Author:
Rahimi Abbas,Gupta Rajesh K.
Abstract
AbstractVoltage scaling, as the most important knob for energy efficiency, is limited by leakage and variability. Variability is arisen from various sources including static manufacturing process, dynamic voltage and temperature fluctuations, and temporal changes over time. To address these variations, designers resort to excessive margins. These margins are increasing rapidly and eventually obliterating any gains due to device scaling. As a consequence, reduction of margins in design has become an important research challenge. We demonstrate how to recover part of these margins through hardware/software codesign with examples in many-core GPUs and FPGAs. This naturally leads to a departure from traditional error-tolerant computing to approximate computing.
Publisher
Springer International Publishing
Reference49 articles.
1. Advanced Micro Devices, Inc: AMD Evergreen Family Instruction Set Architecture
2. Altera sdk for opencl: http://www.altera.com/products/software/opencl/opencl-index.html
3. Amd app sdk v2.9: http://developer.amd.com/tools-and-sdks/opencl-zone/amd-accelerated-parallel-processing-app-sdk/
4. Bernstein, K., Frank, D., Gattiker, A., Haensch, W., Ji, B., Nassif, S., Nowak, E., Pearson, D., Rohrer, N.: High-performance cmos variability in the 65-nm regime and beyond. IBM J. Res. Dev. 50(4.5), 433–449 (2006). https://doi.org/10.1147/rd.504.0433
5. Bhardwaj, S., Wang, W., Vattikonda, R., Cao, Y., Vrudhula, S.: Predictive modeling of the nbti effect for reliable design. In: Custom Integrated Circuits Conference, 2006, CICC ’06, pp. 189–192. IEEE (2006). https://doi.org/10.1109/CICC.2006.320885
Cited by
3 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. Approximate Memory with Protected Static Allocation;2022 IEEE 34th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD);2022-11
2. Dynamic fault-tolerant VLIW processor with heterogeneous Function Units;Microprocessors and Microsystems;2022-09
3. SmartApprox: Learning-based configuration of approximate memories for energy-efficient execution;Sustainable Computing: Informatics and Systems;2022-04