MemVerge Makes Big Memory Apps Sizzle

MILPITAS, Calif., April 6, 2021 /PRNewswire/ — MemVerge™, the pioneers of Massive Reminiscence software program, immediately introduced the discharge of Reminiscence Machine software program model 1.2. The software program delivers Massive Reminiscence efficiency and capability leveraging as much as 40 cores in 3rd Gen Intel Xeon Scalable processors (code named Ice Lake) and as much as 6TB of reminiscence capability per socket with Intel Optane persistent reminiscence 200 collection. The corporate additionally introduced its membership within the CXL™ Consortium, and 5 Massive Reminiscence Labs at Arrow, Intel, MemVerge, Penguin Computing, and WWT that are actually outfitted and accessible for Massive Reminiscence demonstrations, proof-of-concept testing, and software program integration.

“Reminiscence Machine v1.2 is designed to permit software distributors and end-users to take full benefit of Intel’s newest Xeon Scalable processor and Optane reminiscence expertise,” stated Charles Fan, CEO of MemVerge. “We began by offering entry to new ranges of efficiency and capability with out requiring modifications to purposes.”

Massive Reminiscence is Greater, Sooner, and Extra Accessible with Reminiscence Machine v1.2
Pioneered by MemVerge, Massive Reminiscence software program uniquely makes 100% use of accessible reminiscence capability whereas offering new operational capabilities to memory-centric workloads comparable to virtualized cloud infrastructure, in-memory databases, genomics, and animation/VFX. Reminiscence Machine v1.2 provides the next capabilities to make Massive Reminiscence larger, sooner, and extra accessible to an ever-broader set of purposes:

  • Help for third Gen Intel Xeon Scalable Processors – from 8 to 40 cores
  • Help for Intel Optane Persistent Reminiscence 200 Sequence – 32% extra bandwidth and as much as 6TB per socket vs. earlier technology
  • Centralized Reminiscence Administration – configuration, monitoring, administration, and alerts for DRAM and PMEM throughout the information heart
  • Redis & Hazelcast Cluster HA – coordinated in-memory snapshots amongst members of in-memory database clusters permits immediate restoration of the complete cluster.
  • Help for Microsoft SQL Server on Linux – Double the efficiency for OLTP with the identical reminiscence price by leveraging 100% of DRAM and PMEM capability and proprietary reminiscence tiering expertise. 
  • Help for KVM hypervisors – Reminiscence Machine 1.2 helps superior reminiscence administration for QEMU-KVM, enabling utilization of 100% of DRAM+PMEM capability, dynamic tuning of DRAM:PMEM ratio per VM, and minimizes efficiency degradation brought on by noisy neighbors. 
  • Acceleration of Genomic Analytics – Reminiscence Machine v1.1 has confirmed to speed up general single-cell analytics pipelines by 60% by eliminating storage I/O. Learn and write testing with v1.2 is anticipated to enhance these outcomes.
  • Animation and VFX HA and Effectivity – Autosave and in-memory snapshots enable animation and VFX apps to offer “Time Machine” capabilities that enable artists to share workspaces immediately and recuperate from crashes in seconds.

Based on Mark Wright, Expertise Supervisor for Chapeau Studios, “Initially, we opened a poly-dense scene in Maya and it took two-and-a-half minutes. Then, we opened a scene from a snapshot we might taken with Reminiscence Machine and it took eight seconds. Along with opening exponentially sooner, one other good thing about the Reminiscence Machine snapshot is that it will get an artist proper to the spot within the software the place they had been once they created a snapshot. There is not any have to repopulate the complete software.”

Apps Sizzle on Ice Lake in Testing by
Impartial Lab pulled collectively a server configured with Intel Optane 200 Sequence Persistent Reminiscence, Intel Gen 3 Xeon Scalable CPUs, and Reminiscence Machine software program from MemVerge. They carried out bulk insert and skim checks with kdb+, in addition to Redis Fast Restoration with ZeroIO Snapshot and Redis Clone with ZeroIO Snapshot.

In abstract, the configuration with 200 Sequence PMEM, Intel Gen 3 Xeon Scalable CPUs, and Reminiscence Machine software program demonstrated 2x learn efficiency and 3x write efficiency. For extra particulars, learn the complete assessment.

Based on Kevin O’Brien, Lab Director at, “The actual advantages of PMEM present up when you’ll be able to leverage it on the byte degree with the suitable software program. In lots of instances, software builders like SAP, tune their software to have the ability to leverage PMEM. Whereas that works for some purposes, there’s an alternative choice. Leverage a software-defined answer that is constructed from the bottom as much as assist companies leverage all the efficiency and persistence advantages PMEM 200 presents. To check this newest technology of PMEM, that is precisely what we did.”

Massive Reminiscence Labs Speed up New Expertise Analysis and Integration
Massive Reminiscence makes use of Intel Optane persistent reminiscence and MemVerge Reminiscence Machine software program. For IT organizations and distributors that want fast entry to a Massive Reminiscence setting, 5 Massive Reminiscence labs are actually accessible for demonstrations, proof-of-concept testing, and software program integration. Go to the Massive Reminiscence Lab web page at for details about capabilities and scheduling time in every lab: Arrow, Intel, MemVerge, Penguin Computing,, and WWT.

The Massive Reminiscence Alternative
DRAM was invented in 1969. Over 50 years later, it stays costly, scarce, and risky. Clearly DRAM’s velocity of evolution can’t maintain tempo with the demand of contemporary purposes that should course of massive portions of information and ship ends in real-time. Luckily, the invention of Intel 3D XPoint expertise breathed new life into the growing older section, and marked the arrival of the age of Massive Reminiscence Computing.    

By 2024, nearly 1 / 4 of all information created will probably be real-time information, and two-thirds of World 2000 companies may have deployed at the very least one real-time software that’s thought-about mission-critical, in line with IDC.

These next-generation purposes (NGAs) ceaselessly make use of massive information analytics and AI/ML. The actual-time workloads are discovered in lots of industries on-premises and within the cloud. Examples embody in-memory databases and fraud analytics in monetary companies, buyer profiling in social media, advice engines in retail, 3D animation in media & leisure, genomics in well being sciences, and safety forensics, to call only a few.

The result’s explosive adoption of Massive Reminiscence Computing designed for giant and quick information. IDC estimates the market alternative for persistent reminiscence will develop from $65 million to $2.6 billion by 2023. Coughlin & Associates estimates that income for persistent reminiscence will attain $25 billion by 2030, equal to income for DRAM.

About MemVerge
The appearance of persistent reminiscence is sparking a brand new period of Massive Reminiscence Computing the place purposes of any measurement can forgo conventional storage in favor of considerable, persistent and extremely accessible swimming pools of reminiscence. Reminiscence Machine™ software program from MemVerge makes this attainable by virtualizing DRAM and chronic reminiscence to type a platform for enterprise-class in-memory information companies. To study extra about MemVerge, go to

Intel, the Intel brand, and different Intel marks are logos of Intel Company or its subsidiaries. 

Media Contact:
Steve Sturgeon
[email protected]


Associated Hyperlinks