© The Institution of Engineering and Technology
The increasing demand for high-performance computing has emphasised the invocation of sophisticated multi/many-core computing architecture. Graphical Processing Unit (GPU) is considered to be an essential innovation in this regard as GPU offers a significant amount of parallelism in the execution of complex computing applications. The performance of GPUs in reducing the computational time of such applications is worth mentioning. Although GPUs appear to be a problem-solving solution for complex applications yet high power consumption has been a challenging problem, associated with this many-core computer architecture. Efficient resource management is a emerging and promising solution to this challenge; however, reducing the resources would degrade the system's overall performance. On the other hand, reducing the resources based on the analysis of workload can save significant power without degrading the system's overall performance. Therefore, a smart controller to optimise the resources of general purpose-GPU (GP-GPU) architecture is required. AFBRMC-2, a neuro-fuzzy type-2 based controller, is presented for GP-GPU architecture and, based on a feedback mechanism, keeps analysing the stats of processor and manages resources using dynamic voltage frequency scaling and core gating techniques. The proposed controller achieved up to 55% reduction in power consumption against various benchmarks on the NVIDIA TK1 GPU kit.
References
-
-
1)
-
1. Power, M.T.: ‘A first class design constraint for future architecture and automation’. Proc. of the 7th Int. Conf. on High Performance Computing, 2000, pp. 215–224.
-
2)
-
3. Toshiya, K., Hayashi, S., Nakada, T., et al: ‘Power capping of CPU-GPU heterogeneous systems through coordinating DVFS and task mapping’. 2013 IEEE 31st Int. Conf. on computer design (ICCD), 2013, pp. 349–356.
-
3)
-
6. Lungu, A., Bose, P., Buyuktosunoglu, A., et al: ‘Dynamic power gating with quality guarantees’. Proc. of the 2009 ACM/IEEE int. symp. on Low power electronics and design, San Francisco, California, USA, 2009, pp. 377–382.
-
4)
-
10. Ansari, A.Q., Gupta, N.: ‘Back propagation algorithm for neurofuzzy filter’, Int. J. Comput. Cogn. (IJCC), 2011, 9, (3), pp. 41–47.
-
5)
-
8. Mei, X., Wang, Q., Chu, X.: ‘A survey and measurement study of GPU DVFS on energy conservation’, Digit. Commun. Netw., 2017, 3.2, pp. 89–100.
-
6)
-
9. Bridges, R.A., Imam, N., Mintz, T.M.: ‘Understanding GPU power: a survey of profiling, modeling, and simulation methods’, ACM Comput. Surv. (CSUR), 2016, 49, (3), pp. 1–27.
-
7)
-
5. Hong, S., Hyesoon, K.: ‘An integrated GPU power and performance model’. Proc. of the 37th annual int. symp. on Computer architecture, Saint-Malo, France, 2010.
-
8)
-
2. Nugteren, C., Gert-Jan, V.D.B., Henk, C.: ‘Roofline-aware DVFS for GPUs’. Proc. of Int. Workshop on Adaptive Self-tuning Computing Systems, 2014, pp. 8–10.
-
9)
-
7. Najam, S., Ahmed, J., Masood, S., et al: ‘Run-time resource management controller for power efficiency of GP-GPU architecture’, IEEE Access, 2019, 7, pp. 25493–25505.
-
10)
-
4. Daecheol, Y., Chung, K.-S.: ‘Dynamic voltage and frequency scaling framework for low-power embedded GPUs’, Electron. Lett., 2012, 48, (21), pp. 1333–1334.
http://iet.metastore.ingenta.com/content/journals/10.1049/iet-cds.2020.0233
Related content
content/journals/10.1049/iet-cds.2020.0233
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Run-time neuro-fuzzy type-2 controller for power optimisation of GP-GPU architecture