 Unterschiede hervorheben |  PNY Nvidia T400 mit 4 GB GDDR6-RAM |
PNY Nvidia T1000 mit 4 GB GDDR6-RAM |
PNY Nvidia T1000 mit 8 GB GDDR6-RAM |
PNY Nvidia RTX A2000 mit 6 GB GDDR6 ECC-RAM |
PNY Nvidia RTX A2000 mit 12 GB GDDR6 ECC-RAM |
PNY Nvidia RTX A4000 mit 16 GB GDDR6 ECC-RAM |
Nvidia RTX 4000 Ada mit 20 GB GDDR6 ECC-RAM |
PNY Nvidia RTX A4500 mit 20 GB GDDR6 ECC-RAM |
Nvidia RTX 4500 Ada mit 24 GB GDDR6 ECC-RAM |
PNY Nvidia RTX A5000 mit 24 GB GDDR6 ECC-RAM |
Nvidia RTX 5000 Ada mit 32 GB GDDR6 ECC-RAM |
Nvidia RTX A6000 mit 48 GB GDDR6 ECC-RAM |
Nvidia RTX 6000 Ada mit 48 GB GDDR6 ECC-RAM |
| Chipsatz | T400 | T1000 | T1000 | RTX A2000 | RTX A2000 | RTX A4000 | RTX 4000 Ada | RTX A4500 | RTX 4500 Ada | RTX A5000 | RTX 5000 Ada | A100 | RTX 6000 Ada |
| GPU Architektur | Turing | Turing | Turing | Ampere | Ampere | Ampere | Ada | Ampere | Ada | Ampere | Ada | Ampere | Ada Lovelace |
| Speicher | 4 GB | 4 GB | 8 GB | 8 GB | 12 GB | 16 GB | 20 GB | 20 GB | 24 GB | 24 GB | 32 GB | 80GB | 48 GB |
| Speicherart | GDDR6 | GDDR6 | GDDR6 | GDDR6 ECC | GDDR6 ECC | GDDR6 ECC | GDDR6 ECC | GDDR6 ECC | GDDR6 ECC | GDDR6 ECC | GDDR6 ECC | HBM2e ECC on by Default | GDDR6 ECC |
| Speicheranbindung | 64 Bit | 128 Bit | 128 Bit | 192 Bit | 192 Bit | 256 Bit | 160 Bit | 320 Bit | 192 Bit | 384 Bit | 256 Bit | 5120 Bit | 384 Bit |
| Speicherbandbreite | 80 GB/s | 160 GB/s | 160 GB/s | 288 GB/s | 288 GB/s | 448 GB/s | 360 GB/s | 640 GB/s | 432 GB/s | 768 GB/s | 576 GB/s | 1935 GB/s | 960 GB/s |
| GPU-Taktung | 1070 MHz | 1395 MHz | 1395 MHz | 1200 MHz | 1200 MHz | 1560 MHz | 1500 MHz | 1650 MHz | 2070 MHz | 1695 MHz | 1155 MHz | 1065 MHz | 915 MHz |
| CUDA / Streamprozessoren | 384 | 896 | 896 | 3328 | 3328 | 6144 | 6144 | 7168 | 7680 | 8192 | 12800 | 6912 | 18176 |
| Gleitkomma-Rechenleistung (FP32) | 1,094 TFLOPs | 2,13 TFLOPs | 2,50 TFLOPs | 7,987 TFLOPS | 7,987 TFLOPS | 19,2 TFLOPS | 26,73 TFLOPS | 23,7 TFLOPS | 39,60 TFLOPS | 27,8 TFLOPs | 65,31 TFLOPs | 19,49 TFLOPS | 91,06 TFLOPs |
| Steckplatz | PCIe 3.0 x16 | PCIe 3.0 x16 | PCIe 3.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 |
| Slotbreite | 1 Slot | 1 Slot | 1 Slot | 2 Slot | 2 Slot | 1 Slot | 1 Slot | 1 Slot | 2 Slots | 2 Slots | 2 Slots | 2 Slot | 2 Slots |
