ATI Radeon HD 3870 vs NVIDIA GeForce GTS 250
Processeur graphique
| Nom du GPU | RV670 | G92B |
|---|---|---|
| GPU variant | RV670 XT (215-0708005) | G92-426-B1 |
| Architecture | TeraScale | Tesla |
| Fonderie | TSMC | TSMC |
| Taille du processus | 55 nm | 55 nm |
| Transistors | 666 million | 754 million |
| Taille des matrices | 192 mm² | 260 mm² |
Carte graphique
| Date de publication | Nov 19th, 2007 | Mar 4th, 2009 |
|---|---|---|
| Génération | Radeon R600 | GeForce 200 |
| Production | End-of-life | End-of-life |
| Prix de lancement | 269 USD | 199 USD |
| Bus interface | PCIe 2.0 x16 | PCIe 2.0 x16 |
| Critiques | 80 in our database | 58 in our database |
| Prédécesseur | Radeon R500 PCIe | GeForce 9 |
| Successeur | Radeon R700 | GeForce 400 |
Vitesse clock
| GPU clock | 777 MHz | 675 MHz |
|---|---|---|
| Mémoire clock | 1126 MHz 2.3 Gbps effective | 1008 MHz 2 Gbps effective |
| Shader clock | — | 1620 MHz |
Mémoire
| Taille de la mémoire | 512 MB | 1024 MB |
|---|---|---|
| Type de mémoire | GDDR4 | GDDR3 |
| Bus mémoire | 256 bit | 256 bit |
| Bande passante | 72.06 GB/s | 64.51 GB/s |
Render Config
| Shading units | 320 | 128 |
|---|---|---|
| TMUs | 16 | 64 |
| ROPs | 16 | 16 |
| Compute units | 4 | — |
| Cache L2 | 256 KB | 64 KB |
| Comptage SM | — | 16 |
Performance théorique
| Taux de pixel | 12.43 GPixel/s | 10.80 GPixel/s |
|---|---|---|
| Taux de texture | 12.43 GTexel/s | 43.20 GTexel/s |
| FP32 (float) performance | 497.3 GFLOPS | 414.7 GFLOPS |
| FP64 (double) performance | 99.46 GFLOPS (1:5) | — |
Board Design
| Taille Slot | Dual-slot | Dual-slot |
|---|---|---|
| Longueur | 241 mm 9.5 inches | 229 mm 9 inches |
| TDP | 106 W | 150 W |
| Suggestions pour le PSU | 300 W | 450 W |
| Sorties | 2x DVI1x S-Video | 2x DVI |
| Connecteurs de puissance | 1x 6-pin | 1x 6-pin |
| Board number | B339 | P365 |
Caractéristiques graphiques
| DirectX | 10.1 (10_1) | 11.1 (10_0) |
|---|---|---|
| OpenGL | 3.3 (full) 4.0 (partial) | 3.3 |
| OpenCL | — | 1.1 |
| Vulkan | — | — |
| Shader model | 4.1 | 4.0 |
| CUDA | — | 1.1 |
