Tencent, one of the largest cloud providers in the world, is accelerating its data warehousing with the help of MemVerge’s Memory-Converged Infrastructure (MCI) “breakthrough” technology.
MemVerge’s MCI system would offer a new architecture for the enterprise, offering both higher capacity computing memory and faster storage at the same time. MemVerge offers a system design that would provide greater data center reliability and tackle memory, storage and IO bottlenecks – to help massive and complex applications operate smoothly, at memory speed.
Alongside Intel, MemVerge is working closely with Tencent to apply persistent memory technology to accelerate the Tencent Sparkling Data Warehouse Suite, a “high-performance” petabyte-level solution allowing for an “easy-to-use enterprise-grade” distributed cloud data warehouse.
“The memory-speed processing power of MemVerge’s Memory-Converged Infrastructure has proven unique value in accelerating speed of OLAP services and combined with the fact that we can store data on the same system,” said Long Wang, vice president of Tencent Cloud and general manager of Big Data and AI Services. “MemVerge’s technology has huge potential to be a compelling part of the foundation for our next-generation cloud-based data warehouse to provide our services to customers more effectively well into the future.”
Accelerating Data Warehousing Processes
Tencent Cloud runs an elastic cloud-native Spark application built on Tencent Cloud Sparkling Data Warehouse Suite service, and Intel’s non-volatile memory technology would provide end users with speed and flexibility. MemVerge’s Distributed Memory Objects (DMO) software layer builds a high-performance Spark solution with a non-volatile, memory-based converged infrastructure platform.
In terms of the Tencent network topology, the fused memory platform is the host of the Spark data IO. It is deployed on a separate host cluster and separately from the Spark compute cluster, and is interconnected. Data is exchanged with the Spark compute cluster through a “high-speed” general-purpose Ethernet network. MemVerge DMO technology would help to increase performance and make the Spark cluster more elastic, as well as enhance Sparkling to gracefully decommission nodes by decoupling shuffle data to the MemVerge system with the goal of accelerating the data warehousing processes.
“Performance, ease of use and the ability to provide seamless integration is part of the distributed memory objects software layer that MemVerge can provide,” said Charles Fan, CEO and founder of MemVerge. “Our collaboration with Tencent is proving successful because our goals to overcome crippling barriers and eliminate troublesome data bottlenecks are well aligned.”