 |
|
|
|
|
 |
|
|
|
|
|
|
START OF PRINT CONTENT - DO NOT CHANGE THIS LINE
High Performance Computing
LiquidIQ's unique fabric computing architecture delivers industry-leading performance, scalability and manageability for high performance computing applications.
LiquidIQ provides a high performance environment that offers Cray-like, or better, performance at commodity blade server economics, including:
- Performance optimized SMP
- Best-in-class STREAM memory bandwidth performance
- Performance optimized multi-core
- Sustained high bandwidth, low latency fabric communications performance over scale with no bottlenecks or hotspots
- Performance optimized MPI reduces overhead associated with communications
- Hardware-based processor offload for MPI and TCP communications
- HPCC benchmark latency performance beats high-end Cray T3E
- CRASH benchmark performance beats high-end Cray XT3
For application users, LiquidIQ offers:
- Applications that run significantly faster, or require less resources to execute as before, e.g. data warehousing algorithms run up to 3 times faster, reservoir modeling software runs 30% or more faster
- Applications can be scaled to take advantage of more computing resources than was possible with traditional systems due to the elimination of bottlenecks.
Processors spend less time handling communications and more time on computing requiring less resources for a given workload
For system administrators, LiquidIQ offers:
- Industry leading Green IT capabilities to help manage power and cooling challenges
- High availability, self-healing capabilities that mean less downtime
- Automatic and policy-based operations which can reduce operations costs by up to 80%
- The ability to support mixed workloads using SD-RTI capabilities for deploying logical infrastructure, which means that users can share the resource without performance impact or degradation.
END OF PRINT CONTENT - DO NOT CHANGE THIS LINE
|
| |
|
|
|
|
|
