May you live in interesting times, cautions the age-old proverb. As computing chips face the fundamental limitations of miniaturization, it is sure to be interesting times, indeed. One of the most pressing issues facing the scientific community is the inability of today’s best computers to process the large-scale simulations needed for understanding complex physical systems.

“Over the past half century, as supercomputers got faster and more powerful, such simulations became ever more accurate and useful,” states the Defense Advanced Research Projects Agency (DARPA). “But in recent years even the best computer architectures haven’t been able to keep up with demand for the kind of simulation processing power needed to handle exceedingly complex design optimization and related problems.”

To remedy this situation, DARPA is seeking ideas on how to speed up the computation of the complex mathematics that undergird scientific computing. Specifically the agency is looking for assistance with a class of equations, known as partial differential equations. These equations, which describe fundamental physical principles of motion, diffusion and equilibrium, involve continuous rates of change over a large range of physical parameters. These are problems that are not easily broken into discrete parts to be solved by individual CPUs.

“The standard computer cluster equipped with multiple central processing units (CPUs), each programmed to tackle a particular piece of a problem, is just not designed to solve the kinds of equations at the core of large-scale simulations, such as those describing complex fluid dynamics and plasmas,” said Vincent Tang, program manager in DARPA’s Defense Sciences Office.

“A processor specially designed for such equations may enable revolutionary new simulation capabilities for design, prediction, and discovery. But what might that processor look like?” asks the DARPA invitation.

Before the digital era, equations were solved analog-style by manipulating continuously changing values instead of discrete measurements. The analog computer goes back more than 100 years but was displaced when transistor-based digital computers rose to prominence in the 1950s and 1960s based on their ability to solve a wide range of problems.

DARPA suggests that the time is right for taking another look at using analog substrates for the efficient simulation of “systems governed by complex, simultaneous, locally interacting, and non-linear phenomena,” especially given the advances that have been made in microelectromechanical systems, optical engineering, microfluidics, metamaterials and even DNA computing. If the performance advantage is significant enough, the analog coprocessor could be the next big thing in heterogenous computing.

The RFI seeks new processing paradigms that have the potential to overcome current barriers in computing performance – analog, digital, or hybrid approaches are all welcome.

From the announcement:

The RFI invites short responses that address the following needs, either singly or in combination:

• Scalable, controllable, and measurable processes that can be physically instantiated in co-processors for acceleration of computational tasks frequently encountered in scientific simulation.
  
• Algorithms that use analog, non-linear, non-serial, or continuous-variable computational primitives to reduce the time, space, and communicative complexity relative to von Neumann/CPU/GPU processing architectures.
  
• System architectures, schedulers, hybrid and specialized integrated circuits, compute languages, programming models, controller designs, and other elements for efficient problem decomposition, memory access, and task allocation across multi-hybrid co-processors.
  
• Methods for modeling and simulation via direct physical analogy.
  
• Technology development beyond these areas will be considered so long as it supports the RFI’s goals.
• DARPA is particularly interested in engaging nontraditional contributors to help develop leap-ahead technologies in the focus areas above, as well as other technologies that could potentially improve the computational tractability of complex nonlinear systems.

DARPA’s Request for Information (RFI) – titled Analog and Continuous-variable Co-processors for Efficient Scientific Simulation (ACCESS) – is available at: http://go.usa.gov/3CV43. Responses are due by 4:00 p.m. Eastern on April 14, 2015.