NVIDIA PROFESSIONAL GRAPHICS FOR Artificial Intelligence

AI Drives Real Business Success

Customer services, like Pinterest, use deep learning for image recognition, curating rich custom experiences for its users. Financial businesses like PayPal are using NVIDIA GPU-accelerated deep learning for fraud detection. Consulting firm Accenture’s R&D arm and other businesses are using deep learning to detect Internet security threats. And healthcare is one of the most important industries for the application of AI and deep learning. This sweeping change in the landscape of data-driven business is taking place across all industries. It’s changing how farmers herd their animals and forcing us to rethink transportation. The first self-driving public transit system is already up and running, giving us a glimpse into futuristic cities that will exist within our lifetimes.

Fueling All Areas of Business

AECO

Healthcare

Media & Entertainment

Government

Higher Education

Manufacturing

Request More Information

NVIDIA PROFESSIONAL GRAPHICSACCELERATING AI WITH NVIDIA RTX™ PROFESSIONAL GPUs

NVIDIA AI Enterprise

End-to-End Platform for Production AI

NVIDIA Professional Graphics Cards

Powering Professionals to Spearhead Innovation

NVIDIA Data Center GPUs

Accelerate generative AI, Simulations, and More

NVIDIA IGX ORIN

Enhanced Safety & Security for Edge AI

How to train an AI model

Three Steps

What is AI Inferencing?

Compare NVIDIA RTX Graphics Solutions

NVIDIA Ada Lovelace Architecture

	NVIDIA RTX 6000 Ada	NVIDIA RTX 5000 Ada	NVIDIA RTX 4500 Ada	NVIDIA RTX 4000 Ada	RTX 4000 SFF Ada
GPU Memory	48GB GDDR6 ECC	32GB GDDR6 ECC	24GB GDDR6 ECC	20GB GDDR6 ECC	20GB GDDR6 ECC
Memory Bandwidth	960 GB/s	576 GB/s	432 GB/s	360 GB/s	320 GB/s
CUDA Cores	18176	12800	7680	6144	6144
Tensor Cores	568	400	240	192	192
RT Cores	142	100	60	48	48
Dimensions	4.4" x 10.5" FH \| DS	4.4" x 10.5" FH \| DS	4.4" x 10.5" FH \| DS	4.4" x 9.5" FH \| SS	4.7" x 6.6" LP \| DS
Power & Thermal	300 W - Active	250 W - Active	210 W - Active	130 W - Active	70 W - Active
	Learn More HW Sync Version	Learn More HW Sync Version	Learn More HW Sync Version	Learn More HW Sync Version	Learn More HW Sync Version

NVIDIA Ampere Architecture

	NVIDIA RTX A6000	NVIDIA A800 40GB Active	NVIDIA RTX A5500	NVIDIA RTX A5000	NVIDIA RTX A4500	NVIDIA RTX A4000	NVIDIA RTX A2000 12GB
GPU Memory	48GB GDDR6 ECC	40GB HBM2	24GB GDDR6 ECC	24GB GDDR6 ECC	20GB GDDR6 ECC	16GB GDDR6 ECC	12GB GDDR6 ECC
Memory Bandwidth	768 GB/s	1555.2 GB/s	768 GB/s	768 GB/s	640 GB/s	448 GB/s	288 GB/s
CUDA Cores	10752	6912	10240	8192	7168	6144	3328
Tensor Cores	336	432	320	256	224	192	104
RT Cores	84	N/A	80	64	56	48	26
Dimensions	4.4" x 10.5" FH \| DS	4.4" x 10.5" \| DS	4.4" H x 10.5" FH \| DS	4.4" x 10.5" FH \| DS	4.4" x 10.5" FH \| DS	4.44" x 9.5" FH \| SS	2.713" x 6.6" LP \| DS
Power & Thermal	300 W - Active	240W - Active	230 W - Active	230 W - Active	200 W - Active	140 W - Active	70 W - Active
	Learn More HW Sync Version	Learn More	Learn More	Learn More HW Sync Version	Learn More	Learn More HW Sync Version	Learn More

NVIDIA Turing Architecture

	NVIDIA T1000 8GB	NVIDIA T1000	NVIDIA T400 4GB
GPU Memory	8 GB GDDR6	4 GB GDDR6	4 GB GDDR6
Memory Bandwidth	160 GB/s	160 GB/s	80 GB/s
CUDA Cores	896	896	384
Dimensions	2.713" x 6.137" LP \| SS	2.713" x 6.137" LP \| SS	2.713" x 6.137" LP \| SS
Power & Thermal	50 W - Active	50 W - Active	30 W - Active
	Learn More	Learn More	Learn More

Generative AI Comparisons
From Ampere to Ada Lovelace

Select the GPU(s) to compare

Relative Performance

Tests run on an Intel Core i9-12900K Processor @ 3.2GHz (5.2GHz Turbo), 64GB RAM, Windows 11 Enterprise x64, Stable Diffusion WebUI v1.3.1, NVIDIA Driver 536.15. Relative speedup for 512x512 image generation. Performance based on pre-released build, subject to change.

Generative AI models use neural networks to identify the patterns and structures within existing data to generate new and original content. The Ada Lovelace architecture’s Fourth-Generation Tensor Cores are designed to accelerate transformative AI technologies like generative AI.