Nvidia Corp. on Wednesday said that Tesla-powered Tsubame 2.0, a supercomputer located at Tokyo Institute of Technology, is the "greenest" among the highest-performing supercomputer on the planet. The Tsubame 2.0, which is based on Intel Xeon processors and Tesla compute accelerators, is the most energy-efficient PFLOPS-class supercomputer on the planet.
Tsubame 2.0 is comprised of HP ProLiant SL390 servers with Intel six-core Xeon X5670 central processing units accelerated by Nvidia Tesla M2050 accelerators, which provide more than 80% of its horsepower, according to Nvidia, enabling Tsubame 2.0 to achieve high levels of performance with very low power usage. The maximum performance of the Tsubame 2.0 is 1.192 PetaFLOPS, according to the Top 500 list.
Tsubame 2.0 is a heterogeneous supercomputer (combining both microprocessors and GPUs) used to accelerate a range of scientific and industrial research in Japan. With sustained performance of 1.19 petaflops per second while consuming 1.2MW, Tsubame 2.0 delivers 958MFLOPS of processing power per watt of energy. It is 3.4-times more energy efficient than the next-closest x86 CPU-only PFLOPS system, the Cielo Cray supercomputer at Los Alamos National Laboratory, which delivers 278MFLOPS per watt.
In the race to exascale computing, power efficiency has become the defining element of computing performance. Heterogeneous GPU-accelerated systems are inherently more energy efficient than CPU-only systems because applications can take advantage of the different processors for executing different jobs. The sequential parts of the application runs on CPUs, and the data- and compute-intensive parts are accelerated by the massively parallel GPU processor.
The latest Green500 list underscores the energy efficiency of heterogeneous computer design. Five of the world's 10 most efficient systems, and 22 of the top 30 most efficient systems, combine GPUs with CPUs.