This Article 
 Bibliographic References 
 Add to: 
iCFP: Tolerating All-Level Cache Misses in In-Order Processors
January/February 2010 (vol. 30 no. 1)
pp. 12-19
Andrew Hilton, University of Pennsylvania
Santosh Nagarakatte, University of Pennsylvania
Amir Roth, University of Pennsylvania

In-order continual flow pipeline (iCFP) is an in-order pipeline that allows execution to flow around data cache misses. When a cache miss occurs, iCFP executes and speculatively retires miss-independent instructions. It saves miss-dependent instructions in a slice buffer. When the miss returns, iCFP reexecutes the contents of the slice buffer and merges the results into working state. iCFP exploits existing support for multithreading and several novel components.

1. K. Krewell, "Sun's Niagara Pours on the Cores," Microprocessor Report, Sept.13, 2004, pp. 11-13.
2. H.Q. Le et al., "POWER6 Microarchitecture," IBM J. Research and Development, vol. 51, no. 6, 2007, pp. 639-662.
3. J. Dundas and T. Mudge, "Improving Data Cache Performance by Pre-executing Instructions under a Cache Miss," Proc. 11th Int'l Conf. Supercomputing, ACM Press, 1997, pp. 68-75.
4. S. Srinivasan et al., "Continual Flow Pipelines," Proc. 11th Int'l Conf. Architectural Support for Programming Languages and Operating Systems, ACM Press, 2004, pp. 107-119.
5. A. Hilton, S. Nagarakatte, and A. Roth, "iCFP: Tolerating All-Level Cache Misses in In-Order Pipelines," Proc. 15th IEEE Int'l Symp. High Performance Computer Architecture, 2009, IEEE CS Press, pp. 431-442.
6. S. Nekkalapu et al., "A Simple Latency Tolerant Processor," Proc. 26th IEEE Int'l Conf. Computer Design, IEEE CS Press, 2008, pp. 384-389.
7. S. Chaudhry et al., "Simultaneous Speculative Threading: A Novel Pipeline Architecture Implemented in Sun's Rock Processor," Proc. 36th Ann. Int'l Symp. Computer Architecture, ACM Press, 2009, pp. 484-495.
8. O. Ergin et al., "Increasing Processor Performance Through Early Register Release," Proc. 22nd IEEE Int'l Conf. Computer Design, IEEE CS Press, 2004, pp. 480-487.

Index Terms:
in-order processors, data cache misses, latency tolerance, multithreaded execution
Andrew Hilton, Santosh Nagarakatte, Amir Roth, "iCFP: Tolerating All-Level Cache Misses in In-Order Processors," IEEE Micro, vol. 30, no. 1, pp. 12-19, Jan.-Feb. 2010, doi:10.1109/MM.2010.20
Usage of this product signifies your acceptance of the Terms of Use.