1. CUDA parallel processing cores cannot be compared between GPU generations due to several important architectural differences that exist between streaming multiprocessor designs. 2. Maximum possible power consumption including the Dynamic Boost algorithm. For system specific GPU TGP, please consult your OEM/solution provider. 3. Display support varies by system-level implementation. Check with your workstation OEM vendor for system specific configurations. Adaptors available for DVI-SL, DVI-DL, HDMI, and VGA. 4. Peak rates are based on GPU boost clock.  5. Effective TFLOPS using the sparsity feature. NVIDIA Ada Lovelace architecture using FP8 matrix multiply with FP16 or FP32 accumulate; NVIDIA Ampere architecture using FP16 matrix multiply with FP16 or FP32 accumulate. 6. Number of NVENC and NVDEC may vary by GPU. GPU specific details can be found here: https://developer.nvidia.com/video-encode-and-decode-gpu-support-matrix-new 7. Ensures data integrity and reliability by eliminating soft errors on direct random-access memory (DRAM) only. 8. Peak FLOP and display support for NVIDIA RTX A500 Laptop GPU varies by system configuration. Check with your OEM system vendor to confirm which specification is supported. 9. FP8 TOPS with Sparsity.