New GPU servers with professional-grade NVIDIA RTX A5000 / A4000 cards

NVIDIA RTX A4000/A5000 graphics cards are the nearest thing to the RTX 3080/RTX 3080 cards, but they have double the memory and a similar performance when deployed in servers built with RTX3080/3090 cards.

Try Custom GPU Server with A5000 and A4000 cards — closest relative of the RTX 3080 / RTX 3080!

Be at the cutting edge of innovation with your NVIDIA RTX™ A5000 and A4000 powered GPU server, providing you with the right balance between power, performance and reliability when tackling heavy workflows. These servers are built on last gen NVIDIA Ampere architecture including 25 GB of GPU memory on the RTX A5000 and 16 GB on the RTX A4000 cards, delivering the performance needed by engineers, designers and artists to bring their visions of the future to reality today.

Features of GPU servers equipped with A5000 and A4000 video cards

NVIDIA Ampere Architecture-Based CUDA Cores

Accelerate graphics workflows with the latest CUDA® cores for up to 2.5X single-precision floating-point (FP32) performance compared to the previous generation.

Second-Generation RT Cores

Produce more visually accurate renders faster with hardware-accelerated motion blur and up to 2X faster ray-tracing performance than the previous generation.

Third-Generation Tensor Cores

Boost AI and data science model training performance up to 10X faster compared to the previous generation with hardware support for structural sparsity.

24GB / 16GB of GPU Memory

Tackle memory-intensive workloads, from virtual production to engineering simulations, with 24 GB of GDDR6 memory with ECC.

Virtualization Ready

Repurpose your personal workstation into multiple high-performance virtual workstations with support for NVIDIA RTX Virtual Workstation (vWS) software.

Third-Generation NVIDIA NVLink

Scale memory and performance across multiple GPUs with NVIDIA® NVLink™ to tackle larger datasets, models, and scenes.

PCI Express Gen 4

Improve data-transfer speeds from CPU memory for data-intensive tasks with support for PCI Express Gen 4..

We use the NVIDIA NVLink Bridge for multi-GPU scalability

Easily scalable professional applications, memory and performance with multi-GPU server builds employing NVLink technology. NVIDIA NVLink Bridges allow you to connect two RTX A5000 or A4000 cards through an innovative low-profile design that fits seamlessly in any number of systems with an impressive 112 GB/s of bandwidth and a combined 48GB of the latest GDDR6 memory allowing you to tackle even the most memory-intensive workloads.

Видео выше

How fast are the RTX A5000 and RTX A4000?

HOSTKEY engineers compared the performance and cost of GPU servers built with the RTX A5000/A4000, RTX 3090 and GTX 1080Ti / 1080 cards using a machine learning problem as a trial. This study will help you evaluate the capabilities of the cards and their cost-effectiveness in solving a specific task related to the operation of a neural network.

The work of neural networks on servers built with RTX A5000 and RTX A4000 cards is comparable to the speed on servers with RTX 3090/3080 cards. And the cost of training models on servers with RTX A5000 and RTX A4000 cards is up to 12 times cheaper than on cloud GPU platforms such as AWS, Google Cloud and Microsoft Azure.

Learn more about our comparative testing

What server configurations will we provide?

With the help of a convenient online configuration wizard, you can order a custom server. Just select the number of cards, processor, RAM, storage volume, operating system and network settings and it will be delivered within 1 business day. Furthermore, we have already prepared servers built with the most popular configurations which can be delivered within a few hours from the moment of receipt of payment.

Ready to order?
Get a 12% discount!

Order now and save! Increasing discounts when paying for servers for terms of 3, 6 or 12 month – 3%, 5% и 12% respectively.

Due to the high demand for new servers and the shortage of new cards on the market, our first shipment of servers has sold out quickly. And the current shipment is almost gone. Act now and get yours today!

HOSTKEY is the leading provider of GPU solutions in the Netherlands

HOSTKEY provides a wide variety of GPU dedicated servers.

According to research conducted by Data Science experts at Catalys Projects, the cost of training a model on a HOSTKEY server with its "in−house" default cards is almost an order of magnitude lower than with Google or Amazon, though it takes slightly more time.

HOSTKEY also supports the data science community around the world and has recently launched grant programs in support of scientific projects and promising startups, as well as the winners of competitions in the field of Data Science such as Kaggle, Data Driven, IDAO, ZINDI, Signature, and more.