MODULAR SUPPORT: GRID STORAGE AND ARCHIVE SOLUTION (GSAS)
Re-Store is IBM’s only specialist partner with an exclusive focus on technical computing.
Re-Store’s Grid Storage and Archive Solution (GSAS) is a comprehensive, modular solution that delivers on the promise of next-generation Software-Defined Infrastructures (SDI).
Software-defined & Modular SAN Solutions
Software-defined techniques transform static data centers into dynamic resources that drive data-aware customer workloads. Re-Store fuses HPC, cloud (private/public/hybrid), virtualization, clustered filesystem, NAS and mass archive techniques to create compute-intensive and data-centric workflow systems.
Re-Store’s Unified Service Model delivers end-to-end customer success:
- DESIGN – Architectural design and pre-sales planning
- BUILD – Multi-Vendor / Multi-discipline implementation
- SUPPORT – Long-term success with context and continuity
Solution Architects with Spectrum Compute, Spectrum Scale, Spectrum Protect, Spectrum Archive, Spectrum Virtualize and everything that’s possible with this suite of tools.
While aesthetically pleasing rack diagrams may lead some to believe in the idea of a turnkey, one-size-fits-all supercomputer appliance, at Re-Store we understand every supercomputer is a new machine, worthy of its own name.
Every supercomputer is unique. They are no longer collections of storage, nodes, and network. Vendors sell many of the components of a supercomputer and do a great job supporting the individual components in the machine, but they are not positioned to support the supercomputer as an integrated entity producing science.
THE RE-STORE DIFFERENCE
Customer success depends on the access to knowledge and experience to solve “single-instance” problems that arise for one-of-a-kind machines, executing unique computations to achieve new insight.
Re-Store takes a personalized, one-on-one approach to building and operating supercomputers. We collaborate with each and every customer to find the most cost-effective solution to meet the organization’s needs.
Re-Store is a boutique provider of the skills and experience to design, resell, install, support, monitor, and operate the finished supercomputing machine. Our services range from simple resale to daily operation of the cluster.
Our goal is to fill the gaps in the customer’s experience and knowledge to ensure success. Bioinformatics is the important multi-disciplinary bond between the science and the computational resource. We offer the knowledge and experience of doctoral-level supercomputer engineers and bioinformaticists to augment the customer’s resources.
We’re not staff augmentation, but rather knowledge as a service (KaaS) that becomes a part of the customer team, instead of replacing the customer team.
Accepting the idea that a clever rack of hardware can simplify this process is anathema to the solution and will lead to failure.
Re-Store can add value beyond what vendors can provide by offering:
- AutonomousHPC: Our custom-built machine learning system that understands the behavior of a machine to more effectively manage it. Take a look at our AutonomousHPC video for a high-level overview of how our system leverages machine learning using billions of data points on the telemetry collected from the machine.
- AfterCare: Our custom-designed knowledge and skills augmentation service for the customer. AfterCare services range from occasional health checks to comprehensive operation of a production genomic pipeline.
Re-Store builds, designs, and operates clusters in a variety of disciplines, including genomics, computational neuroscience, autonomous car, and insurance. As proof of performance, we can arrange a call with the HPC Director at a major clinical genomics laboratory to document the success of Re-Store’s daily operation of the pipeline.