GPU procurement complexity has been increasing with more providers offering GPU cloud options. AIMultiple analyzed GPU cloud provider across most relevant dimensions to facilitate cloud GPU procurement.
While listing pros and cons for each provider, we relied on user reviews on G2, other online reviews as well as our assessment.
Amazon Web Services (AWS)
AWS is the largest cloud platform provider and a leading cloud GPU provider.1Amazon EC2 (Elastic Compute Cloud) offers GPU-powered virtual machine instances facilitating accelerated computations for deep learning tasks.
Pros
Offers seamless integration with other popular AWS solutions like:
- SageMaker, used for creating, training, deploying, and large-scale application of ML models
- Simple Storage Service (Amazon S3), Amazon RDS (Relational Database Services) or other AWS storage services, which can serve as a storage solution for training data
Cons
- AWS offers fewer GPU options than some other players like Azure.
- UI is found to be complex by users
- On-demand pricing per hour is higher than other big cloud providers. Like other cloud providers, AWS offers volume discounts.
Microsoft Azure
Microsoft Azure, the second largest cloud provider, provides a cloud-based GPU service known as Azure N-Series Virtual Machines, which leverages NVIDIA GPUs like other providers to deliver high-performance computing capabilities.4 This service is particularly suited for demanding applications such as deep learning, simulations, rendering and the training of AI models.
Pros
- Microsoft Azure is offering a larger set of GPU options than most other providers
- Free plan offers 12 months of access to some services
- Azure’s intuitive user interface is praised for its ease of use
Cons
- Some users find that certain advanced features within Azure require a high level of technical expertise to configure and manage effectively
- Some users find Azure’s pricing structure complex to navigate and stress the importance of careful planning to avoid unexpected costs
Google Cloud Platform (GCP)
Google Cloud Platform (GCP) is the third biggest cloud platform.6 GCP offers GPU instances that can be attached to existing virtual machines (VMs) or can be part of a new VM setup.
Pros
- UI is easier than other common platforms such as AWS
- Offers limited free GPU options for Kaggle and Colab users
- Customers can use 20+ products for free, up to monthly usage limits
Cons
- GPUs must be attached to standard VMs, making pricing confusing
- Like AWS, GCP offers fewer GPU options than some players like Azure
NVIDIA DGX Cloud
NVIDIA is the leader in the GPU hardware market. NVIDIA launched its GPU cloud offering, DGX Cloud, by leasing space in leading cloud providers’ (e.g. OCI, Azure and GCP) data centers.
DGX cloud offers NVIDIA Base Command™, NVIDIA AI Enterprise and NVIDIA networking platforms. DGX Cloud instances featured 8 NVIDIA H100 or A100 80GB Tensor Core GPUs at launch.
An initial customer’s, Amgen’s, research team claims 3x faster training of protein LLMs with BioNeMo and up to 100x faster post-training analysis with NVIDIA RAPIDS.8
Oracle Cloud Infrastructure (OCI)
Oracle ramped up its GPU offering after formalizing its partnership with NVIDIA.9
Oracle provides GPU instances in both bare-metal and virtual machine formats for quick, cost-effective, and high-efficiency computing. Oracle’s Bare-Metal instances offer customers the capability to execute tasks in non-virtualized settings. These instances are accessible in regions such as the United States, Germany, and the United Kingdom, with availability under both on-demand and interruptible pricing models.
Pros
- Wide range of cloud products and services. Among the tech giants’ cloud services, only OCI offers bare metal GPUs.10 For GPU cluster users, only OCI offers RoCE v2 for its cluster technology among the tech giants’ cloud services.11
- Cost-effective compared to other major cloud providers
- Offers provision for free trial period and some free-forever products
Cons
- User interface perceived as clunky and slow by users
- Some users find the documentation difficult to understand
- The process of starting to use Oracle Cloud compute services was viewed as bureaucratic, complicated, and time-consuming by some users
CoreWeave
CoreWeave is a specialized GPU cloud provider. NVIDIA is one of CoreWeave’s investors. CoreWeave claims to have 45,000 GPUs and to be selected as the first first Elite level cloud services provider by NVIDIA.14
Jarvis Labs
Jarvis Labs, established in 2019 and based in India, specializes in facilitating swift and straightforward training of deep learning models on GPU compute instances. With its data centers located in India, Jarvis Labs is recognized for its user-friendly setup that enables users to start operations promptly.
Pros
- No credit card required to register
- A simple interface for beginners
Cons
- Although gaining momentum, Jarvis Labs is not a good option for enterprise-level and time-consuming tasks
Lambda Labs
Originally, Lambda Labs was a hardware company offering GPU desktop assembly and server hardware solutions. Since 2018, Lambda Labs offer Lambda Cloud as a GPU platform. The virtual machines they offer are pre-equipped with predominant deep learning frameworks, CUDA drivers, and a dedicated Jupyter notebook. Users can connect to these instances through the web terminal in the cloud dashboard or directly using the given SSH keys.
Pros
- Purely GPU focused offering
Paperspace CORE
Paperspace is a cloud computing platform that offers GPU-accelerated virtual machines, among other services. The company is well-regarded for its focus on GPU-intensive workloads and provides a cloud platform for developing, training, and deploying machine learning models.
Pros
- Offers a wide range of GPUs compared to other providers
- Users find the prices fair for the computing power provided
- Users find the customer service to be friendly and responsive
Cons
- Some users complain about machine availability, both in terms of the free virtual machines and specific machine types not being available in all regions
- The integrated Jupyter interface is criticized and lacks some keyboard shortcuts, although a native Jupyter Notebook interface is offered
- Longer loading or creation times for machines
- Monthly subscription fee on top of machine costs can be a downside, and multi-GPU training can be expensive
External links
- Big Three Dominate the Global Cloud Market, Statista, Retrieved July 19, 2023
- https://www.g2.com/products/amazon-ec2/reviews/amazon-ec2-review-8154729
- https://www.g2.com/products/aws-cloud/reviews/aws-cloud-review-8271023
- Same Statista source as above
- https://www.g2.com/products/azure-virtual-machines/reviews/azure-virtual-machines-review-8145738
- Same Statista source as above
- “NVIDIA Launches DGX Cloud, Giving Every Enterprise Instant Access to AI Supercomputer From a Browser“. NVIDIA. March 21, 2023. Retrieved September 26, 2023.
The offering is enterprise focused with the list price of DGX Cloud instances starting at $36,999 per instance per month at launch.
Pros
- Support from NVIDIA engineers
Cons
- Offering is not suitable for firms with limited GPU needs
- The service is provided on top of cloud providers’ physical infrastructure. Therefore buyer needs to pay for the margins of both the cloud provider and NVIDIA.
IBM Cloud
The GPU offered by IBM Cloud allows for a flexible process of selecting servers, and it has a seamless integration with the architecture, applications, and APIs of IBM Cloud. This is accomplished via a globally distributed network of data centers that are interconnected.
Pros
- Powerful integration with IBM Cloud architecture and applications
- Worldwide distributed data centers increases data protection
Cons
This post originally appeared on TechToday.