Cornell Looks to Cloud for On-Demand Research
The Cornell University Center for Advanced Computing (CAC) needed a cloud computing platform configured to meet the unique needs of research scientists and engineers from Cornell, other academic institutions, and industry. The solution was to build “Red Cloud”— a cloud computing service designed specifically for research that provides Infrastructure-as-a-Service (IaaS) using Dell PowerEdge C* servers equipped with the Intel® Xeon® processor 5600 series and Intel® Ethernet 10 Gigabit Server Adapters. With this new service, CAC is helping researchers in a broad range of fields to access the outstanding compute performance they need, when they need it, without excessive costs.
Challenges
• Provide on-demand HPC resources. Facilitate cutting-edge research, education, and outreach in a variety of fields by offering cost-effective access to on-demand resources.
• Maximize flexibility and scalability. Complement existing traditional HPC clusters to fulfill research computing needs—from temporarily extending HPC clusters and hosting research wikis and science gateways to providing virtual workstations and computer labs. Red Cloud runs on Eucalyptus to ensure compatibility with Amazon Elastic Cloud Compute (EC2)* so that users can scale beyond the capacity of Red Cloud as needed.
Solution
• Red Cloud with Dell servers and Intel® Xeon® processors. The Red Cloud IaaS service uses Dell PowerEdge C servers equipped with the Intel Xeon processor 5600 series and Intel® Ethernet 10 Gigabit Server Adapters.
Technology results
• Outstanding performance. Built on Intel Xeon processor–based servers, Red Cloud helps researchers accelerate results from days to hours compared with existing departmental resources.
Business value
• On-demand flexibility. The IaaS offering provides on-demand resources for short- and long-term needs without requiring research groups to purchase and maintain their own infrastructure. A second offering—Red Cloud with MATLAB—is a Software-as-a-Service (SaaS) targeting MATLAB users.
Seamless scalability. Researchers can easily scale up to Red Cloud from departmental workstations and scale out from Red Cloud to larger commercial cloud environments.
Read the full Cornell Looks to Cloud for On-Demand Research case study: http://www.intel.com/content/www/us/en/high-performance-computing/high-performance-computing-xeon-5600-cornell-study.html
469KB
Veuillez nous excuser, ce PDF peut uniquement être téléchargé
Cornell Looks to Cloud for On-Demand Research
The Cornell University Center for Advanced Computing (CAC) needed a cloud computing platform configured to meet the unique needs of research scientists and engineers from Cornell, other academic institutions, and industry. The solution was to build “Red Cloud”— a cloud computing service designed specifically for research that provides Infrastructure-as-a-Service (IaaS) using Dell PowerEdge C* servers equipped with the Intel® Xeon® processor 5600 series and Intel® Ethernet 10 Gigabit Server Adapters. With this new service, CAC is helping researchers in a broad range of fields to access the outstanding compute performance they need, when they need it, without excessive costs.
Challenges
• Provide on-demand HPC resources. Facilitate cutting-edge research, education, and outreach in a variety of fields by offering cost-effective access to on-demand resources.
• Maximize flexibility and scalability. Complement existing traditional HPC clusters to fulfill research computing needs—from temporarily extending HPC clusters and hosting research wikis and science gateways to providing virtual workstations and computer labs. Red Cloud runs on Eucalyptus to ensure compatibility with Amazon Elastic Cloud Compute (EC2)* so that users can scale beyond the capacity of Red Cloud as needed.
Solution
• Red Cloud with Dell servers and Intel® Xeon® processors. The Red Cloud IaaS service uses Dell PowerEdge C servers equipped with the Intel Xeon processor 5600 series and Intel® Ethernet 10 Gigabit Server Adapters.
Technology results
• Outstanding performance. Built on Intel Xeon processor–based servers, Red Cloud helps researchers accelerate results from days to hours compared with existing departmental resources.
Business value
• On-demand flexibility. The IaaS offering provides on-demand resources for short- and long-term needs without requiring research groups to purchase and maintain their own infrastructure. A second offering—Red Cloud with MATLAB—is a Software-as-a-Service (SaaS) targeting MATLAB users.
Seamless scalability. Researchers can easily scale up to Red Cloud from departmental workstations and scale out from Red Cloud to larger commercial cloud environments.
Read the full Cornell Looks to Cloud for On-Demand Research case study: http://www.intel.com/content/www/us/en/high-performance-computing/high-performance-computing-xeon-5600-cornell-study.html
