NVIDIA · GPU
RTX PRO 4000 Blackwell SFF Edition RTX PRO 4000 Blackwell SFF Edition — Specifications and benchmarks. .
These specifications are supported by the linked sources. Core counts, clocks and memory capacities do not predict workload performance. Check benchmark versions, environments and sample counts.
Evidence review in progress · Search indexing deferred
Compare this product →
Maximum supported memory differs from installed RAM. Unified or shared memory is not dedicated VRAM. TDP, base power and maximum power have different definitions and conditions. System measurements do not establish the performance of an isolated chip. Power Parameter Value and conditions Source-reported maximum GPU power 70 W Source and conditions View source ↗
{
"state": "known",
"raw_label": "Max Power Consumption",
"raw_value": "70 W",
"qualifiers": {
"reference_configuration": true,
"source_architecture_section": "NVIDIA Blackwell Architecture",
"source_field": "Max Power Consumption",
"source_row": 7
},
"document_id": "nvidia-workstation-specs-d65c4fe926a4dde0"
}
Compute Parameter Value and conditions Architecture NVIDIA Blackwell Architecture Source and conditions View source ↗
{
"state": "known",
"raw_label": "Architecture",
"raw_value": "NVIDIA Blackwell Architecture",
"qualifiers": {
"reference_configuration": true,
"source_architecture_section": "NVIDIA Blackwell Architecture",
"source_field": "Architecture",
"source_row": 7
},
"document_id": "nvidia-workstation-specs-d65c4fe926a4dde0"
}
Memory Parameter Value and conditions Graphics memory type GDDR7 Source and conditions View source ↗
{
"state": "known",
"raw_label": "GPU Memory",
"raw_value": "GDDR7",
"qualifiers": {
"reference_configuration": true,
"source_architecture_section": "NVIDIA Blackwell Architecture",
"source_combined_value": "24 GB GDDR7 with ECC",
"source_field": "GPU Memory",
"source_row": 7
},
"document_id": "nvidia-workstation-specs-d65c4fe926a4dde0"
} Dedicated VRAM 24 GB Source and conditions View source ↗
{
"state": "known",
"raw_label": "GPU Memory",
"raw_value": "24 GB",
"qualifiers": {
"reference_configuration": true,
"source_architecture_section": "NVIDIA Blackwell Architecture",
"source_combined_value": "24 GB GDDR7 with ECC",
"source_field": "GPU Memory",
"source_row": 7
},
"document_id": "nvidia-workstation-specs-d65c4fe926a4dde0"
} 24 GB Source and conditions View source ↗
{
"state": "known",
"raw_label": "Graphics Memory Size",
"raw_value": "24",
"qualifiers": {
"reference_configuration": true,
"source_architecture_section": "NVIDIA Blackwell Architecture",
"source_field": "Graphics Memory Size",
"source_row": 7
},
"document_id": "nvidia-workstation-specs-d65c4fe926a4dde0"
}
Connectivity and features Parameter Value and conditions CUDA compute capability 12.0 Source and conditions View source ↗
{
"state": "known",
"raw_label": "CUDA Compute Capability",
"raw_value": "12.0",
"qualifiers": {
"evidence_scope": "compute-capability-listing; not runtime-compatibility guarantee",
"source_field": "CUDA Compute Capability",
"source_row": 10
},
"document_id": "nvidia-cuda-current-adea1d9c5294e1ce"
}
Public benchmarks Compare results with matching versions, scenes, settings and statistics. A run count of 0 identifies an aggregate or reported value, not a newly measured run.
V-Ray (1) V-Ray 6 (build unknown) v6-current-cuda 2,988 vpaths
Run count: 0 (aggregate / reported value)
View source ↗
{
"comparison_readiness": "source-aggregate-only",
"condition": {
"benchmark": "V-Ray",
"environment": {
"backend": "CUDA",
"configuration": [
{
"deviceType": "GpuDevice",
"name": "NVIDIA RTX PRO 4000 Blackwell SFF Edition 24466MB",
"tag": "Device"
}
],
"evidence_type": "source-aggregate",
"population": "Chaos public submitted runs; mixed systems",
"scene_identity_status": "current public v6 scene; asset checksum unavailable",
"settings_status": "unknown",
"source_configuration_id": 3040,
"source_device_name": "NVIDIA RTX PRO 4000 Blackwell SFF Edition 24466MB",
"source_max": 2988,
"source_median": 2988,
"source_model_url": "https://benchmark.chaos.com/vray/v6/vray-gpu/3040",
"source_sample_count": 1,
"statistic": "mean"
},
"scene": "v6-current-cuda",
"scene_checksum": "chaos-v6-current-workload:cuda",
"version": "6 (build unknown)"
},
"condition_id": "ac31b79c7e2b02fb7a66",
"document_id": "bench-64817401c564f226-015baafd3482368d",
"example_record_ids": [
"3040"
],
"measured_from": null,
"measured_to": null,
"parameter_id": "benchmark.vray6_cuda",
"population": "Chaos public submitted runs; mixed systems",
"product_id": "rtx-pro-4000-blackwell-sff-edition",
"record_count": 1,
"reported_value": 2988,
"run_count": 0,
"sample_status": "source-aggregate",
"source_sample_count": 1,
"statistic": "mean"
} Sources Source terms and retrieval dates are listed below. Published values and conditions may change.
View source ↗ 2026-10-09 Official NVIDIA workstation comparison tables; architecture sections and explicit memory variants retained View source ↗ 2026-10-09 Official NVIDIA CUDA compute-capability tables View source ↗ 2026-10-09 Published numeric/categorical facts; original page retained locally; exact catalog model matching only View source ↗ 2026-10-09 Official NVIDIA current and legacy supported-chip tables; device names and PCI identifiers only View source ↗ 2026-10-09 Public benchmark facts; source aggregates retained; original snapshot local