Sandia Labs in the US has been reviewing the multi-core processor stagnation problem - simulating key algorithms for deriving knowledge from large data sets.
The simulations show a significant increase in speed going from two to four multicores, but an insignificant increase from four to eight multicores. "Exceeding eight multicores causes a decrease in speed. Sixteen multicores perform barely as well as two, and after that, a steep decline is registered as more cores are added," said Sandia.
The problem is the lack of memory bandwidth as well as contention between processors over the memory bus available to each processor.
"To some extent, it is pointing out the obvious - many of our applications have been memory-bandwidth-limited even on a single core," said researcher Arun Rodrigues. "However, it is not an issue to which industry has a known solution, and the problem is often ignored."
According to Sandia, the obvious answer is more GHz, a solution which has already proved too power hungry.
"A more natural goal of researchers would be to increase the clock speed of single cores, since the vast majority of applications are designed for single-core performance on word processors, music, and video applications," said the Lab. "But power consumption, increased heat, and basic laws of physics involving parasitic currents meant that designers were reaching their limit in improving chip speed for common silicon processes."
The overall conclusion is that there is no right answer, although progress is being made.
"The chip design community didn't go with multicores because they were without flaw," says Mike Heroux of Sandia's scalable algorithms department. "The community couldn't see a better approach. It was desperate. Presently we are seeing memory system designs that provide a dramatic improvement over what was available 12 months ago, but the fundamental problem still exists."
Sandia has an investment in message-passing programs and its Institute for Advanced Architectures is intended to prepare the way for exaflop computing. It is working on answers.