Once you’ve gotten your head around what HPC is (you may find What is HPC helpful if you haven’t already read it) you might find yourself wondering whether it applies to you. After all, if you’re not building rocket engines or working on a design for a new airplane or car, you might think that HPC doesn’t have much to offer you.
That was probably true 25 years ago when the only high performance computers around were supercomputers, multi-million dollar behemoths with legions of specialists taking care of them. But the same forces that have brought us $300 laptops have also put high performance computers into the hands of all kinds of people who would have never had access before, including thousands of small and medium-sized businesses all around the world.
That said, high performance computing still isn’t “cheap” — while you can get into a small cluster that will greatly increase your computing capability for $10,000 or so, that is still a large investment for most smaller companies, and you’ll want to do some thinking (and lots of planning) before you open up your checkbook.
The first step in that process is to develop answers to the question: “How could my business grow if I had access to 100x more computing power?”
A powerful question
That is a deceptively powerful question. The answer is definitely not “more email” or “more powerpoint slides,” — even today’s cheap computers are ridiculously over-powered for most of these office functions. And if your business focuses on hand-crafts sold in very small quantities, then the answer may well be that more computing power would have no impact on your business.
But I encourage you to spend some time thinking seriously about the question before you give up on the idea of growing your business by growing your ability to compute. There are almost as many uses for HPC in solving problems as there are problems to be solved.
For example, a local chain of four gas stations could use trend analysis software on a small cluster to run through the past 5 years of sales data and give you insights into what products are selling when. Using a computer for this problem rather than the owner’s intuition can reveal some surprising trends in your data that are regular and predictable, and which you can use to get the right merchandise into the store at the right time. Add shelf placement data to that mix and now you can start figuring out what shelves are selling the fastest, and improve sales by targeting high margin merchandise to those areas.
A small engineering and design firm is a natural business to consider HPC. Many of these companies are already using advanced computer simulation software on engineer workstations in their office. Perhaps you’ve been avoiding bidding on new contracts because you don’t have the resources to do the computer simulations in the window of time available before the response is due. HPC can help there, too, and the benefit in this situation is that these kinds of businesses aren’t adding computing to the business, they are just adding more computing to a business that already depends upon simulation.
The same arguments could apply to small manufacturing firms as well that use computer aid design and manufacturing software to optimize molds for casting, mixing, materials flows, or any number of industrial uses.
What about financial analysis? Let’s say that you want to give your clients an investment strategy tailored for the current market and their risk tolerance, but don’t have the horsepower to hindcast more than a few years of market data? HPC is there to help you, too, and there are even solutions available today that will let you use your tried and true Microsoft Excel worksheets on a cluster.
The key is having a clear picture of the problem you are trying to solve
The key to knowing whether an HPC investment makes sense for you is knowing what you want to accomplish. We are big believers in the power of HPC and supercomputing — we’ve spent our whole careers doing it! But buying an HPC cluster for your business with no clear idea of what you want from it doesn’t make any more sense than buying a forklift for a knitting company. Unless you are moving a lot of yarn, you probably aren’t going to get much out of that investment.