inference/research
verifiedhigh · 0.85updated 2026-10-04

NVIDIA RTX PRO 4500 Blackwell Workstation Edition datasheet (NVIDIA-hosted PDF)

spec-sheet

Fields

id
sources/aa-nv-rtx-pro-4500-blackwell-datasheet
Publisher
NVIDIA
Kind
spec-sheet
Accessed
2026-10-04
notes
FIRST-PARTY nvidia.com-hosted copy - unlike nv-blackwell-datasheet and nv-blackwell-ultra-datasheet, which the repo could only reach on a reseller domain. Document code 5108623. APR26. Extracted spec table verbatim: 'GPU architecture NVIDIA Blackwell', 'CUDA Cores 10,496', 'Tensor Cores 5th Generation', 'AI TOPS 1,617 AI TOPS 1,2', 'Single-precision performance 51 TFLOPS 1', 'RT Core performance 153 TFLOPS 1', 'GPU memory 32 GB GDDR7 with ECC', 'Memory interface 256-bit', 'Memory bandwidth 896 GB/s', 'System interface PCIe 5.0 x16', 'Display connectors 4x DisplayPort 2.1b', 'Power consumption Total board power: 200 W', 'Form factor 4.4in H x 10.5in L, dual slot, full height'. Footnotes: 1. 'Peak rates based on GPU Boost Clock.' 2. 'Effective FP4 TOPS using sparsity.' (footnote 2 text confirmed on the product page as 'Theoretical FP4 TOPS using sparsity'). Added by audit 2026-10-04 because the part had no accelerator record despite shipping in the same MLPerf v6.1 wave as the RTX PRO 6000 and the Arc Pro B70.

Notes

FIRST-PARTY nvidia.com-hosted copy - unlike nv-blackwell-datasheet and nv-blackwell-ultra-datasheet, which the repo could only reach on a reseller domain. Document code 5108623. APR26. Extracted spec table verbatim: 'GPU architecture NVIDIA Blackwell', 'CUDA Cores 10,496', 'Tensor Cores 5th Generation', 'AI TOPS 1,617 AI TOPS 1,2', 'Single-precision performance 51 TFLOPS 1', 'RT Core performance 153 TFLOPS 1', 'GPU memory 32 GB GDDR7 with ECC', 'Memory interface 256-bit', 'Memory bandwidth 896 GB/s', 'System interface PCIe 5.0 x16', 'Display connectors 4x DisplayPort 2.1b', 'Power consumption Total board power: 200 W', 'Form factor 4.4in H x 10.5in L, dual slot, full height'. Footnotes: 1. 'Peak rates based on GPU Boost Clock.' 2. 'Effective FP4 TOPS using sparsity.' (footnote 2 text confirmed on the product page as 'Theoretical FP4 TOPS using sparsity'). Added by audit 2026-10-04 because the part had no accelerator record despite shipping in the same MLPerf v6.1 wave as the RTX PRO 6000 and the Arc Pro B70.

Cross-references

This record references no other records.