Digital Ocean Additional GPU - A100, $ 2,232.80/mo. on Linux Cloud


Additional GPU - A100 has been added on Oct 14, 2026, Aggregate Rating (2 out of 10 from 2 reviews)
digitalocean.com
Digital Ocean
105 Edgeview Drive, Suite 425
Broomfield , CO 80021
US
☎ Phone +1.6463978051
☎ Phone +1 646 827 4366 [Investor relations]
  • 💡 Plan Name: Additional GPU - A100 (sku #p159204)
  • 🔧 Category: Cloud / Linux
  • 💰 Price:$ 2,232.80/mo.
  • 💿 Disk Space: 50 GB
  • 📶 Traffic bandwidth: unmetered
  • 💲 Setup Fee: free
💪 CPU/Cores :12 vCPUs
🔋 RAM :90000 MB
🆓 free domains :0
📌 Dedicated IPs :1
💳 Payment Methods :Credit / Debit / Prepaid CardsPayPalAliPayBitcoin
🔨 Control Panel :[In-house]
✍️ Support Options :EmailHelp Desk
🌏 Server Locations :United States
⚑ Targeting :AU FR NL SG US
screenshot of Additional GPU - A100 from digitalocean.com

See also initial Additional GPU - A100 plan location on their website!

*📜 Plan description

Additional GPU - A100 includes NVIDIA A100, 90 GiB host RAM, 12 vCPUs, a 50 GB persistent disk and a public IPv4 address. GPU compute costs $3.09/hour, or $2224.80 for 720 powered-on hours. Adding the $5/month disk cap and $3/month IPv4 cap gives $2232.80 for this complete SSH-accessible configuration.

The selected configuration uses a Linux machine with the minimum listed 50 GB persistent block disk and one public IPv4 address for SSH access. The disk costs $0.0074/hour up to $5/month; public IPv4 costs $0.0045/hour up to $3/month. The monthly amount combines 720 hours of powered-on GPU compute with these two monthly resource caps. Compute stops being charged when the machine is powered off; retained storage and IP resources keep accruing charges until removed. Ingress and egress bandwidth are free and unmetered, but sustained transfer of at least 1 Gbps for more than ten seconds is rate-limited to 200 Mbps unless an increase is granted. This is a Paperspace machine, not a GPU Droplet, so GPU Droplet scratch storage and SLA values are not assumed. Windows templates are unavailable to new users; eligible accounts created before July 2024 can still use their legacy Windows templates.

📄 Editorial Review

DigitalOcean is an American cloud infrastructure and platform services provider founded in 2012, specializing in virtual private servers, cloud computing, managed hosting, container orchestration, databases, storage, networking, and artificial intelligence infrastructure. The company serves developers, startups, software companies, and small to medium-sized businesses seeking scalable computing resources without the operational complexity commonly associated with larger enterprise cloud platforms.

The company was founded by Ben Uretsky, Moisey Uretsky, Mitch Wainer, Alec Hartman, and Jeff Carr. Its early development was supported by the Techstars accelerator program, and its SSD-powered virtual machines gained considerable popularity in 2013. DigitalOcean initially established its reputation through straightforward Linux VPS hosting, affordable cloud servers, and a developer-friendly management interface.

DigitalOcean has since developed into a publicly traded technology company. Its parent organization, DigitalOcean Holdings, Inc., trades on the New York Stock Exchange under the symbol DOCN, following its initial public offering in March 2021. The corporate headquarters is in Broomfield, Colorado, having previously been located in New York City. The company serves more than 640,000 customers across approximately 190 countries.

Its expansion has included several significant acquisitions. In 2021, DigitalOcean acquired Nimbella to strengthen its serverless computing capabilities. In 2022, it acquired Cloudways, extending its reach into managed website hosting and agency-oriented services. The 2023 acquisition of Paperspace introduced GPU computing and machine learning infrastructure. In April 2026, DigitalOcean acquired Katanemo Labs, incorporating agentic AI infrastructure technology and the Plano open-source project into its growing inference platform.

