DRC is the leading provider of programmable coprocessor systems critical for compute-intensive applications. DRC accelerates essential algorithms and total applications in high-performance computing for business and science.
Modern business problems continually demand cheaper, faster and more efficient answers—whether performing risk analysis for financial institutions, generating accurate seismic images for oil and gas exploration, or mapping a genome. There is an equally pressing demand for industry-standard platforms and simple-to-use tools to run existing applications with little or no modification.
DRC has met this demand by creating the optimum way to directly connect a programmable unit a processor bus and system memory. By moving application processing from software running on a general-purpose CPU to specialized programmable hardware, DRC’s products deliver dramatic performance improvements. Additionally, DRC leverages industry standard systems for even greater ease of implementation and price/performance advantage.
The original work on accelerating applications in hardware began with Steve Casselman’s recognition that general-purpose computing could be transformed by using FPGAs. After perfecting the technology and with five patents issued pertaining to the use of reconfigurable logic in high performance systems, the time is right for release of a modular coprocessor system.
DRC’s implementation that leverages an industry-standard processor socket and open interconnect architectures has emerged as the ideal solution for today’s markets. DRC has also developed the core logic, system software, and user APIs that complete the system required to achieve results in real-world environments.
The DRC vision, therefore, is to continually innovate in the field of acceleration technologies to advance the performance of HPC applications. DRC anticipates broad adoption across industries for CPU-intensive applications such as cryptography, synthesis, searching/sequencing, modeling, simulation, rendering, compression, and protocol offload.
