TidalScale Updates Its Software-Defined, Composable Server

Instead of taking a standard server and slicing it up into smaller pieces for individual applications, TidalScale aggregates computing power from various commodity-type boxes from within a system and puts it all into a single computing lake of DRAM.


Server-making startup TidalScale, which introduced itself and its new product a year ago at the Open Compute Project Summit, has unveiled the next generation of its software-defined server—what some analysts have called the missing piece of the composable software-defined data center.

Composable infrastructure is a trendy architecture that was given visibility by Hewlett Packard Enterprise's entry into the genre a few years ago. It is designed to ensure that the exact amount of computing resources—from processing power and storage to network fabric and virtualization—can be rapidly pulled together from a single resource pool to support an application, and then returned to the pool when they're no longer needed for the workload.

It’s a just-in-time approach that enables optimization of resources. This is similar to aggregating various storage silos into one huge data repository.

Fellow server-making newcomer Liqid also is plowing the field in the composable architecture sector and is gaining attention from some first-mover enterprises. eWEEK will continue to follow these data-center equipment advancements.

TidalScale's Composable World

TidalScale is doing something called inverse virtualization. Instead of taking a standard server and slicing it up into smaller pieces for individual applications, TidalScale aggregates computing power from various commodity-type boxes from within a system and corrals it all into a single computing ocean of DRAM (dynamic random-access memory) computing power.

As a result, Campbell, Calif.-based TidalScale claims its new-generation server offers high performance, scalability and manageability to go with its broad platform support. Organizations ostensibly gain business insights sooner with faster performance and agile workloads.

“Maintaining competitive advantage depends on an organization’s ability to quickly process and gain insight from real-time information, but the rapid growth of those data sets is outpacing the performance improvements in server hardware,” President and CEO Gary Smerdon said in a media advisory.

“The release of TidalScale’s next-generation software-defined server addresses these challenges and takes a giant leap toward a future where software-defined servers are the foundation of modern data centers.” 

Key features include:

Real-time machine learning engine: TidalScale deploys real-time machine learning beneath the operating system and applications to constantly optimize server performance and utilization.

Fast: A TidalScale software-defined server with sufficient memory and CPU resources often outperforms an undersized traditional server by 10X-1000X, the company said. In addition, users can access this “right-sized” server in minutes –a radical improvement over the weeks or months it takes to deploy traditional large servers.

Easy-to-use UI: TidalScale’s WaveRunner orchestration software has been enhanced to expand manageability and speed deployment. Using WaveRunner’s point-and-click control panel, creating an software-defined server of any size from a set of commodity servers takes three clicks and only five minutes.

Broad operating system and application support: TidalScale runs applications and operating systems with no modifications. Applications such as SAP HANA, Oracle Database, R and SPARK all benefit from in-memory performance on TidalScale servers. TidalScale is certified by enterprise Linux providers that include Red Hat, Ubuntu and SuSe Enterprise Linux. 

“The ongoing adoption of software-defined infrastructure is as inevitable as that of x86 server virtualization a decade or more ago, and will achieve similar levels of penetration,” John Morency, Vice-President of Gartner Research, said.

“Hardware will be viewed as being increasingly commoditized as the SDI abstraction layers mask the native capabilities and differentiators. As data center infrastructure becomes increasingly software-centric, new and agile software companies will disrupt many existing data center providers.”

For more information, go here.

Chris Preimesberger

Chris J. Preimesberger

Chris J. Preimesberger is Editor-in-Chief of eWEEK and responsible for all the publication's coverage. In his 15 years and more than 4,000 articles at eWEEK, he has distinguished himself in reporting...