Peak power reduction has been a critical challenge in the design of integrated circuits impacting the chip's performance and reliability. The reduction of peak power also reduces the power density of integrated circuits. Due to large IR-voltage drops in circuits, transistor switching slows down giving rise to timing violations and logic failures. In this paper, we present a new clock control strategy for peak-power reduction in VLSI circuits. In the proposed method, the simultaneous switching of combinational paths is minimized by taking advantage of the delay slacks among the paths and clustering the paths with similar slack values. Once the paths are identified based on the path delays and their slack values, the clustering algorithm determines the ideal number of clusters for the given circuit and for each cluster the maximum possible phase shift that can be applied to the clock. The paths are assigned to clusters in a load balanced manner based on the slack values and each cluster will have a phase shift possible on its clock depending on the slack. Thus, the proposed register-transfer level (RTL) method takes advantage of the logic-path timing slack to re-schedule circuit activities at optimal intervals within the unaltered clock period. When switching activities are redistributed more evenly across the clock period, the IC supply-current consumption is also spread across a wider range of time within the clock period. This has the beneficial effect of reducing peak-current draw in addition to reducing RMS power draw without having to change the operating frequency and without utilizing additional power supply voltages as in dual or multi VT approaches. The proposed method is implemented and tested through simulations using an experimental setup with Synopsys Tools Suite and Cadence Tools on the ISCAS'85 benchmark circuits, OpenCore circuits and LEON processor multiplier circuit. Experimental results indicate that peak power can be reduced significantly to at least 72% depending on the number of clusters and the phase-shifted clock identified as suitable for the given circuit by the proposed algorithms. Although the proposed method incurs some power overhead compared to the traditional clocking method, the overhead can be made negligible compared to the peak-power reduction as seen in the experimental results presented.

In the nanometer era, process, voltage, and temperature variations are dominating circuit performance, power, and yield. Over the past few years, statistical optimization methods have been effective in improving yield in the presence of uncertainty due to process variations. However, statistical methods overconsume resources, even in the absence of variations. Hence, to facilitate a better performance-power-yield trade-off, techniques that can dynamically enable variation compensation are becoming necessary. In this article, we propose a dynamic technique that controls the instance of data capture in critical path memory flops, by delaying the clock edge trigger. The methodology employs a dynamic delay detection circuit to identify the uncertainty in delay due to variations and stretches the clock in the destination flip-flops. The delay detection circuit uses a latch and set of combinational gates to dynamically detect and create the slack needed to accommodate the delay due to variations. The Clock Stretching Logic (CSL) is added only to paths, which have a high probability of failure in the presence of variations. The proposed methodology improves the timing yield of the circuit without significant overcompensation. The methodology approach was simulated using Synopsys design tools for circuit synthesis and Cadence tools for placement and routing of the design. Extraction of parasitic of timing information was parsed using Perl scripts and simulated using a simulation program generated in C++. Experimental results based on Monte-Carlo simulations on benchmark circuits indicate considerable improvement in timing yield with negligible area overhead.

The trends in technology scaling and the reduction in supply voltages have significantly improved the performance and energy consumption in modern microprocessors. Microprocessors are being built with higher degrees of spatial parallelism and deeper pipelines to improve performance, which, however, makes them more susceptible to transient faults. Radiation causes "transient faults” or "single-event transients” in logic, which, once propagated and latched, become full cycle errors or soft errors. If radiation hits memory elements, this is usually called an "single-event upset” or "soft error” as it can further propagate as a full cycle error. The problem of soft errors is further exacerbated in large multiprocessors employed in servers in which reliability is a key concern. In the past, the technique of lockstep execution of the original and the duplicate instructions has been used for error detection in multiprocessors. However, the execution of redundant threads in the on-chip multiprocessor (CMP) provides error detection at lower overheads, since the branch outcomes of the leading thread can be exploited during the execution of the trailing thread, and also because the interprocessor communication latency is a key concern for lockstepping. In this paper, we show that by mining various redundancies inherent within a single core, the interprocessor communication can be brought down to a minimum. Toward this, we propose techniques based on 1) temporal redundancy, 2) data value redundancy, and 3) information redundancy for error detection in multicore designs. We exploit temporal redundancy by using the "latency slack cycles” (LSC) of an instruction, which we define as the number of cycles before the computed result from the instruction becomes the source operand of a subsequent instruction. The value-based detection technique is explored by exploiting the width of the operands with small data values and information redundancy is exploited by the generation of residue code check bits for the source operands. We show that with a clustered core multiprocessor, the interprocessor communication overhead can be significantly reduced. In our proposed multicore design, when a soft error is detected, error correction is achieved by rolling back the execution to a previous checkpoint state and re-executing the instructions. The proposed techniques have been implemented on the RSIM simulation framework and validated using the SPLASH benchmarks. Experimental results indicate that the soft error detection schemes proposed in this work, can be implemented, on the average, with less than 10 percent increase in CPI on modern multicore designs.

With the continuous decrease in the minimum feature size and increase in the chip density, modern processors are being increasingly susceptible to soft errors. In the past, the technique of lockstep execution with redundant threads on duplicated pipelines have been used for soft error rate reduction which can achieve high error coverage but at the cost of large overheads in terms of area and performance. In this paper, we propose techniques for protection against soft errors in multi-core designs using (i) the properties of spatial and temporal redundancy and (ii) value based detection. We utilize temporal redundancy by using the ldquolatency use slackrdquo (LSC) of an instruction, which we define as the number of cycles before the computed result from the instruction becomes the source operand of a subsequent instruction, while spatial redundancy is exploited by duplicating the instruction to a nearby idle processor core. Further, the value based detection technique is explored by exploiting the width of the operands with small data values and the generation of residue code check bits for the source operands. When a soft error is detected, error correction is achieved by rolling back the execution to a previous checkpoint state and re-executing the instructions. The proposed techniques have been implemented on the RSIM simulation framework and validated using the SPLASH benchmarks. Our results indicate that the soft error detection schemes proposed in this work, can be implemented, on average, with less than 10% increase in CPI on modern multi-core designs.