GPU Plans

Amazon EC2 P6 instances, powered by NVIDIA Blackwell GPUs, are designed for high-performance AI training and inference. Featuring 5th Generation Intel Xeon Scalable processors, these instances provide substantial leaps in GPU memory and compute throughput compared to previous generations. They are optimized for large-scale distributed AI workloads, including mixture of experts and trillion-parameter models, utilizing Elastic Fabric Adapter (EFA) networking and high-bandwidth local NVMe storage to accelerate data-intensive machine learning and high-performance computing applications efficiently in the cloud.

G5 instances feature up to 8 NVIDIA A10G Tensor Core GPUs and second generation AMD EPYC processors. They also support up to 192 vCPUs, up to 100 Gbps of network bandwidth, and up to 7.6 TB of local NVMe SSD storage.

Amazon Elastic Compute Cloud (Amazon EC2) P4d instances deliver high performance for machine learning (ML) training and high performance computing (HPC) applications in the cloud. P4d instances are powered by NVIDIA A100 Tensor Core GPUs and deliver industry-leading high throughput and low-latency networking. These instances support 400 Gbps instance networking. P4d instances provide up to 60% lower cost to train ML models, including an average of 2.5x better performance for deep learning models compared to previous-generation P3 and P3dn instances.

G6 instances feature up to 8 NVIDIA L4 Tensor Core GPUs with 24 GB of memory per GPU and third generation AMD EPYC processors. They also support up to 192 vCPUs, up to 100 Gbps of network bandwidth, and up to 7.52 TB of local NVMe SSD storage.

P5 instances provide up to 8 NVIDIA H100 GPUs with a total of up to 640 GB HBM3 GPU memory per instance. P5e and P5en instances provide up to 8 NVIDIA H200 GPUs with a total of up to 1128 GB HBM3e GPU memory per instance. Both instances support up to 900 GB/s of NVSwitch GPU interconnect (total of 3.6 TB/s bisectional bandwidth in each instance), so each GPU can communicate with every other GPU in the same instance with single-hop latency.

Amazon EC2 G4 instances are the industry’s most cost-effective and versatile GPU instances for deploying machine learning models such as image classification, object detection, and speech recognition, and for graphics-intensive applications such as remote graphics workstations, game streaming, and graphics rendering. G4 instances are available with a choice of NVIDIA GPUs (G4dn) or AMD GPUs (G4ad).

Dedicated servers with Xeon CPUs and various GPU types. One month billing minimum.

Reliably run training and inference on AI/ML models, process large data sets and complex neural networks for deep learning use cases, and serve additional use cases like high-performance computing (HPC)

A3 VMs are preconfigured in one shape with eight of NVIDIA H100 GPUs

G2 VMs come with NVIDIA L4 GPUs.

Virtualized GPU technology for sharing graphics processing & computing power.

Balanced price-performance for machine learning, data analytics, and gaming harnessing the power of CUDA, Tensor, and RT cores. Currently available in limited core compute regions. Contact our sales team to get started.

Dedicated virtual machines equipped with NVIDIA GPUs to speed up complex compute jobs.

The ND H200 v5-series is a flagship Azure virtual machine designed for high-end AI training and generative inference. It features eight NVIDIA H200 Tensor Core GPUs with 141GB of HBM3e memory each, providing 4.8 TB/s of bandwidth. Powered by Sapphire Rapids processors, it offers 96 vCPUs and 1850 GiB of RAM. With 3.2 Tb/s of InfiniBand interconnect, it is optimized for large-scale, low-latency AI clusters and complex scientific computing workloads.

NCv3-series VMs are powered by NVIDIA Tesla V100 GPUs. From 6 to 24 Intel Xeon E5-2690 v4 vCPUs.

The NCasT4_v3-series virtual machines are powered by Nvidia Tesla T4 GPUs and AMD EPYC 7V12(Rome) CPUs. The VMs feature up to 4 NVIDIA T4 GPUs with 16 GB of memory each, up to 64 non-multithreaded AMD EPYC 7V12 (Rome) processor cores(base frequency of 2.45 GHz, all-cores peak frequency of 3.1 GHz and single-core peak frequency of 3.3 GHz) and 440 GiB of system memory.

The NVv3-series virtual machines are powered by NVIDIA Tesla M60 GPUs and NVIDIA GRID technology with Intel E5-2690 v4 (Broadwell) CPUs and Intel Hyper-Threading Technology. These virtual machines are targeted for GPU accelerated graphics applications and virtual desktops where customers want to visualize their data, simulate results to view, work on CAD, or render and stream content

The NCads H100 v5-series virtual machines are part of the Azure GPU family designed for Applied AI training and batch inference workloads. Powered by NVIDIA H100 NVL GPUs and 4th-generation AMD EPYC Genoa processors, these instances provide high-performance compute capabilities. They feature up to 2 GPUs with 94GB memory each and up to 96 non-multithreaded processor cores. This series is ideal for GPU-accelerated analytics, video processing, and autonomy model training, supporting high-throughput local NVMe storage and accelerated networking.

The ND H100 v5-series virtual machines are Azure's flagship generative AI infrastructure, designed for massive-scale AI training and inference. These instances feature eight NVIDIA H100 Tensor Core GPUs interconnected via 400 Gb/s NVIDIA Quantum-2 InfiniBand. They are powered by 4th Gen Intel Xeon Scalable processors and provide significant performance leaps for large language models. With high-speed DDR5 memory and local NVMe storage, they deliver the throughput necessary for the most demanding deep learning workloads and high-performance computing applications.

Cloud servers specially designed for processing AI, graphics and massively parallel tasks

The most powerful Public Cloud instances for parallel processing: up to 1,000 times faster than a CPU.

The A100 is based on NVIDIA’s Ampere architecture, which brings significant performance improvements over previous generations. It features advanced Tensor Cores that accelerate deep learning computations, enabling faster training and inference times.

Premium AMD CPUs and up to 100k IOPS with MaxIOPS.

Breakthrough multi-workload acceleration for large language model (LLM) inference and training, graphics, and video applications based on the latest Ada Lovelace architecture. The prices shown in the table below reflect on-demand pricing. 36-month prepaid pricing for the NVIDIA L40S starts at $0.848/GPU/hr.

Designed specifically for AI and HPC workloads, featuring fourth-generation Tensor Cores and the Transformer Engine with FP8 precision for accelerated performance. Hourly price is calculated based on 730 hours per month. Contact us for additional contract duration and payment options.

Combining professional graphics with powerful compute and AI, to meet today's design, creative, and scientific challenges.

Providing remarkable acceleration with breakthrough memory capacity and memory bandwidth. The prices shown in the table below reflect on-demand preemptible instances for AMD MI355X GPUs.