| Anschlüsse | 3x mDP 1.4a | 4x mDP 1.4 | 4x mDP 1.4a | 4x mDP 1.4a | 4x mDP 1.4a | 4x DP 1.4 | 4x DP 1.4a | 4x DP 1.4a | 4x DP 1.4a | 4x DP 1.4 | 4x DP 1.4a | 4x DP 1.4a + VirtualLink | |
| Kühlung | active | active | active | active | active | active | active | active | active | active | active | passive | active |
| SLI / CrossFire | NVLink™2 | NVLink | NVLink | ||||||||||
| Verbrauch (Watt) | 30W | 50W | 50W | 70W | 70W | 140W | 130W | 200W | 210W | 230W | 250W | 300W | 300W |
| DirectX | 12 | 12 | 12 | 12.07 | 12.07 | 12_2 | 12_2 Ultimate | 12_2 | 12_2 Ultimate | 12_2 | 12_2 Ultimate | 12_2 Ultimate | |
| Open-GL | 4.5 | 4.5 | 4.5 | 4.68 | 4.68 | 4.6 | 4.6 | 4.6 | 4.6 | 4.6 | 4.6 | 4.68 | |
| Open-CL | 2.0 | 2.0 | 2.0 | 2.0 | 2.0 | 2.0 | 3.0 | 2.0 | 3.0 | 2.0 | 3.0 | 3.0 | 3.0 |
| Vulkan | 1.0 | 1.0 | 1.0 | 1.2 | 1.2 | 1.2 | 1.3 | 1.2 | 1.3 | 1.2 | 1.3 | 1.3 | |
| Shader Model | 5.1 | 5.1 | 5.1 | 6.6 | 6.6 | 6.5 | 6.7 | 6.5 | 6.7 | 6.5 | 6.7 | 6.7 | |
| Max. Auflösung (Pixel) | 5120×2880 | 7680x4320 | 7680x4320 | 7680x4320 | 7680x4320 | 7680x4320 | 7680x4320 | 7680x4320 | 7680x4320 | 7680x4320 | 7680x4320 | 7680x4320 | |
| Max. Anzahl Monitore | 3 | 4 | 4 | 4 | 4 | 4 | 4 | 4 | 4 | 4 | 4 | 4 | |
| Zusatzfunktionen | Nvidia CUDA C, CUDA C++, DirectCompute 5.0, OpenCL, Java, Python, and Fortran | Nvidia CUDA C, CUDA C++, DirectCompute 5.0, OpenCL, Java, Python, and Fortran | Nvidia CUDA C, CUDA C++, DirectCompute 5.0, OpenCL, Java, Python, and Fortran | CUDA, DirectCompute, OpenCL, HDR, 4K, Mosaic | NVLink, CUDA, DirectCompute, OpenCL, HDR, 5K, Mosaic | NVLink, CUDA, DirectCompute, OpenCL, HDR, 5K, Mosaic | 48 RT Cores, 192 Tensor Cores, CUDA, RTX IO, DirectCompute, GPUDirect, OpenCL, Mosaic, 2x encode/decode (+AV1) | NVLink, CUDA, DirectCompute, OpenCL, HDR, 5K, Mosaic | 60 RT Cores, 240 Tensor Cores, CUDA, RTX IO, DirectCompute, GPUDirect, OpenCL, Mosaic, 2x encode/decode (+AV1) | NVLink, CUDA, DirectCompute, OpenCL, HDR, 5K, Mosaic | 100 RT Cores, 400 Tensor Cores, CUDA, NVIDIA vPC/vApps/vWS, RTX IO, DirectCompute, GPUDirect, OpenCL, Mosaic, 2x encode/decode (+AV1) | NVIDIA Virtual Compute Server with MIG support | NVIDIA vPC/vApps/vWS, NVIDIA vGPU, NVIDIA GPUDirect, NVIDIA GPU Boost, NVIDIA RTX IO, NVIDIA Mosaic, AV1 encode and decode |
