Live Webcast 15th Annual Charm++ Workshop

-->

PPL/Charm++ at SC17

Comparing the Power and Performance of Intel’s SCC to State-of-the-Art CPUs and GPUs
| Ehsan Totoni | Babak Behzad | Swapnil Ghike | Josep Torrellas
IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS) 2012
Publication Type: Paper
Repository URL:
Abstract
Power dissipation and energy consumption are becoming increasingly important architectural design constraints in different types of processors (from embedded systems to large-scale supercomputers). To continue the scaling of performance, it is essential that we build parallel processor chips that make the best use of exponentially increasing number of transistors within the power and energy budget. Intel SCC is as an appealing option for future many-core architectures. In this paper, as a quantitative study, we compare and analyze the performance, power consumption and energy efficiency of different cutting-edge platforms that differ in architectural build using various scalable applications. These platforms include: 1. Intel Single-Chip Cloud Computer(SCC): many-core, 2. Intel Core i7: general purpose multicore, 3. Intel Atom: low-power processor, and 4. Nvidia ION: GPGPU. Our results show that GPGPUs have outstanding results in performance, power consumption and energy efficiency for many applications, but they require significant programming effort and are not general enough to show the same level of efficiency for all applications. “Light-weight” many-cores present an opportunity for better performance per watt over “heavy-weight” multi-cores, although, multicores are still very effective for some sophisticated applications. In addition, low power designs are not necessarily energy-efficient since the runtime delay effect can be much greater than the power savings.
TextRef
People
Research Areas