DigitalOcean retains a strong focus on conventional cloud hosting while expanding into AI-native infrastructure. Its portfolio combines Infrastructure as a Service (IaaS), Platform as a Service (PaaS), managed data systems, and GPU-powered AI solutions.

Droplets – Cloud VPS Hosting remains the company's flagship product. These Linux-based virtual machines use KVM virtualization and SSD or NVMe storage, depending on the configuration. Customers can deploy instances within minutes, selecting the operating system, hardware configuration, geographical location, and optional software images.

Basic Droplets use shared CPU resources and are suitable for smaller websites, development environments, lightweight APIs, testing, and applications with fluctuating processor requirements. Dedicated CPU configurations offer more consistent computing performance for applications requiring predictable processing capacity.

The principal Droplet families include Basic, General Purpose, CPU-Optimized, Memory-Optimized, and Storage-Optimized configurations. Each family targets particular workloads, ranging from modest websites and application servers to demanding databases and storage-intensive business applications.

General Purpose instances balance dedicated CPU resources with memory, while CPU-Optimized servers provide additional processing capacity for media encoding, analytics, and batch operations. Memory-Optimized Droplets are useful for databases, caching systems, and applications requiring substantial RAM. Storage-Optimized instances incorporate high-performance NVMe storage for applications requiring considerable local capacity and disk throughput.

A significant addition in August 2026 was the introduction of v5 Droplets, powered by fifth-generation AMD EPYC processors. These instances allow customers to configure CPU, memory, and storage resources more independently rather than relying exclusively on predetermined hardware combinations. This provides greater flexibility for specialized applications, data processing, and demanding AI-related workloads.

Premium processor configurations are also available, combining modern Intel or AMD hardware with NVMe storage and higher network throughput. Selected Premium dedicated-CPU configurations support up to 10 Gbps outbound connectivity.

Private Droplets provide another deployment option for workloads that do not require a public IP address. Applications can operate within private network environments, reducing unnecessary exposure to the public internet.

Droplets remain largely self-managed. Customers are responsible for operating system maintenance, application deployment, security updates, web server configuration, and administrative tasks. This makes the service particularly suitable for developers and system administrators with Linux experience.

App Platform – Managed Application Hosting provides a simplified deployment environment for websites, APIs, background workers, and applications built using popular programming languages and frameworks.

Applications can be deployed directly from GitHub, GitLab, or container registries. The platform manages infrastructure provisioning, application builds, runtime configuration, HTTPS certificates, deployments, and scaling.

A free tier supports up to three static-site applications with global CDN delivery, automatic HTTPS, custom domains, and DDoS mitigation. Paid application containers provide additional resources, managed runtime features, deployment revisions, logs, and monitoring.

Depending on the instance class, customers can use shared or dedicated CPU resources, vertical scaling, horizontal scaling, and automated scaling. App Platform is particularly suitable for developers who prefer concentrating on application code rather than maintaining a traditional Linux server.

DigitalOcean Kubernetes (DOKS) delivers managed Kubernetes clusters for deploying containerized applications and microservices. The service includes a managed control plane, worker node pools, automated upgrades, autoscaling, integrated networking, and persistent storage options.

Customers can combine shared or dedicated CPU worker nodes with GPU resources where regional availability permits. Advanced capabilities include cluster autoscaling, high availability, and scaling eligible node pools down to zero during idle periods.

The standard control plane is included without an additional charge, while worker nodes are billed according to their underlying computing resources. A paid high-availability control plane option provides an additional reliability commitment.

DOKS is well suited to development teams operating container-based production applications that require native Kubernetes compatibility without directly administering all control plane components.

Managed Databases provide fully managed hosting for PostgreSQL, MySQL, MongoDB, Apache Kafka, Valkey, and OpenSearch.

These services reduce the administrative work involved in provisioning, patching, upgrading, monitoring, backing up, and maintaining database clusters. Features include automated daily backups, point-in-time recovery options, metrics, private networking, encryption, configurable storage, and automated failover with suitable cluster configurations.

Database resources can be expanded by adding processing power, memory, storage, or additional nodes. Storage autoscaling further reduces the likelihood of applications unexpectedly exhausting their database capacity.

