HPC has become a critical-path option for computational applications, and CIOs need to provide HPC services to their existing users, while planning for usage growth. They want to adopt the latest HPC technologies and cloud architectures to drive efficiency and effectiveness of the resource across the entire user community.
“We help organizations think through sizing for capacity, performance and long-term scalability,” highlights Sue. Whether it is the need to support big data analysis, more sophisticated modeling or other types of workloads, Silicon Mechanics helps customers evaluate the possibilities for their specific HPC architecture. Later, the company will deliver and support a purpose-built, turn-key HPC cluster or hybrid cloud solution, depending on the client’s requirements.
The company’s proprietary methodology offers customers its‘ dynamic assessment tool’, allowing customers to change the environmental variables such as heat, cooling, power, footprint or performance variables like bandwidth, IO, CPU cores to visualize the impacts of their system configurations. Customers can manipulate characteristics of the solution, see the impact on their budget as they choose their priorities.
“Our job as a system integrator is to present all of the pertinent options to customers,” notes Sue. With over 15 years of experience, Silicon Mechanics constantly assesses and tests current and future hardware and software components from a wide network of technology component manufacturers. This combined with its ‘Expert included’ in-house technical expertise allows the company to present customers with a customized series of options, which is targeted at their specific needs and ultimately “we can offer purpose-built clusters for the customers,” adds Sue.
Our expertise is in helping companies take advantage of today’s most innovative HPC architectures and technologies, while giving them the best value for their requirements
Recently, Silicon Mechanics partnered with the Van Andel Institute (VAI), a nonprofit biomedical research and science education organization. VAI had virtually no ability to handle big data, and as a result large-scale scientific computing projects were not processed due to lack of an on-site infrastructure to support them. Silicon Mechanics helped architect a high performance cluster to enable processing via VAI’s existing infrastructure along with cloud capability for bursting compute workloads when necessary. By engaging with Silicon Mechanics, VAI was able to reduce their own project deployment schedule by over a year versus working to build and implement themselves.
Silicon Mechanics ensures each custom high performance cluster is designed, built and supported for both short and long term optimization. By offering their consulting, professional services and methodologies, Silicon Mechanics assists its customers in optimizing hybrid cloud HPC architectures for flexible compute workloads, performance storage, data sharing and multi-level security.
“Customers are very interested in hybrid cloud as a solution for supporting flexible compute capacity and data storage requirements for today’s computational applications across multiple industries. Silicon Mechanics will continue to refresh our reference architectures to meet the evolving and unique needs of high performance workloads including storage and security, within key industries such as federal government, higher education, and research,” concludes Sue. “Our experience and skills with hybrid cloud deployments can really help clients shorten the cost of assessing this approach, and can accelerate their path to a fully-deployed solution.”