By Paul Abram, Data Center Practice Leader, and Nick Turk, Solution Architect
What if I told you that your dream of having supercomputing capabilities within your organization is no longer a fantasy?
Affordable supercomputing is now available to organizations of all sizes. Yes, you read that right. Supercomputing is no longer something only well-funded industries can take advantage of.
Naturally, we understand this news brings up questions, such as:
How can I use this technology today? When can I get started? Can I really afford it?
Before we answer those and others, let’s first take a moment to understand how we got here.
Supercomputing: A brief history
Supercomputing began in 1964 when the room-sized CDC 6600, the first supercomputer, was released. It could process 500 kiloflops on standard math operations and was the world’s fastest supercomputer from 1964–1969.1
In the following decades, several more supercomputers were released, each one breaking previous performance records. Toward the end of the 1980s, the Cray Y-MP was released and could process 333 megaflops at its peak.2
The 1990s saw drastic change. First, supercomputers could now be made with thousands of processors and off-the-self components, meaning anyone with a deep pocketbook could start building supercomputers.3 Second, the Japanese officially overtook the United States’ nearly 40-year dominance of the industry, when they released the SR2201 in 1996, which soon became the fastest in the world. It contained 2048 processors and could process an impressive 600 gigaflops.3
That brings us to the 2000s. As industry leaders, such as IBM, found ways to condense processing power, the price of supercomputing technology began to fall and footprints began to shrink. At the same time, process performance continued to increase to speeds never before seen.
So here we are today: smaller, faster, cheaper.
Workloads that used to require a giant mainframe to process can now be processed on computers so small that they can sit on someone’s desk. In fact, IBM PowerSystems is one that organizations are using to perform advanced computing and complete other groundbreaking work. Which brings us back to one of the questions initially posed:
How can I use this technology today?
Organizations of all types and sizes are using supercomputing to realize faster time to insights, accelerate time to market for new innovations and increase competitiveness.
Here are ways our customers are using their new high-performance data analytics (HPDA) and high-performance computing (HPC) capabilities:
- Healthcare: Analyzing data faster to improve patient outcomes and accelerating time to groundbreaking discoveries.
- Finance/Insurance: Making high-frequency trading even faster, running complex simulations and mitigating rising risk.
- Retail: Crunching customer data to improve sales forecasts and enhance existing relationships.
- Government/Higher Ed: Improving citizen services or boosting research power and capabilities in academia.
- Manufacturing: Taking manufacturing to the next level and accelerating innovation.
Now that you have some ideas for how organizations are using affordable supercomputing, let’s answer those remaining questions from above:
Can I really afford it? And when can I get started?
The answers to those questions are simply, yes and now.
Because of the advancements in technology, as we discussed in our brief history, one of the main barriers to supercomputing—a high price tag—has been eliminated. That means with the right partner, such as Logicalis and IBM, you can affordably own the processing power that enables HPDA/HPC.
In our next post, “4 steps to realizing supercomputing,” we’ll show you why getting started on your supercomputing journey immediately is possible, as well as share all the details about how to make supercomputing happen for your organization.
In the meantime, if you’d like more information about bringing supercomputing to your organization, contact one of our reps.