Gary Smerdon: Making In-Memory Computing Accessible with Software-Defined Servers

There’s no denying the fact that in today’s digital world, data is the new gold. Businesses across the globe are struggling with processing their increasing amounts of data and noticing that their IT infrastructure cannot handle their growing workloads. They keep coming back to the same questions: What systems will they need to maximize the value of their data…today, tomorrow, and 5 years from now? How can they keep their data in-memory to ensure optimal performance? How can they deploy their systems quickly and agility to respond to evolving requirements? How can they do all of this while keeping their costs down?

This is where TidalScale comes into the picture. The company’s industry leading software solution aggregates the memory, cores, and I/O of multiple physical servers to create a single, software-defined server. These software-defined servers enable in-memory performance for large workloads, eliminating the need for costly scale-up solutions or complex scale-out infrastructure—thereby allowing customers to run their same workloads for HALF the TCO compared to alternative options.

Software-defined servers represent the world’s first use of machine learning below the application layer. They are built upon the standard commodity hardware, require no changes to applications and operating systems, and can be deployed within minutes on-premises or in the cloud. TidalScale simplifies and speeds up deployment by up to 25 percent.

Revolutionizing the Industry

TidalScale was founded by Ike Nassi, former Chief Scientist at SAP. At the time, in-memory databases came with a fundamental flaw: the amount of DRAM that could be physically put in a server was limited by how much a processor can physically access. The limitation was determined by the number of addresses, data, and control pins one could put on a modern processor. The sad truth was that there were physical limits on the number of pins on a processor, which prevented users from putting unlimited amounts of DRAM directly attached to a processor.

Thus, a new interest in distributed computing was born. Distributed computing involves multiple servers connected over a network. However, distributed computing solutions required writing custom applications to access data on different servers, and businesses had to figure out in advance how to partition data amongst a set of servers and then maintain that partitioning.

In founding TidalScale, Ike came up with a solution that was fundamentally different and new. It hides the complexity of distributing computing among servers. The old methodologies are still in place i.e. the servers are connected over the network and take full advantage of the economies of scale of modern commodity servers. However, TidalScale allows users to create one single software-defined machine running across multiple servers.

About TidalScale’s CEO

At present, Gary holds the duties of being the President and CEO at TidalScale. Throughout his 35-year tech career, Gary has successfully led business units, corporate strategy, marketing, and engineering for large global organizations and has steered multiple start-ups from early concepts to significant valuations and acquisitions.

Previously, Gary was EVP & Chief Strategy Officer at Fusion-io, where he drove their $1.3 billion-dollar acquisition by SanDisk. Prior to that, Gary founded LSI and served as the SVP and GM of their Accelerated Solutions Division, leading the start-up to over $100M in revenue within two years.

“Run your largest database and analytics workloads at HALF the cost. The future of computing is here”

Gary has been a disruptor and change-maker throughout his career, having been directly involved with the first high-speed personal computer and conceiving/executing on the breakthrough concept of networking as a standard feature in PCs. He holds three patents and graduated from Duke University with a BSEE.

Under his remarkable leadership, TidalScale’s groundbreaking software solutions have revolutionized how servers can be deployed to fit user needs, enabling large-scale computing power with minimized costs and overhead.

TidalScale’s Unique Value

Businesses running SAP and other large databases must continually ask themselves: how much memory is needed to hold their growing data sets—today, and tomorrow? What data growth rate is expected? What compression rate is expected? What size system is needed for dev, test, QA, production, HA, and disaster recovery? The list of factors to consider is endless.

TidalScale believes that businesses shouldn’t have to undergo months of costly planning and still be forced to guess at these answers. Hence, it empowers organizations to deploy a system that adapts flexibly to evolving needs, quickens time-to-deployment by 25X, and cuts TCO in half.

By allowing IT organizations to achieve in-memory performance for even their largest database workloads—and to do so at HALF the TCO compared to traditional alternatives—TidalScale is making powerful in-memory compute accessible to almost all businesses. Amidst these unprecedented times of economic uncertainty, this value proposition is especially powerful.

Solutions for Tomorrow

At present, TidalScale is heavily concentrated on empowering its partners to identify the right businesses that could benefit from the solution and helping these firms bring the technology to their clients.

Throughout the next decade, TidalScale will work with its network of partners to help the industry understand how software-defined servers work and how they can benefit IT organizations across the globe.