INDUSTRY

Liquid Computing Corp., a developer of a new class of scalable computing system, today announced the general availability of the world's first Interconnect Driven Server, LiquidIQ. LiquidIQ converges computing, networking and broadband to deliver unprecedented performance scalability and system control that is not available from today’s server and network equipment providers. In spite of point solution technologies such as Infiniband switches, blade servers, monitoring agents and grid software designed to help scale systems, users continue to be under pressure to address the inherent performance and system control bottlenecks associated with pre-integrated or user-integrated legacy servers, switches and software adjuncts. LiquidIQ’s Interconnect Driven Server architecture leverages the advantages of AMD64 technology with Direct Connect Architecture and HyperTransport™ technology to directly connect today’s powerful processors, memory, I/O and commodity switches into a unified system platform that is highly integrated, optimized and controlled. “As technical computing users expand their use of scalable computers, their requirements for scalable application performance in line with system size, manageable system complexity, controlled power and cooling costs, and assured resources availability is also growing,” said Dr. Earl Joseph, Research Vice-President of IDC's High Performance Systems Division. “Companies designing products that reduce or help manage the complexity of scalable systems will see strong interest from users looking to expand their computing options.” With over 15 current and upcoming trials in the oil and gas, financial services, government research, defense and software as a service sectors, LiquidIQ is helping organizations around the world gain a competitive edge in their Enterprise High Performance Computing and Software as a Service endeavors. Initial testing has demonstrated that LiquidIQ can deliver a 14 times or greater improvement over data center operations costs. Service Providers and Enterprise High Performance Computing users in government labs and corporations look at sustained performance scalability, operations simplification and material reductions in system footprint size, power and cooling costs. The combination of LiquidIQ’s integrated system architecture, transparent optimization software and granular system control are what make this architecture so unique. Leveraging new technologies such as HyperTransport technology, LiquidIQ can deliver linear, independent scaling of compute, interconnect and I/O resources without degradation across 12 chassis. “Liquid Computing has created a distinctive architecture that will further advance the ability of enterprises to perform compute-intensive processing,” stated Kevin Knox, Vice President, Commercial Business, AMD. “LiquidIQ has challenged and extended HyperTransport technology to bring sustained performance scalability and granular control over virtualized system resources.” LiquidIQ includes tightly integrated performance optimization libraries that are transparent to standards based applications and codes, but deliver up to five times or greater sustained communications bandwidth and low latency over commercial off-the-shelf servers and Infiniband switches. With a telecom grade control system, every system component is detected upon insertion and continuously monitored. Should a component fail or require servicing it is immediately isolated by the system and replaced through software driven sparing. The event driven control plane communicates over a separate network, has its own microprocessors, operating system, sensors and database to store the status and state of any module or virtual server configuration across the system. “Liquid Computing is dedicated to delivering sustained performance scalability and a fundamentally new control environment to Enterprise High Performance Computing customers,” said Brian Hurley, CEO and President of Liquid Computing. “LiquidIQ is bringing proven telecommunications reliability, availability and serviceability to scalable computing users across industries that have been thirsting for this type of performance and control.”