CSMT: Simultaneous Multithreading for Clustered VLIW Processors
Manoj Gupta, Fermín Sánchez Carracedo, Josep Llosa
Abstract
Manoj Gupta, Fermín Sánchez Carracedo, Josep Llosa
Abstract
Simultaneous MultiThreading (SMT) is a well-known technique that improves resource utilization by exploiting thread-level parallelism at the instruction grain level. However, implementing SMT for VLIWs requires complex structures, which is contrary to the VLIW philosophy of hardware simplicity. In this paper, we propose Cluster-level Simultaneous MultiThreading (CSMT) to allow some degree of SMT in clustered VLIW processors with low hardware cost and complexity. CSMT considers the set of operations that execute simultaneously in a given cluster as the assignment unit. To minimize cluster conflicts between threads, a very simple hardware-based cluster renaming mechanism is proposed. The hardware required to implement CSMT is cheap, realistic, and practical for a clustered VLIW processor. An analysis of the hardware required to implement CSMT shows that it is quite scalable, with up to eight threads easily supported at low hardware cost. The experimental results show that CSMT significantly improves performance when compared with other multithreading approaches suited for VLIW. For instance, with four threads, CSMT shows an average speedup of 110 percent over a single-thread VLIW architecture and 40 percent over Interleaved MultiThreading (IMT). In some cases, speedup can be as high as 225 percent over single-thread architecture and 84 percent over IMT.
OpenAlex reports 4 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Simultaneous MultiThreading (SMT) is a well-known technique that improves resource utilization by exploiting thread-level parallelism at the instruction grain level. However, implementing SMT for VLIWs requires complex structures, which is contrary to the VLIW philosophy of hardware simplicity. In this paper, we propose Cluster-level Simultaneous MultiThreading (CSMT) to allow some degree of SMT in clustered VLIW processors with low hardware cost and complexity. CSMT considers the set of operations that execute simultaneously in a given cluster as the assignment unit. To minimize cluster conflicts between threads, a very simple hardware-based cluster renaming mechanism is proposed. The hardware required to implement CSMT is cheap, realistic, and practical for a clustered VLIW processor. An analysis of the hardware required to implement CSMT shows that it is quite scalable, with up to eight threads easily supported at low hardware cost. The experimental results show that CSMT significantly improves performance when compared with other multithreading approaches suited for VLIW. For instance, with four threads, CSMT shows an average speedup of 110 percent over a single-thread VLIW architecture and 40 percent over Interleaved MultiThreading (IMT). In some cases, speedup can be as high as 225 percent over single-thread architecture and 84 percent over IMT.
Key concepts: Very long instruction word, Multithreading, Computer science, Simultaneous multithreading, Speedup, Thread (computing), Parallel computing, Instruction set