Expanding Access to Full-Performance HPC Through the Cloud
In 2009, Penguin Computing engineers saw first-hand that the dramatic proliferation of data and corresponding increase in demand for high-performance computing (HPC) required more ways to access compute power. From this customer need, one of the first real HPC-as-a-service options, PODTM (Penguin Computing On-Demand), was born.
Built on a secure, high-performance bare metal server platform with supercomputing-grade, non-blocking InfiniBand interconnects infrastructure, POD can handle the most challenging simulation and analytics. But, because of access via the cloud (from either a traditional Linux command line interface (CLI) or a secure web portal) you get both instant accesses and extreme scalability—without having to invest in on-premise infrastructure or the associated operational costs.
With included free support from HPC experts, transparent pricing, and a variety of financing options, POD meets the needs of both power users who need to expand capacity quickly as well as those building their organizations on a budget.
Choose from public cloud, private cloud, or a hybrid cloud solution. Confidently run manufacturing, earth sciences, bio-informatics, energy, design, finance or other applications, which you’ll likely find pre-installed (if not, just ask us to load them).
WEBINAR: Cloud Computing with HELYX on POD
Penguin Computing and ENGYS have partnered to bring HELYX open-source CFD solutions to the POD HPC Cloud. Join this webinar to learn how you can mesh, setup, solve, and post-process with HELYX on 3D remote workstations, then submit directly to a non-virtualized, bare metal HPC cluster.
Webinar March 26th, 10 AM CDT
There are hundreds of applications pre-installed on POD
Get Full Service HPC Cloud for a Clear and Predictable Low Price
POD offers a transparent pricing model, so you know your actual costs and can predict future costs.
Determine Your Costs
Use the Cost Calculator to understand your potential POD costs. Simply select a compute queue and its corresponding storage. Each account comes with:
- A free login node
- 2D Workstation
- Larger login nodes and 3D Workstations are also available for power users
- Industry-leading support from our team of HPC experts is included in the price
For more detailed pricing, view the detailed price list. Support from HPC experts who can assist you in running applications, and managing workflows is included free of charge.
Pay Only for What You Use
POD usage is metered in three second increments with:
- No network charges
- No data download fees
- No minimum charges
- No fees to open or maintain an account
- No charges for idle time
Reporting & Billing
The POD reporting dashboard gives you a timely, real-time billing that accurately reflects usage. It shows both live reporting of your actual usage and charges as well as billing history. The POD reporting dashboard also granularity so that you can get a snapshot of billing status, compartmentalized by users, groups or projects.
All of these details are available by downloading a spreadsheet, or through the POD API.
Cluster Design Options
Select the HPC configuration that best meets your computing needs:
|Compute Queue Name||Compute Nodes||Cores/Node||Ram/Node||High Speed Interconnect||Storage|
|S30||Dual Intel® Xeon® Gold 6148 (Skylake)||40||384 GB||Intel Omni-Path||Lustre File System|
|B30||Dual Intel® E5-2600v4 Series (Broadwell)||28||256 GB||Intel Omni-Path||Lustre File System|
|T30||Dual Intel® E5-2600v3 Series (Haswell)||20||128 GB||QDR InfiniBand||Private NAS Volume|
|M40||Dual Intel® X5600 Series (Westmere)||12||48 GB||QDR InfiniBand||Private NAS Volume|
|H30||Dual Intel® E5-2600 Series (Sandy Bridge)||16||64 GB||QDR InfiniBand||Private NAS Volume|
|Compute Queue Name||GPU Nodes||Cores GPUs/Node||Ram/Node||High Speed Interconnect||Storage|
|H30G||Dual NVIDIA® Tesla® K40 GPU||2 GPUs/Node||64 GB||QDR InfiniBand||Private NAS Volume|
Set up Your Workspace, Your Way
Register for a POD account
- Choose your storage and login node preferences (free login nodes are available)
- Launch your login node and get secure access to the compute cluster
Choose Among User Interface Options
- Power users can opt to use the traditional Linux command-line interface environment
- Others may want to leverage POD’s secure Web GUI and API web services
- Consider our browser-based remote visualization desktop for graphically interacting data or running GUI-based applications
Run Virtually Any HPC Application
- Load any of our vast set of pre-installed applications that are ready to use
- POD supports applications with ready-to-run cloud licenses (i.e., LSTC LS-DYNA, CD-Adapco StarCCM+, or Mathworks MATLAB). However, POD also supports secure VPN tunnels, which allow you to leverage your existing, in-house licenses directly on POD
- If the application or version you need isn’t already installed (see selected list), Penguin’s support team will install it for you
Move Data In/Out Securely & Quickly
- Take advantage of POD’s redundant, high-speed data links and begin quickly uploading and downloading data to POD without any fees for network transfers
- For your security, POD only supports secure protocol transfers in and out of POD. This includes secure access through our web interface, APIs, remote GUI tools such as WinSCP, and standard CLI transfer tools such as SCP and rsync via SSH.
