Available GPUs

Volta-generation data center GPU with 32 GB HBM2 and Tensor Cores for AI training and inference.

• VRAM 32 GB HBM2
• FP16 31.3 TFLOPS

AMD flagship data center GPU with CDNA 2 architecture and 128 GB HBM2e for HPC and AI workloads.

• VRAM 64 GB HBM2e
• FP16 383 TFLOPS

High-performance Ampere consumer GPU with 10 GB GDDR6X and 2nd-gen RT Cores for gaming and rendering.

• VRAM 20 GB GDDR6X
• FP16 29.8 TFLOPS

Budget-friendly Ampere desktop GPU with 6 GB GDDR6 and RT Cores for entry-level ray tracing.

• VRAM 6 GB GDDR6
• FP16 6.8 TFLOPS

Professional mobile workstation GPU with 4 GB ECC GDDR6 and certified drivers for ISV applications.

• VRAM 8 GB GDDR6
• FP16 8.6 TFLOPS

Entry-level Ampere mobile GPU for thin-and-light gaming laptops with RT Cores and DLSS.

• VRAM 8 GB GDDR6
• FP16 7.1 TFLOPS

Flagship Blackwell consumer GPU with 32 GB GDDR7 and next-gen RT Cores for extreme performance.

• VRAM 32 GB GDDR7
• FP16 104.8 TFLOPS

High-end Blackwell GPU with 24 GB GDDR7 and 4th-gen RT Cores for enthusiast gaming and AI.

• VRAM 16 GB GDDR7
• FP16 56.3 TFLOPS

Mid-range Blackwell GPU with 16 GB GDDR7, ideal for 1440p gaming and local AI workloads.

• VRAM 16 GB GDDR7
• FP16 23.7 TFLOPS

Budget-friendly RDNA 2 desktop GPU with 4 GB GDDR6 and high clock speeds for 1080p gaming.

• VRAM 4 GB GDDR6
• FP16 11.5 TFLOPS

Low-power Turing data center GPU with 16 GB GDDR6 and Tensor Cores optimized for inference.

• VRAM 16 GB GDDR6
• FP16 65.1 TFLOPS

Entry-level Turing workstation GPU with 8 GB GDDR6 and low-profile form factor for CAD and office use.

• VRAM 8 GB GDDR6
• FP16 5 TFLOPS

Efficient entry-level mobile GPU based on Ada Lovelace architecture with 6 GB GDDR6 and AV1 encoding.

• VRAM 6 GB GDDR6
• FP16 17.8 TFLOPS