Posted by on
Categories: Netlist Super computing

IRVINE, Calif., Nov. 10, 2017 /PRNewswire/ — @Netlist, Inc. (NASDAQ: NLST) announced today its participation in the #SuperComputing2017 conference at the Colorado Convention Center in Denver, CO from November 13-16. Demonstrations showcasing the Company’s #HybriDIMM™, a next-generation storage class Memory solution that lowers cost and improves performance of big data analytics and in-memory computing, will take place at Netlist’s booth #2069. The company will also be demonstrating its #NVvault™ DDR4 (NV4) NVDIMM-N in a #hyperscale GPU storage appliance in cooperation with industry partners. The HybriDIMM demonstration will showcase two widely used open-source databases, for the first time, running concurrently on a Storage Class Memory (SCM) solution: @MongoDB exercises HybriDIMM’s in-memory channel block storage capabilities while the Redis in-memory database utilizes memory semantics to achieve a true persistent memory solution. With the flexibility to operate as main memory, block storage and persistent memory in a standard server, a single HybriDIMM can serve as both a memory or storage solution. HybriDIMM combines existing memory semiconductors, DRAM and NAND Flash, with intelligent “on-DIMM” co-processing to seamlessly manage the data transfer between on-DIMM storage and the system. HybriDIMM is recognized by any x86 server as a standard LRDIMM with no BIOS modifications or system changes. Operating modes include Main Memory access via on board DRAM and Block Storage and Persistent Memory accessed via on-board co-processors with NAND Flash. With HybriDIMM’s cost-effective fast storage and memory extension datacenter operators can capitalize on the promise of storage class memory by lowering the cost of running database applications, online transaction processing, big data analytics, and in-memory computing. The NV4 demonstration will showcase the latest NVDIMM-N non-volatile memory in a hyperscale GPU storage application providing sustained high throughput while enabling levels of storage resilience and data integrity impossible with traditional CPU or RAID-based solutions.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.