Leverage Integrated Pre- and Post-Processing
- POD is designed to handle your complete HPC workflow – pre-processing, computing, and post-processing. Accelerate your time to results by leveraging POD’s remote visualization, 3D desktop interface allowing you to visualize your cloud data directly through a web browser.
Difference Between Traditional Bare Metal and Virtualized HPC Clusters
When considering whether to choose a bare metal or virtualized server in the cloud for Linux high-performance computing (HPC), consider which criteria are most important to you in a computing platform.
Bare metal HPC clusters excel at:
- Providing the highest performance available from the hardware
- Predictable high-speed low latency interconnections
- Flexibility in security/regulatory compliance
- True high availability
- Low latency
- Faster ‘time-to-computing’
- Scalability for tightly coupled HPC workloads
- Dedicated resources
- Ability to customize the platform
Cloud-based virtualized servers are typically chosen for:
- Scalability for embarrassingly parallel workloads
- Organizational agility
- Flexibility to re-provision
- A sandbox to run new projects
- Pricing structure reflecting a shared resource
- Contingencies against power outages, equipment failure, or other disruptions
Bare metal HPC cloud provides:
- All the benefits of bare metal as well as the benefits of a cloud-based solutions that are typically associated with virtualized servers
Feel Confident with a Secure Environment
To ensure you have the most secure environment possible to protect your data, secure your identity, and restrict access to only authorized users, Penguin:
- Uses a Single-Tenant Compute Environment – Unlike multi-tenant cloud environments, POD’s compute environment is segregated to protect your privacy.
- Layers on Physical Security – POD is hosted at a highly secure, Tier III facility that undergoes an annual SSAE18 SOC 2 Type II audit.
- Ensures Data Security – All customer data lives solely at our United States, Tier III datacenter facility. Customers have complete control over access and persistence of their cloud data.
- Encrypted Connections – All access to POD is over secure, industry standard, encrypted protocols: HTTPS+SSL, SSH through AES-128 or AES-256, or customized private VPN connections through AES-256.
POD Support and Services
POD provides full-service support to help organizations optimize their high-performance computing environments. Our highly-technical and highly-skilled experts engage with customers on the ground in a quick and efficient manner—working with them to fine tune workflows, integrate HPC applications and edit network policies and procedures.
Our highly-skilled experts have many years combined HPC expertise and have backgrounds in various science disciplines (computational physicals and weather forecasting are just two examples). More than just tech experts, this domain experience helps us to better engage with and support the unique needs of our research customers. Specifically, we can help you get started on research, optimize workflow and get the most out of your HPC environment.
Professional Services That Go Above and Beyond
While our basic support package is more robust and inclusive than traditional service offerings, POD also understands that additional services may be required for highly customized projects. These service offerings include custom and branded portal interfaces, more in-depth app support and integration with on-premise clusters or hybrid cloud environments. No matter what is next for your organization, we can help get you there.
“It has been a pleasure using Penguin’s POD service. Penguin is more like a collaborator ensuring my applications run and run efficiently. It’s also trivial to scale up and offers significant savings compared to the typical cloud providers.”
“After evaluating multiple cluster management products, we chose Scyld ClusterWare for the processing of large amounts of image data. Scyld ClusterWare contains a robust, easy-to-use and unique Single System Image architecture, which we believe will resonate well with researchers using the SOLiD platform for a wide variety of applications.”
“We replaced our legacy IBM system with a TOP100 cluster from Penguin in a ‘rolling’ update without significant downtime. A daunting task that Penguin was very successful at completing. We are very happy.”
"The price (of POD) was every bit as advertised and that undersold the other service we were using by 50%. That is NOT a small number. The servers proved more powerful than our other out-sourced analysis company, and (POD) was overall easier to use."
"We are pleased to work with Penguin Computing to enable users to run their applications faster and solve larger, more intricate problems without increasing the IT cost associated with maintaining high-performance capacity."
"We appreciate the consistent quality of the cores that are being made available to us by Penguin Computing, enabling us flexibility in terms of volume load. The interactive framework is convenient to implement and allows us to use POD not only for solving, but also for pre- and post-processing."
"I am really glad we made the choice to go with your service. Your tech support guys have been really amazing and I know we would have had a lot more headaches with any other services we were considering."
"Penguin's HPC expertise and knowledge of LS-DYNA is outstanding. POD is an easy and cost effective cloud bursting solution."
"Penguin has gone out of its way many times to accommodate our requests for technical help and to meet our fast-changing needs."
100% on-time project completion by solving models during peak seasons on Penguin Computing On-Demand HPC cloud platform