nVidia just seem to have released a GT1030 specially against the Radeon RX 550. What is so special about the Radeon RX 500 series?
Let's start with the RX 550 (original can be seen
here, you might have to register with AMD Users)
Radeon RX 500 Series
Model Type | GPU | Fab (nm) | Shaders | TMU | ROP | GFLOP SP | GFLOP DP | GFLOP HP | TBP | Gflop (SP) Watt | Gflop (DP) Watt | Gflop (HP) Watt |
Radeon RX 540 | Lexa | 14 | 512 | 32 | 16 | 1151 | 71.9 | 1151 | 50 | 23.02 | 1.44 | 23.02 |
Radeon RX 550 | Polaris 12 | 14 | 512 | 32 | 16 | 1211 | 75.7 | 1211 | 50 | 24.22 | 1.51 | 24.22 |
Radeon RX 560 | Polaris 11 | 14 | 1024 | 64 | 16 | 2406 | 150.4 | 2406 | 80 | 30.01 | 1.88 | 30.01 |
Radeon RX 570 | Polaris 20 | 14 | 2048 | 128 | 32 | 4784 | 300.0 | 4784 | 150 | 31.89 | 2.00 | 31.89 |
Radeon RX 580 | Polaris 20 | 14 | 2304 | 144 | 32 | 5792 | 362.0 | 5792 | 185 | 31.31 | 2.13 | 31.31 |
Radeon Pro Duo Polaris | 2x Polaris 20 | 14 | 4608 | 288 | 64 | 11456 | 716.0 | 11456 | 250 | 45.82 | 2.86 | 45.82 |
The RX 550 is a very capable card in view of its power needs, a mere 50 Watt. Those willing to spend a little more on purchase and power bills can choose for the even more capable RX 560
The TBP (Typical Board Power) of the RX 560 is actually 60-80 Watt, so the GFLOPS/Watt may be even better yet.
What does nVidia offer?
Geforce 10 Series
Model Type | GPU | Fab (nm) | Shaders | TMU | ROP | GFLOP SP | GFLOP DP | GFLOP HP | TBP | Gflop (SP) Watt | Gflop (DP) Watt | Gflop (HP) Watt |
GeForce GT 1030 | GP108-300-A1 | 16 | 384 | 24 | 8 | 952 | 30 | 15 | 30 | 31.73 | 1.00 | 0.50 |
GeForce GTX 1050 | GP107-300-A1 | 14 | 640 | 40 | 32 | 1733 | 54 | 27 | 75 | 23.11 | 0.72 | 0.36 |
GeForce GTX 1050 Ti | GP107-400-A1 | 14 | 768 | 48 | 32 | 1981 | 62 | 31 | 75 | 26.41 | 0.83 | 0.41 |
GeForce GTX 1060 3GB | GP106-300-A1 | 16 | 1152 | 72 | 48 | 3470 | 108 | 54 | 120 | 28.92 | 0.90 | 0.45 |
GeForce GTX 1060 6GB | GP106-400/410-A1 | 16 | 1280 | 80 | 48 | 3855 | 120 | 60 | 120 | 32.13 | 1.00 | 0.50 |
GeForce GTX 1070 | GP104-200-A1 | 16 | 1920 | 120 | 64 | 5783 | 181 | 90 | 150 | 38.56 | 1.21 | 0.60 |
GeForce GTX 1080 | GP104-400/410-A1 | 16 | 2560 | 160 | 64 | 8228 | 257 | 128 | 180 | 45.71 | 1.43 | 0.71 |
GeForce GTX 1080 Ti | GP102-350-K1-A1 | 16 | 3584 | 224 | 88 | 10609 | 332 | 166 | 250 | 42.44 | 1.33 | 0.66 |
nVidia Titan X | GP102-400-A1 | 16 | 3584 | 224 | 96 | 10157 | 317 | 159 | 250 | 40.63 | 1.27 | 0.64 |
nVidia Titan Xp | GP102-450-A1 | 16 | 3840 | 240 | 96 | 12150* | 380* | 190* | 250 | 48.61 | 1.52 | 0.76 |
*=Normal values unknown yet. For the Titan Xp I only know of the 'Turbo boost' values, at the moment.
Considering the fact that most projects only ask for Single precision, the GT 1030 offers real value. But when we take double precision performance (e.g. MilkyWay) into account, nVidia is fighting a lost battle as they have sacrificed that feature to boost the sales of their professional (Quadro/Tesla) cards. I do not know yet of the effects of the half-precision performance, but nVidia seems to have made the wrong choice there too by offering Half-precision at 1/64th of the Single precision while AMD offers it at the same GFLOPS as the single precision. There ought to be projects where AMD cards smash the nVidia cards, performance-wise.