 Unterschiede hervorheben |  Nvidia RTX A400 mit 4 GB GDDR6-RAM |
Nvidia RTX A1000 mit 8 GB GDDR6-RAM |
Nvidia RTX 2000 Ada mit 16 GB GDDR6 ECC-RAM |
Nvidia RTX PRO 2000 Blackwell mit 16 GB GDDR7 ECC-RAM |
Nvidia L4 mit 24 GB GDDR6-RAM |
| Chipsatz | RTX A400 | RTX A1000 | RTX 2000 Ada | RTX PRO 2000 | L4 |
| GPU Architektur | Ampere | Ampere | Ada | Blackwell | Ada |
| Speicher | 4 GB | 8 GB | 16 GB | 16 GB | 24 GB |
| Speicherart | GDDR6 | GDDR6 | GDDR6 ECC | GDDR7 ECC | GDDR6 |
| Speicheranbindung | 64 Bit | 128 Bit | 128 Bit | 128 Bit | 192 Bit |
| Speicherbandbreite | 96 GB/s | 192 GB/s | 224 GB/s | 288 GB/s | 300 GB/s |
| GPU-Taktung | 727 MHz | 727 MHz | 1620 MHz | 1957 MHz | 795 MHz |
| CUDA / Streamprozessoren | 768 | 2304 | 2816 | 4352 | 7680 |
| Gleitkomma-Rechenleistung (FP32) | 2,706 TFLOPS | 6,737 TFLOPS | 12 TFLOPS | 17,03 TFLOPS | 30,31 TFLOPs |
| Steckplatz | PCIe 4.0 x16 (x8) | PCIe 4.0 x16 (x8) | PCIe 4.0 x16 | PCIe 5.0 x16 (x8) | PCIe 4.0 x16 |
| Slotbreite | 1-Slot LP | 1-Slot LP | 2 Slots LP | 2 Slots LP | 1 Slot |
| Anschlüsse | 4x mDP 1.4a | 4x mDP 1.4a | 4x mDP 1.4a | 4x mDP 2.1b | |
| Kühlung | active | active | active | active | passive |
| Verbrauch (Watt) | 50W | 50W | 70W | 70W | 72W |
| DirectX | 12_2 Ultimate | 12_2 Ultimate | 12_2 Ultimate | 12_2 Ultimate | 12_2 Ultimate |
| Open-GL | 4.6 | 4.6 | 4.6 | 4.6 | 4.6 |
| Open-CL | 3.0 | 3.0 | 3.0 | 3.0 | 3.0 |
| Vulkan | 1.3 | 1.3 | 1.3 | 1.4 | 1.3 |
| Shader Model | 6.8 | 6.8 | 6.8 | 6.8 | 6.7 |
| Max. Auflösung (Pixel) | 5120×2880 | 5120×2880 | 7680x4320 | 7680x4320 | 0 |
| Max. Anzahl Monitore | 4 | 4 | 4 | 4 | |
| Zusatzfunktionen | 6 RT Cores, 16 ROPs, 24 Tensor Cores, 6 SM Count, 24 TMUs, RTX IO, DirectCompute, Mosaic, 1x encode/decode (+AV1) | 18 RT Cores, 32 ROPs, 72 Tensor Cores, 18 SM Count, 72 TMUs, RTX IO, DirectCompute, Mosaic, 1x encode, 2x decode (+AV1 decode) | 22 RT Cores, 48 ROPs, 88 Tensor Cores, NVIDIA AI, DLSS 3, RTX IO, DirectCompute, GPUDirect, OpenCL, Mosaic, 1x encode/decode (+AV1 encode and decode) | 545 AI TOPS, 34 RT Cores, 64 ROPs, 136 Tensor Cores (5th Gen), DLSS 4, FP4, NVIDIA AI, RTX IO, DirectCompute, GPUDirect, OpenCL, AV1 encode (9th Gen) and decode (6th Gen), VP11 | 60 RT Cores, 240 Tensor Cores, NVIDIA vPC/vApps/vWS, NVIDIA vGPU, NVIDIA vCS, NVIDIA AI Enterprise, DLSS 3, 2 NVENC, 4 NVDEC, 4 JPEG decoders, AV1 |
