What is the memory capacity and type of the NVIDIA 900-5G132-1770-000?

The RTX A5500 features 24GB GDDR6 memory with ECC support on a 384-bit interface, delivering 768 GB/s bandwidth. ECC prevents silent data corruption critical for scientific computing, rendering farms, and AI training where reproducibility is essential.

How many CUDA, Tensor, and RT cores does it have?

It includes 10,240 CUDA cores for general compute, 320 third-gen Tensor cores for 272.8 TFLOPS AI acceleration, and 80 RT cores for 66.6 TFLOPS ray tracing. This enables balanced performance in DCC tools like Maya, Blender for real-time previews.

What interfaces and form factor does the RTX A5500 use?

PCIe 4.0 x16 system interface, dual-slot form factor (4.4in H x 10.5in L), 4x DisplayPort 1.4a outputs up to 8K@60Hz, and 1x 8-pin power connector at 230W max. Fits standard workstation chassis with NVLink for multi-GPU.

Is NVLink supported and what bandwidth?

Yes, 2-way NVLink with low-profile bridges provides 112.5 GB/s bidirectional bandwidth, scaling to 48GB combined memory for larger models in NLP, simulations, and VFX scenes requiring massive VRAM.

What are ideal workloads for this GPU?

Optimized for professional visualization, AI inference/training, ray-traced rendering, simulations, and VFX. Ampere architecture with ECC suits data centers, engineering workstations handling complex geometry, high-res textures, and compute-intensive tasks.

What display and encoding capabilities?

4x DisplayPort 1.4a supports 7680x4320@60Hz, NVIDIA Quadro Sync II for frame lock, 1x NVENC encoder, 2x NVDEC decoders with AV1 support for pro video workflows and multi-monitor setups.

900-5G132-1770-000 - NVIDIA 900-5G132-1770-000 24GB 10K Cores

Name: NVIDIA 900-5G132-1770-000 24GB 10K Cores
Brand: NVIDIA
SKU: 900-5G132-1770-000

900-5G132-1770-000 - NVIDIA 900-5G132-1770-000 24GB 10K Cores | US IT Hardware

Product overview

The NVIDIA RTX A5500 (SKU 900-5G132-1770-000) is a high-end professional workstation GPU built on the Ampere architecture, featuring 24GB of GDDR6 memory with ECC for error-free compute in demanding environments. It includes 10,240 CUDA cores, 320 third-generation Tensor cores for AI acceleration up to 272.8 TFLOPS, and 80 RT cores delivering 66.6 TFLOPS for real-time ray tracing. With a 384-bit memory interface providing 768 GB/s bandwidth and PCIe 4.0 x16 connectivity, it supports high-throughput data transfer ideal for large datasets. The dual-slot design with active cooling handles 230W TDP via single 8-pin power, fitting standard workstations. NVLink enables 2-way scaling with 112.5 GB/s bidirectional bandwidth for multi-GPU setups. Key applications include 3D rendering, machine learning model training, architectural visualization, VFX production, scientific simulations, and data analytics, offering precision, stability, and performance for enterprise professionals.

24GB GDDR6 ECC memory ensures data integrity in long compute jobs.
10,240 CUDA cores deliver 34.1 TFLOPS single-precision performance.
384-bit interface provides 768 GB/s bandwidth for large datasets.
PCIe 4.0 x16 enables fast CPU-GPU data transfer in modern systems.
NVLink supports 112.5 GB/s 2-way GPU interconnect scalability.
230W TDP with active dual-slot cooling for sustained workloads.

Expedited and overnight shipping options are available at checkout, priced by destination and service.
Purchase orders accepted from qualified institutions.
Refunds within 14 days and replacements within 30 days of delivery with an RMA (25% restocking fee on customer-initiated returns), and every part carries our Standard 30-Day Warranty - see our return policy or contact US IT Hardware.

Featuring

Thermal Solution
Manufacturer part #	900-5G132-1770-000
Brand	NVIDIA
Condition	New
Die Size	628.4 mm²
RT Cores	80 (2nd gen)
CUDA Cores	10,240
Transistors	28.3 billion
Architecture	NVIDIA Ampere
Tensor Cores	320 (3rd gen)
RT Performance	66.6 TFLOPS
Single Precision	34.1 TFLOPS

Compatible systems

Compatible with modern PCIe 4.0 workstations; requires adequate PSU and chassis airflow for sustained loads.

Dell Precision
Model: 7000 Series
Form factor: Tower/Workstation
PCIe 4.0 slot required for full performance.
HP Z Series
Model: Z8/Z6
Form factor: Workstation
NVLink compatible in select models.
Lenovo ThinkStation
Model: P620
Form factor: Tower
Supports Threadripper PRO with PCIe 4.0.

Best for

3D Rendering
Accelerates complex scene rendering with RT cores and high memory capacity for animation pipelines.
AI/ML Training
Tensor cores enable fast model training on large datasets with ECC memory for accuracy.
Architectural Viz
Real-time ray tracing for photorealistic building simulations and walkthroughs.
VFX Production
High TFLOPS performance for effects compositing and motion graphics workflows.

Details

Shipping, warranty & returns

Customer reviews

No reviews yet — be the first to review this part.

NVLink	2-way 112.5 GB/s
NVENC/NVDEC	1 encoder \| 2 decoders + AV1
Max Resolution	7680x4320@60Hz
Display Outputs	4x DisplayPort 1.4a
Power Connector	1x 8-pin PCIe
System Interface	PCIe 4.0 x16

NVIDIA 900-5G132-1770-000 24GB 10K Cores

Product overview

Featuring

Compatible systems

Best for

Details

Shipping, warranty & returns

Customer reviews

Storage

Power Consumption

Physical Characteristics

Display

Interface/Ports

Frequently asked questions

GPU Memory	24GB GDDR6 ECC
Memory Bandwidth	768 GB/s
Memory Interface	384-bit