• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.43 no.2 s.308
• /
• pp.1-11
• /
• 2006

In this paper, we propose a new register file architecture called the Register File Extension for Multi-words or Long-word Operation (RFEMLO) to accelerate both symmetric and asymmetric cryptographic algorithms. Based on the idea that most of cryptographic algorithms heavily use multi-words or long-word operations, RFEMLO allows multiple contiguous registers to be specified as a single operand. Thus, a single instruction can specify a SIMD-style multi-word operation or a long-word operation. RFEMLO can be applied to general purpose processors by adding instruction set for multi-words or long-word operands and functional units for additional instruction set. To evaluate the performance of RFEMLO, we use Simplescalar/ARM 3.0 (with gcc 2.95.2) and run detailed simulations on various symmetric and asymmetric cryptographic algorithms. By applying RFEMLO, we could get maximum 62% and 70% reductions in the total instruction count of symmetric and asymmetric cryptographic algorithms respectively. Also, performance results show that a speedup of 1.4 to 2.6 can be obtained in symmetric cryptographic algorithms and a speedup of 2.5 to 3.3 can be obtained for asymmetric cryptographic algorithms when we apply RFEMLO to a processor with an in-order pipeline. We also found that RFEMLO can effectively improve the performance of these cryptographic algorithms with much less cost compared to issue-width increase available in Superscalar implementations. Moreover, the RFEMLO can also be applied to Superscalar processor, leading to additional 83% and 138% performance gain in symmetric and asymmetric cryptographic algorithms.

• Journal of KIISE:Computer Systems and Theory
• /
• v.27 no.2
• /
• pp.200-208
• /
• 2000

Multi-ports register could shared several instructions at the same time in read operation. We address a formal methods for describing timing analysis and resource restriction in pipeline super scalar process that having multi-Port registers. First, we specify in-order pipeline instructions, and then, extend timing analysis in out-of-order super-scalar. In this case, we find instruction pairs in any cycle which can execute same time, We use ACSR(Algebra of Communicating Shared Resources), a branch of formal methods based on process algebra, for instruction specification and modelling.

• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.1_2
• /
• pp.67-77
• /
• 2004

Value prediction is a technique to obtain performance gains by supplying earlier source values of its data dependent instructions using predicted value of a instruction. To fully exploit the potential of value speculation, however, the efficient recovery mechanism is necessary in case of value misprediction. In this paper, we propose a sequential and selective recovery mechanism for value misprediction. It searches data dependency chain of the mispredicted instruction sequentially without pipeline stalls and adverse impact on clock cycle time. In our scheme, only the dependent instructions on the predicted instruction is selectively squashed and reissued in case of value misprediction.

• Journal of KIISE:Computer Systems and Theory
• /
• v.30 no.10
• /
• pp.569-578
• /
• 2003

Data dependencies are one of major hurdles to limit ILP(Instruction Level Parallelism), so several related works have suggested that the limit imposed by data dependencies can be overcome to some extent with use of the data value prediction. Hybrid value predictor can obtain the high prediction accuracy using advantages of various predictors, but it has a defect that same instruction has overlapping entries in all predictor. In this paper, we propose a new hybrid value predictor which achieves high performance by using the information of static and dynamic classification. The proposed predictor can enhance the prediction accuracy and efficiently decrease the prediction table size of predictor, because it allocates each instruction into single best-suited predictor during the fetch stage by using the information of static classification. Also, it can enhance the prediction accuracy because it selects a best- suited prediction method for the “Unknown”pattern instructions by using the dynamic classification mechanism. Simulation results based on the SimpleScalar/PISA tool set and the SPECint95 benchmarks show the average correct prediction rate of 85.1% by using the static classification mechanism. Also, we achieve the average correction prediction rate of 87.6% by using static and dynamic classification mechanism.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.44 no.2
• /
• pp.61-71
• /
• 2007

The instruction level parallelism, which has been used to improve the performance of processors, expose its limit. The change of a control flow by a branch miss prediction is one of the obstacles that restrict the instruction level parallelism. The single chip multiprocessors have been developed to utilize the thread level parallelism. However, we could not use the maximum performance of the single chip multiprocessor in case of executing the coded programs without considering the multi-thread. In order to overcome the two performance degradation factors, in this paper, we suggest the concurrent branch execution method that applies to the multi-path execution method at a single chip multiprocessor. We executes all two flows of the conditional branch using the idle core processor. Through this, we can improve the processor's efficiency with blocking the control flow termination by the branch instruction and reducing the idle time. We analyze the effects of concurrent branch execution proposed in this paper through the simulation. As a result of that, concurrent branch execution reduces about 20% of idle time and improves the maximum 10% of the branch prediction accuracy. We show that our scheme improves the overall performance of maximum 39% compared to the normal single chip multiprocessor and maximum 27% compared to the superscalar processor.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.45 no.11
• /
• pp.56-62
• /
• 2008

Modern microprocessors must deliver high application performance, while the design process should not subordinate power. In terms of performance and power tradeoff, the instructions window is particularly important. This is because a large instruction window leads to achieve high performance. However, naive scaling conventional instruction window can severely affect the complexity and power consumption. This paper explores an architecture level approach to reduce power dissipation. We propose a low power issue logic with an efficient tag translation. The direct lookup table (DTL) issue logic eliminates the associative wake-up of conventional instruction window. The tag translation scheme deals with data dependencies and resource conflicts by using bit-vector based structure. Experimental results show that, for SPEC2000 benchmarks, the proposed design reduces power consumption by 24.45% on average over conventional approach.

• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.1_2
• /
• pp.125-134
• /
• 2004

Recently, an on-chip compressed cache system was presented to alleviate the processor-memory Performance gap by reducing on-chip cache miss rate and expanding memory bandwidth. This research Presents an extended on-chip compressed cache system which also significantly expands main memory capacity. Several techniques are attempted to expand main memory capacity, on-chip cache capacity, and memory bandwidth as well as reduce decompression time and metadata size. To evaluate the performance of our proposed system over existing systems, we use execution-driven simulation method by modifying a superscalar microprocessor simulator. Our experimental methodology has higher accuracy than previous trace-driven simulation method. The simulation results show that our proposed system reduces execution time by 4-23% compared with conventional memory system without considering the benefits obtained from main memory expansion. The expansion rates of data and code areas of main memory are 57-120% and 27-36%, respectively.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.42 no.1
• /
• pp.9-26
• /
• 2005

In the pursuit of ever higher levels of performance, recent computer systems have made use of deep pipeline, dynamic scheduling and multi-issue superscalar processor technologies. In this situations, branch prediction schemes are an essential part of modem microarchitectures because the penalty for a branch misprediction increases as pipelines deepen and the number of instructions issued per cycle increases. In this paper, we propose a novel branch prediction scheme, direction-gshare(d-gshare), to improve the prediction accuracy. At first, we model a neural network with the components that possibly affect the branch prediction accuracy, and analyze the variation of their weights based on the neural network information. Then, we newly add the component that has a high weight value to an original gshare scheme. We simulate our branch prediction scheme using Simple Scalar, a powerful event-driven simulator, and analyze the simulation results. Our results show that, compared to bimodal, two-level adaptive and gshare predictor, direction-gshare predictor(d-gshare. 3) outperforms, without additional hardware costs, by up to 4.1% and 1.5% in average for the default mont of embedded direction, and 11.8% in maximum and 3.7% in average for the optimal one.