Fresh out of recession, the finance and manufacturing sectors are looking to outcompute to outcompete this year. Spurred on by government encouragement and the monumental rise of big data, is super or High Performance Computing (HPC) at a tipping point?
HPC has been around since since the 1960s, used for a whole variety of purposes from weather forecasting to medical research. Yet, because of the cost and the skills needed to develop individual applications - as well as the recent recession - the HPC community has largely remained secluded in academia and research-focused.
But there are signs that the attractions of big data and necessity for increased computing power are pulling HPC out of its academic cloisters and into business.
The UK government’s recent investments in HPC suggest that it also recognises the commercial advantages that HPC can bring and the boost it could give to the economy. And it’s the financial and manufacturing sectors where take up has been quickest.
Financial sector drive
IDC, predicts that the highest growth in HPC up to 2016 will be the financial sector. Their breakdown is as follows:
11.1% in the financial sector
9.4% in computer-aided engineering
6.8% in geosciences
6.9% across the whole of industry.
For the financial industry, the rationale is obvious. The banking crisis left the industry searching for new ways to respond to its markets and big data analytics has become the weapon of choice, enabling banks and other financial institutions to target customers more precisely, analyse trends and mitigate risk.
However, other businesses such as security and retail are fast catching on too.
In the US, it’s security which is the driver behind a bill aiming to boost the U.S. supercomputer industry through the development of exascale computing, which is proposed as part of President Barack Obama’s 2015 budget.
The bill would set up partnerships between industry and national laboratories to research and develop exascale computing machines with the need for domestically-made exascale machines for national security to the fore. However research in energy, health, the environment and energy would also benefit.
Seeking to take advantage of the green shoots of recovery, manufacturing and engineering-based industries that use complex modelling and simulation are looking for new ways to meet global competition.
In Japan, manufacturers are increasingly turning to the nation’s fastest supercomputers – such as the 10-petaflop K supercomputer, installed at Japan’s RIKEN research institute – to gain a competitive advantage. According to the Nikkei Asian Review
, there are a number of projects in place, which are expected to yield results within a couple of years.
Software developed for K is being used by carmakers Toyota, Suzuki and tyre manufacturer, Bridgestone, to help them design their next-generation of products. Much of the advantage gained is the possibility for manufacturers to meet their prototyping needs without having to build full-scale physical designs.
Not only is the digital approach less costly and time-consuming, it enables greater innovation as new ideas can be tried out with a few clicks of the keyboard. Testing a large number of design parameters in a physical format just wouldn’t be feasible from an economic or time standpoint.
In car manufacturing for instance, digital modeling can enable engineers to determine the most aerodynamic shapes. Software can simulate the air resistance created by a car by interpreting the space around it as a grid of 2.3 billion segments. The computer simulation reflects how the air movement is affected by different driving conditions, for example a passing vehicle of a strong crosswind, with lower wind resistance enabling vehicles to be more fuel-efficient and increasing steering efficiency.
In the past, such accurate simulation has only achievable through constructing large wind tunnels and running tests with full-scale models.
In addition to the auto industry, Japan is also expanding its supercomputing efforts into the shipbuilding field. Here, software developed by the Shipbuilding Research Center of Japan can demonstrate how a ship’s movement creates turbulence. It needs as little water as 1mm to operate and can reduce design costs by up to 50 percent.
Materials science and pharmaceutical research are also beneficiaries of the K supercomputer, thus enabling a faster pace of discovery. While, in this sector, the main user base is still universities and labs, business project proposals have nearly doubled over the past year. Interest levels have encouraged Japan’s science ministry to work towards developing an exascale supercomputer, 100 times faster than the K, by 2020.
UK manufacturers are also taking advantage of supercomputing. By shortening the time taken to solve complex problems, manufacturers can curtail the the design phase, spend more time perfecting a product and still reduce the time it takes to get it to market.
With (according to CBI), UK manufacturers’ order books hitting an 18-year high at the end of 2013, speed and productivity need to step up.
Swansea-based medical supply company Calon Cardio-Technology is focused on developing smaller and more efficient blood pumps for use as an alternative to heart transplants in patients with chronic cardiac failure. Designing these artificial hearts so that they are both safe and effective requires a thorough understanding of how blood flows through the pump. This is a sophisticated computational fluid dynamics problem that is beyond the scope of most desktop computers.
Calon researchers are carrying out their design work with the help of Swansea University’s Advanced Sustainable Manufacturing Technologies centre. New designs are simulated on a supercomputer cluster managed by HPC Wales. Using a well-equipped desktop computer, each 3D simulation, requiring a “mesh” of approximately two million elements, would take two to three days to complete, but the supercomputer shortens that time significantly.
Making a difference
However, for HPC to truly make a difference in the 2014, continued governmental support is a pre-requisite. Computing skills shortages must be addressed and an awareness drive to drive home the advantages for SMEs.
Despite the competitive advantages conferred by HPC and a number of programs aimed at opening up access, the technology still remains out of reach to many ordinary businesses.
Hybrid computing using accelerators (NVIDIA GPUs) or coprocessors (Intel Phi) may offer one path and GPU maker NVIDIA is working with Imperial College to super-charge desktop computers with NVIDIA’s latest-generation Tesla parts. Each of these chips has about 2,500 compute “cores” while most desktop computers have perhaps two or four. The UK’s most powerful GPU-powered supercomputer, Emerald, has 372 Tesla M2090 GPUs.
And there are other ways of buying supercomputing access. HPC-on-Demand means businesses can buy HPC prowess as they would any other cloud-based solution. It can provide remote facilities with pre-installed and configured applications and open source codes, available within a low-latency, high bandwidth, compute environment.
These new developments can open up the world of supercomputers to expanding organisations that want to quickly take on more work and exceed capacity when needed. This means that SMEs, that have never previously had access to this kind of power, can now use it on a project basis over a defined period of time and only pay for what they use. They don’t waste investment by having infrastructure running idle. Supercomputing for SMEs has never been such a lucrative possibility.