Home > Manycore programming > Why Manycore

Manycore programming

Why Manycore

Or how to combine multicore with GPUs

why manycore

 

Homogeneous multicore processors proposed by Intel or AMD have become mainstream. However heterogeneous architectures - also denoted manycore -  that combine a large number of cores with GPUs or with any hardware specialized co-processors usually offer a much higher peak performance/watt ratio. When high performance and power efficiency has to be achieved, specialized hardware is often the way to go. Combining general-purpose cores with highly parallel coprocessors like GPUs allows to both achieve high speedups on parallel sections of code while maintaining high performance on control sections.

Manycore performance opens HPC to many new scientific and consumer applications. New multimedia, medical and scientific applications will be developed by hundreds of thousands of engineers across the world.

However, the programming of heterogeneous architectures, also denoted hybrid programming, is quite a challenge for any application developer. The embedded market has been living with it for decades but at a very high programming cost. The general-purpose computing is now entering this era.

 

Other links

Fermi, FireStream, Larrabee, FPGA, OpenCL, DirectX, Cuda, Brook+, Top500, Roadrunner,  Tsubame, Cray CS1, Bull, QPACE, GPU Clusters