The Analysis of Generic SIMT Scheduling Model Extracted from GPU

To improve the performance of processor, more and more companies during the industrial circle put the single instruction multi-threads (SIMT) scheduling technology into the processor architecture now, which can develop the multicore processor multi-thread parallel performance through promote the ability of processor multi-thread parallel processing. In order to research and develop the technology of SIMT, this article extracts a generic SIMT scheduling model from Graphic Processing Unit (GPU) which is a kind of processor that used in the field of high performance computing. Through analyzing the performance of this scheduling model, this article shows the attributes of this model and can be an important reference for the use and optimizing of this model in other processors.

[1]  Matei Ripeanu,et al.  StoreGPU: exploiting graphics processing units to accelerate distributed storage systems , 2008, HPDC '08.

[2]  Norman P. Jouppi,et al.  Optimizing NUCA Organizations and Wiring Alternatives for Large Caches with CACTI 6.0 , 2007, 40th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO 2007).

[3]  Kevin Skadron,et al.  Dynamic warp subdivision for integrated branch and memory divergence tolerance , 2010, ISCA.

[4]  Pradeep Dubey,et al.  Debunking the 100X GPU vs. CPU myth: an evaluation of throughput computing on CPU and GPU , 2010, ISCA.

[5]  William J. Dally,et al.  The GPU Computing Era , 2010, IEEE Micro.

[6]  S.A. Manavski,et al.  CUDA Compatible GPU as an Efficient Hardware Accelerator for AES Cryptography , 2007, 2007 IEEE International Conference on Signal Processing and Communications.

[7]  Greg Humphreys,et al.  How GPUs Work , 2007, Computer.

[8]  Tor M. Aamodt,et al.  Dynamic Warp Formation and Scheduling for Efficient GPU Control Flow , 2007, 40th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO 2007).