The newer PostgreSQL Advanced Edition provides additional performance, administration, and observability capabilities for demanding production environments. Managed databases are particularly useful for companies requiring relational databases, document stores, caching systems, search infrastructure, or event-stream processing without maintaining dedicated database administrators for each component.

Spaces Object Storage offers S3-compatible storage for files, website assets, application-generated content, backups, images, videos, and large datasets. It includes integrated CDN functionality and can be managed through compatible S3 clients and APIs.

Spaces also includes a lower-cost cold-storage option for infrequently accessed data. This is useful for archival workloads, historical backups, and large datasets requiring occasional retrieval rather than continuous access.

Volumes Block Storage provides expandable network-attached block devices for Droplets. These volumes offer persistent storage that can be resized and attached to compatible computing resources.

Network File Storage delivers shared, POSIX-compatible NFS storage for multiple servers or containers. Standard and high-performance tiers accommodate conventional shared-storage requirements as well as more demanding AI workloads.

Automated Droplet Backups, snapshots, and custom images provide additional options for data protection, recovery, and migration. Backup frequencies include weekly, daily, and more frequent schedules, with configurable retention periods on applicable plans.

A private Container Registry is also available for storing and distributing container images, with a limited free tier and additional paid capacity.

DigitalOcean's Networking and Security portfolio includes Virtual Private Clouds, cloud firewalls, reserved IP addresses, IPv6, load balancers, VPC peering, private networking, and hybrid connectivity options.

Regional and global Load Balancers distribute application traffic across multiple instances, improving availability and scalability. Features include HTTPS support, Let's Encrypt integration, and HTTP/3 support.

Virtual Private Clouds isolate application resources from unrelated environments. Customers can also use NAT gateways, reserved public addresses, and private network interconnections to build more sophisticated infrastructure.

Security capabilities include DDoS protection, role-based access control, identity and access management, two-factor authentication, and Cloud Security Posture Management (CSPM). The latter assists with detecting configuration issues and improving visibility into cloud security risks.

DigitalOcean also provides free DNS management for hosted domains. It is not primarily a retail domain registration company; customers generally register domain names through independent registrars and connect them to their DigitalOcean infrastructure.

GPU Hosting and Artificial Intelligence represent one of the company's fastest-growing business areas. DigitalOcean has considerably expanded its GPU Droplets and Bare Metal GPU products to support machine learning, artificial intelligence, high-performance computing, model training, and inference workloads.

Hardware choices include NVIDIA and AMD accelerators, including NVIDIA H100, H200, Blackwell-generation B300 configurations, and AMD Instinct MI300X, MI325X, and MI350X systems. Availability depends on the product, geographical location, deployment model, and capacity.

GPU resources can be purchased through on-demand, reserved contractual, or interruptible Spot arrangements. Bare Metal GPU hosting provides dedicated hardware for applications requiring direct access to powerful computing infrastructure.

The Inference Engine combines managed serverless inference, dedicated inference endpoints, batch processing, model selection, and intelligent inference routing. Developers can access numerous AI models through APIs without maintaining the underlying GPU infrastructure themselves.

Managed Agents provides tools for building and operating production AI agents. Its broader ecosystem includes the DigitalOcean Harness Runtime, Action Gateway, Launchpad, managed knowledge bases, model evaluations, and integrations for retrieval-augmented generation.

The Harness Runtime supports isolated, persistent execution environments for coding agents. Action Gateway enables governed access to external tools and APIs, while the Inference Router can optimize model selection according to cost, latency, and workload requirements.

Some newer AI capabilities remain in public preview, making product maturity, regional availability, and production readiness important considerations for businesses adopting these technologies.

DigitalOcean also serves customers seeking more conventional Managed Web Hosting through its Cloudways subsidiary, particularly for WordPress, PHP applications, ecommerce websites, and digital agencies.

Cloudways adds managed application environments, staging functionality, caching technologies, simplified website administration, and deployment workflows requiring less direct server maintenance than standard Droplets.

