You may recall our recent blog post introducing the NVIDIA TITAN X™ GPU, which made desktop supercomputing a much more affordable proposition. Well, NVIDIA has just raised the bar again with the GTX 1080 Ti. The GTX 1080 Ti looks to be a compelling alternative to the Titan X. It’s much more affordable, and in many aspects,…
Today’s high performance computing applications are becoming as resource-intensive and complex as they are diverse. Across disciplines such as Earth Sciences, Financial Services, Physics, and Machine Learning, demands for CPU power and memory bandwidth have increased exponentially. Within this set of workloads a common component is the need for highly parallel processing to maximize performance. However,…
Density and intelligence need not be mutually exclusive. For instance, take the Pogo Linux J4600S storage enclosure. With this solution, an IT admin can decrease both their storage footprint and their power consumption — saving precious power and rack space in the data center. Using a combination of an advanced JBOD design and HGST Ultrastar® drives, the…
For maximum IOPS and minimum latency without a ridiculous power bill, it’s time for a better host controller interface to your flash storage! Thanks to the NVMe™ (Non-Volatile Memory express) specification, flash devices can now use the speed and bandwidth of a PCIe connection, without incurring the overhead of emulating a SAS interface. The I/O…
NVIDIA has recently introduced the new TITAN X™, the biggest and most advanced GPU ever built. Blaise Pascal said, “Imagination decides everything.” This certainly applies to the hardware engineering team at NVIDIA. Not content with the already “irresponsible amount of performance” of their GTX 1080, they imagined breaking the 10 teraflops barrier for computing on a single chip….
