Abstract
Starting in 2004, the microprocessor industry has shifted to multicore scaling---increasing the number of cores per die each generation---as its principal strategy for continuing performance growth. Many in the research community believe that this exponential core scaling will continue into the hundreds or thousands of cores per chip, auguring a parallelism revolution in hardware or software. However, while transistor count increases continue at traditional Moore's Law rates, the per-transistor speed and energy efficiency improvements have slowed dramatically. Under these conditions, more cores are only possible if the cores are slower, simpler, or less utilized with each additional technology generation. This paper brings together transistor technology, processor core, and application models to understand whether multicore scaling can sustain the historical exponential performance growth in this energy-limited era. As the number of cores increases, power constraints may prevent powering of all cores at their full speed, requiring a fraction of the cores to be powered off at all times. According to our models, the fraction of these chips that is "dark" may be as much as 50% within three process generations. The low utility of this "dark silicon" may prevent both scaling to higher core counts and ultimately the economic viability of continued silicon scaling. Our results show that core count scaling provides much less performance gain than conventional wisdom suggests. Under (highly) optimistic scaling assumptions---for parallel workloads---multicore scaling provides a 7.9× (23% per year) over ten years. Under more conservative (realistic) assumptions, multicore scaling provides a total performance gain of 3.7× (14% per year) over ten years, and obviously less when sufficiently parallel workloads are unavailable. Without a breakthrough in process technology or microarchitecture, other directions are needed to continue the historical rate of performance improvement.
- Amdahl, G.M. Validity of the single processor approach to achieving. large-scale computing capabilities. In AFIPS (1967). Google ScholarDigital Library
- Azizi, O., Mahesri, A., Lee, B.C., Patel, S.J., Horowitz, M. Energy-performance tradeoffs in processor architecture and circuit design: A marginal cost analysis. In ISCA (2010). Google ScholarDigital Library
- Bhadauria, M., Weaver, V., McKee, S. Understanding PARSEC performance on contemporary CMPs. In IISWC (2009). Google ScholarDigital Library
- Bienia, C., Kumar, S., Singh, J.P., Li, K. The PARSEC benchmark suite: Characterization and architectural implications. In PACT (2008). Google ScholarDigital Library
- Borkar, S. The exascale challenge. Keynote at VLSI-DAT (2010).Google Scholar
- Chakraborty, K. Over-provisioned multicore systems. PhD thesis, University of Wisconsin-Madison (2008). Google ScholarDigital Library
- Chung, E.S., Milder, P.A., Hoe, J.C., Mai, K. Single-chip heterogeneous computing: Does the future include custom logic, FPGAs, and GPUs? In MICRO (2010). Google ScholarDigital Library
- Dennard, R.H., Cai, J., Kumar, A. A perspective on today's scaling challenges and possible future directions. Solid-State Electron., 5, 4 (Apr. 2007).Google Scholar
- Dennard, R.H., Gaensslen, F.H., Rideout, V.L., Bassous, E., LeBlanc, A.R. Design of ion-implanted mosfet's with very small physical dimensions. IEEE J. Solid-State Circuits 9 (Oct. 1974).Google ScholarCross Ref
- Guz, Z., Bolotin, E., Keidar, I., Kolodny, A., Mendelson, A., Weiser, U.C. Many-core vs. many-thread machines: Stay away from the valley. IEEE Comput. Archit. Lett. 8 (Jan. 2009). Google ScholarDigital Library
- Hameed, R., Qadeer, W., Wachs, M., Azizi, O., Solomatnikov, A., Lee, B.C., Richardson, S., Kozyrakis, C., Horowitz, M. Understanding sources of inefficiency in general-purpose chips. In ISCA (2010). Google ScholarDigital Library
- Hardavellas, N., Ferdman, M., Falsafi, B., Ailamaki, A. Toward dark silicon in servers. IEEE Micro 31, 4 (Jul--Aug. 2011). Google ScholarDigital Library
- Hempstead, M., Wei, G.Y., Brooks, D. Navigo: An early-stage model to study power-contrained architectures and specialization. In MoBS (2009).Google Scholar
- Hill, M.D., Marty, M.R. Amdahl's law in the multicore era. Computer 41, 7 (Jul. 2008). Google ScholarDigital Library
- Ipek, E., Kirman, M., Kirman, N., Martinez, J.F. Core fusion: Accommodating software diversity in chip multiprocessors. In ISCA (2007). Google ScholarDigital Library
- ITRS. International technology roadmap for semiconductors, the 2010 update, http://www.itrs.net (2011).Google Scholar
- Kim, C., Sethumadhavan, S., Govindan, M.S., Ranganathan, N., Gulati, D., Burger, D., Keckler, S.W Composable light weight processors. In MICRO (2007). Google ScholarDigital Library
- Moore, G.E. Cramming more components onto integrated circuits. Electronics 38, 8 (Apr. 1965).Google Scholar
- Pollack, F. New microarchitecture challenges in the coming generations of CMOS process technologies. Keynote at MICRO (1999). Google ScholarDigital Library
- SPEC. Standard performance evaluation corporation, http://www.spec.org (2011).Google Scholar
- Suleman, A.M., Mutlu, O., Qureshi, M.K., Patt, Y.N. Accelerating critical section execution with asymmetric multi-core architectures. In ASPLOS (2009). Google ScholarDigital Library
- Venkatesh, G., Sampson, J., Goulding, N., Garcia, S., Bryksin, V., Lugo-Martinez, J., Swanson, S., Taylor, M.B. Conservation cores: Reducing the energy of mature computations. In ASPLOS (2010). Google ScholarDigital Library
Index Terms
- Power challenges may end the multicore era
Recommendations
Dark silicon and the end of multicore scaling
ISCA '11: Proceedings of the 38th annual international symposium on Computer architectureSince 2005, processor designers have increased core counts to exploit Moore's Law scaling, rather than focusing on single-core performance. The failure of Dennard scaling, to which the shift to multicore parts is partially a response, may soon limit ...
Power Limitations and Dark Silicon Challenge the Future of Multicore
Since 2004, processor designers have increased core counts to exploit Moore’s Law scaling, rather than focusing on single-core performance. The failure of Dennard scaling, to which the shift to multicore parts is partially a response, may soon limit ...
Dark silicon and the end of multicore scaling
ISCA '11Since 2005, processor designers have increased core counts to exploit Moore's Law scaling, rather than focusing on single-core performance. The failure of Dennard scaling, to which the shift to multicore parts is partially a response, may soon limit ...
Comments