply in a summary like this. I would at the very least like to see more games being used to calculate these results.
I'll bring up the idea of adding resolutions and games. I think the purpose here was to add a better-quantified alternative to the old legacy chart (which is still included underneath), but to start backing some of the current-gen rankings with data. I am at least satisfied that the three genres/chosen APIs do match our review guidance. However, the suggestions are appreciated and, again, I'll bring them up in our next meeting to see if I can get the green light for more data.
I was unable to compare my card with the new ones.
Tom's hierarchy lists and techreport x-y price performance charts are some of the most valuable info on hardware reviews.
I wouldn't really call that favoring AMD. What's happened is that game developers are now spending as much time optimizing for AMD as they are for Nvidia. Older games were optimized for Nvidia but not for AMD. So the FPS in current games is a more of a level playing field between the capabilities of cards from the two manufacturers. i.e. Older games favored Nvidia. Current games don't favor either.
That said, to address your point, this really needs to be turned into a dynamic chart (and graph I mentioned above), where the user can select which game they want to use for the comparison. The website should just send a bunch of FPS data to the browser, and a script on the browser can dynamically generate the chart based on which game the user selects.
Good to add the performance difference between GTX1060 3gb and 6gb.
Perhaps you could add a column for integrated graphics.
Many will want to know just where HD630 fits