DigitalOcean Functions provides serverless code execution for event-driven applications without maintaining persistent application servers. The broader ecosystem also includes a software Marketplace, one-click application images, developer tutorials, APIs, Terraform integrations, and startup support programs.

The company's website and primary technical documentation are in English, with a modern, responsive interface and straightforward navigation between products, documentation, pricing, and account management. Its engineering and product blog remains active, with regular announcements and technical articles, including updates during October 2026.

📉 Similar hosting plans from other companies in the same category & country location ≡

  • Files.com
    files.com logo
    🏆 Alexa Rating0

    • 💡 Plan: Starter
    • 🔧 Category: Cloud
    • 💻 OS Type: Linux/Windows
    • 💰 Price: $ 2,099.00/yr.
    • 💿 Disk Space: 1000 GB
    • 📶 Traffic bandwidth: 1 TB
    • 💲 Setup Fee: free
    • 📅 Added:
    • 🌏 Server in:
    • screenshot of Starter from files.com
    Starter is the entry commercial tier and is only available to new customers. It includes the full platform across the eight major capability areas without the heaviest enterprise features. The [...]
  • Lambda.ai
    lambda.ai logo
    🏆 Alexa Rating0

    • 💡 Plan: On-demand 2x NVIDIA A100 PCIe 40 GB
    • 🔧 Category: Cloud
    • 💻 OS Type: Linux
    • 💰 Price: $ 1,857.60/mo.
    • 💿 Disk Space: 1000 GB
    • 📶 Traffic bandwidth: unmetered
    • 💲 Setup Fee: free
    • 📅 Added:
    • 🌏 Server in:
    • screenshot of On-demand 2x NVIDIA A100 PCIe 40 GB from lambda.ai
    2x A100 PCIe 40 GB with 60 vCPUs, 450 GB RAM and 1 TB SSD. At 1.29 per GPU hour a continuous month is about 1857.60. Flexible two GPU rig for small training runs and accelerated serving.
    All [...]
  • Atlantic.Net
    atlantic.net logo
    🏆 Alexa Rating64,852 ▲
    👤 User Rating 👉 Total Reviews: 18
    🙌 Average Rating: 7.9 / 10
    👍 Good Reviews: 15
    👎 Bad Reviews: 3
    👈 Official Responses: 1

    • 💡 Plan: AH100NVL.240GB
    • 🔧 Category: Cloud
    • 💻 OS Type: Linux/Windows
    • 💰 Price: $ 2,407.80/mo.
    • 💿 Disk Space: 2400 GB
    • 📶 Traffic bandwidth: 15 TB
    • 💲 Setup Fee: free
    • 📅 Added:
    • 🌏 Server in:
    • Aggregate Rating (7.9 out of 10 from 18 reviews)
    • screenshot of AH100NVL.240GB from atlantic.net
    GPU Cloud instance AH100NVL.240GB includes 28 vCPU, 240GB RAM, 1 x H100 NVL, 2400GB SSD, and 15 TB transfer. Term discounts available; price shown reflects 3-Year. Built for demanding AI inference [...]
  • InMotion Hosting
    inmotionhosting.com logo
    🏆 Alexa Rating2,859 ▲
    👤 User Rating 👉 Total Reviews: 49
    🙌 Average Rating: 7 / 10
    👍 Good Reviews: 34
    👎 Bad Reviews: 15
    👈 Official Responses: 6

    • 💡 Plan: InMotion Cloud Medium
    • 🔧 Category: Cloud
    • 💻 OS Type: Linux
    • 💰 Price: $ 2,669.00/mo.
    • 💿 Disk Space: 10500 GB
    • 📶 Traffic bandwidth: unmetered
    • 💲 Setup Fee: free
    • 📅 Added:
    • Aggregate Rating (7 out of 10 from 49 reviews)
    • screenshot of InMotion Cloud Medium from inmotionhosting.com
    InMotion Cloud Medium costs a fixed $2669.00/month for 150 vCPU, 520 GB RAM and 10.5 TB storage. The tier is designed for growing applications and multiple team environments. This is a private-cloud [...]