Close this search box.

The NVIDIA A6000 & A40 GPUs

Built on the NVIDIA Ampere architecture

The NVIDIA Ampere Architecture

The Ampere Architecture, NVIDIA’s largest nanometer chip marks an inflection point for the company. Being the first ever smallest chip crafted by NVIDIA, it allows more room for increased number of transistors to be packed in a smaller area. Designed with 54 billion transistors, the 7 nanometer (nm) chip features 6 core pathbreaking innovations. The process shrink provides a great oppurtunity for NVIDIA to drastically improve on the previous generation RTX 20-series hardware and technologies. NVIDIA’s Ampere architecture is the second generation of consumer ray tracing and third geenration deep learning hardware. 

ampere architecture

The NVIDIA Ampere architecture simplifies and accelerates AI adoption to extract new insights from voluminous datasets, both on the Edge and in the cloud. Buit on the third generation tensor core technology, it brings new precisions – Tensor Float 32 (TF32) and floating point 64 (FP64), that radically brings down the training time, from hours to weeks, redering rapid acceleration to inference. Even without needing to make any code changes, it delivers a speedup of up to 20x for AI and double the performance. As every application may not require performance of a full GPU, the architecture prompts the distinct workloads to share the GPU with its Multi-Instance GPU (MIG) functionality. Additionally, the NVIDIA NVLink increases the GPU-to-GPU direct bandwidth to 600 gigabytes per second (GB/s), making it double and almost 10 times better than PCIe Gen4.

With the ability to convert it into a sparse model, whereby the unused parameters are converted to “zeros”, it ensures to boost the application performance while maintaining the accuracy. Crafted for the age of elastic computing, Ampere architecture delivers unparalled and unmatched speed at every scale for professionals to harness the power of AI for their most crucial and demanding industrial and scientific workloads. The architecture provides a massive leap in the performance, moving on top og the charts of the list of all the GPUs.

NVIDIA RTX A6000 Graphics Card

The NVIDIA RTX A6000 is an extremely powerful visual computing GPU for desktop workstations to meet the increasing demand of today’s innovative engineers, designers and artists. With the GPUs advanced features and cutting-edge performance, professionals can unlock the speed of inspiration for their path-breaking innovations, revolutionary designs, and mesmeric graphics and entertainment. Backed by the latest NVIDIA Ampere architecture, 84 second-generation RT Cores, 336 third-generation Tensor Cores, and 10,752 CUDA cores with 48 GB of graphics memory, NVIDIA RTX A6000 enables the professionals to design futuristic products.

With support for NVIDIA virtual GPU (vGPU) software, remote users are enabled to share resources to develop their state-of-the-art designs with AI, as their personal workstations are remodeled into various virtual workstations, each delivering incredible application performance concurrently. Combining second-generation RT Cores and third-generation Tensor Cores, NVIDIA RTX A6000 doubles the throughput and makes training throughput fivefold, comares to the previous generation, to speedup complex workloads like architectural design evaluations, photorealistic rendering of movie content, and virtual prototyping of product designs. Designed to fit into multiple types of workstations from global OEM vendors, the GPU is crafted with a power efficient design and delivers twice the power over Turing GPUs.

As work from home becomes rampant and the need for professional visualization advances expeditiously across different industries, the GPU continues to deliver enhanced performance for complex requirements such as powerful compute, engineering simulation, real time ray tracing, AI, immensive VR, and many more to name a few. Access accelerated speed with high performance by connecting two RTX A6000s wuth NVIDIA NVLink for 96 GB of combined GPU memory to unlock some of the most powerful graphic solutions.

NVIDIA RTX A6000 Graphics Card 48 GB GDDR6 GA102 GPU 9

Technical Specifications: NVIDIA A6000 GPU

GPU FeaturesNVIDIA RTX A6000
GPU Memory48 GB GDDR6 with error-correcting code (ECC)
Display Ports4x Display Port 1.4*
Max Power Consumption300 W
Graphics BusPCI Express Gen 4 x 16
Form Factor4.4” (H) x 10.5” (L) dual slot
NVLink2-way low profile (2-slot and 3-slot bridges) Connect 2 RTX A6000
vGPU Software SupportNVIDIA vPC/vApps, NVIDIA RTX Virtual Workstation, NVIDIA Virtual Compute Server
vGPU Profiles SupportedSee the Virtual GPU Licensing Guide
VR ReadyYes

NVIDIA RTX A40 Graphics Card

The NVIDIA A40 is an extremely powerful data center GPU that accelerates the most challenging visual computing workloads from the data center, by integrating powerful compute, professional graphics, and artificial intelligence speed. Backed by the latest NVIDIA Ampere architecture, RT Cores, CUDACores with 48 GB of graphics memory, and Tensor Cores, NVIDIA A40 brings the next generation NVIDIA RTX technology to the data center to deliver exceptional performance for multi-workload capabilities, ranging from dedicated render nodes to powerful virtual workstations that can be accessed from anywhere. The GPU brings to the table unique features for professionals demanding high performance and premier quality graphics for simulation, ray-traced rendering, virtual production and many more to name a few.

nvidia A40

Being the world’s strongest data center GPU for visual computing, NVIDIA A40 combines its ultra-modern features and accelerated performance for VR, large screen display experiences, broadcast-grade streaming, and many more. The GPU provides a strong multi-display with support for 8K monitors and allows the use of multi-stream video applications for broadcast with its video encoder (NVENC) and decoder engines (NVDEC). Its Quadro Sync technology enables the projectors to create large-scale visualizations, along with providing an immersive virtual reality (VR) and augmented reality (AR) experience with peak performance and increased display bandwidth.

The next generation updates with NVIDIA virtual GPU (vGPU) software make room for bigger and stronger virtual workstation instances through its 48 GB GDDR6 GPU Memory Bandwidth. Its Second-Generation RT Cores and Third-Generation Tensor Cores deliver massive speeds for concurrent demanding workloads and increased throughput for inferencing, combined with strong AI and data science.

Technical Specifications: NVIDIA A40 GPU

GPU Memory48 GB GDDR6 with error-correcting code (ECC)
GPU Memory Bandwidth696 GB/s
InterconnectNVIDIA NVLink 112.5 GB/s (bidirectional) PCIE Gen4 x16 31.5 GB/s (bidirectional)
NVLink2-way low profile (2-slot)
Display Ports3x DisplayPort 1.4*
Max Power Consumption300 W
Form Factor4.4" (H) x 10.5" (L) Dual Slot
vGPU Software SupportNVIDIA vPC/vApps, NVIDIA RTX Virtual Workstation, NVIDIA Virtual Compute Server