NVIDIA RTX PRO 5500 Blackwell Workstation Edition NEW

Overview
The NVIDIA RTX PRO 5500 Blackwell Workstation Edition is a professional Blackwell desktop card with 84 GB of GDDR7 and ECC, 1,398 GB/s of memory bandwidth and a 600 W board power, announced in September 2026 and not yet shipping. It slots into the gap between the RTX PRO 5000 Blackwell, which tops out at 72 GB and 300 W, and the RTX PRO 6000 Blackwell Workstation Edition at 96 GB and 600 W, a gap that had been wide enough to leave buyers choosing between half the memory or the full price. Its 21,760 CUDA cores are exactly the count in the GeForce RTX 5090, on the same full-size GB202 die, but with 84 GB of ECC memory in place of the 5090's 32 GB, professional driver certification and Multi-Instance GPU. That last part matters for the way NVIDIA is positioning the card: it is built for rack-mounted deployment rather than a desk, so an IT team can put the cards in a managed rack, split each one into two isolated 42 GB instances and hand workstation-class graphics and AI to people who never see the hardware. What NVIDIA has not published is any throughput figure. There is no single-precision number and no AI TOPS number on either NVIDIA's page or PNY's, so Flopper shows the datatypes the card supports and leaves the rates blank until a datasheet exists.
Performance
Peak theoretical throughput by precision type
| Precision | Dense | 2:4 Sparse |
|---|---|---|
FP64 64-bit floating point | No verified data available | Structured sparsity is a tensor-core feature; this vector precision has no sparse form |
FP32 32-bit floating point | No verified data available | Structured sparsity is a tensor-core feature; this vector precision has no sparse form |
TF32 TensorFloat-32 | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
BF16 Brain Float 16 | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
FP16 16-bit floating point | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
FP8 8-bit floating point | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
FP6 | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
FP4 4-bit floating point | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
INT8 8-bit integer | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
Flopper holds no verified throughput figures for this part yet. The rows stay so the gaps are visible.
Dense peak divided by accelerator TDP. Board power only: excludes host CPUs, networking, cooling and facility overhead. TDP for this part: 600 W.
Specifications
Architecture
Blackwell
Form Factor
PCIe
Launch Year
2026
Process Node
4nm
Memory
84 GB GDDR7
Bandwidth
1,398 GB/s
TDP
600 W
Max power
600 W
Transistors
92.2 billion
CUDA Cores
21,760
Spec Confidence
Vendor claimed
Full Specifications
| Compute Engine | |
|---|---|
| CUDA Cores | 21,760 |
| Streaming Multiprocessors | 170 |
| Chip Design | |
| Transistors | 92.2 billion |
| Die Size | 750 mm² |
| Process Node | 4nm |
| Memory | |
| Memory | 84 GB |
| Memory Type | GDDR7 |
| Bandwidth | 1.4 TB/s |
| Interface Width | 416-bit |
| Interconnect & I/O | |
| PCIe | 5.0 x16 |
| Power & Thermal | |
| TDP | 600 W |
| Max power | 600 W |
| Power Connector | 1x PCIe CEM5 16-pin |
| Cooling | Active |
| Enterprise Features | |
| ECC Memory | Yes |
| MIG Support | Up to 2x 42GB, Up to 1x 84GB |
| Sparsity | Yes |
| Physical & Media | |
| Card Length | 281.9 mm |
| Width | 2-slot |
| NVENC Engines | 3 |
| NVDEC Engines | 3 |
| General | |
| Form Factor | PCIe |
| Architecture | Blackwell |
| Process Node | 4nm |
| Launch Year | 2026 |
Datasheet & Resources
Flopper Datasheet
NVIDIA RTX PRO 5500 Blackwell Workstation Edition specifications, generated from our database. Printable.
NVIDIA RTX PRO 5500 Blackwell Workstation Edition product page
NVIDIA · Latest version
NVIDIA Developer Documentation
Technical specs, programming guides
Data Provenance
Every figure traced to a source
Primary Source
- Publisher
- NVIDIA
- Published
- No verified data available
Data Quality
- Spec confidence
- Vendor claimed
- Clock basis
- Boost
- Core precisions with figures
- 0 of 9
- Normalization
- All values in TFLOPS
Compare cloud providers offering on-demand GPU instances for AI training, inference, and HPC workloads.
Browse GPU Cloud ProvidersSimilar GPUs
NVIDIA RTX PRO 6000 Blackwell Workstation Edition
Frequently Asked Questions
How many TFLOPS does the NVIDIA RTX PRO 5500 Blackwell have?
Flopper does not currently have any verified throughput figures for the NVIDIA RTX PRO 5500 Blackwell.
What is the power consumption of the NVIDIA RTX PRO 5500 Blackwell?
The NVIDIA RTX PRO 5500 Blackwell has a TDP (Thermal Design Power) rating of 600 watts.
How much memory does the NVIDIA RTX PRO 5500 Blackwell have?
The NVIDIA RTX PRO 5500 Blackwell is equipped with 84 GB of memory with 1,398 GB/s of memory bandwidth.
What architecture is the NVIDIA RTX PRO 5500 Blackwell based on?
The NVIDIA RTX PRO 5500 Blackwell is based on the Blackwell architecture, launched in 2026.
Stay Updated on GPU Releases
Get notified when new GPUs are added or specifications are updated.
No spam, unsubscribe anytime.