Circuit techniques for reducing the minimum supply voltage (VMIN) of last-level and intermediate static random-access memory (SRAM) caches enhance processor energy efficiency. For the first time at a 16nm technology node, projections of a high-density 6-transistor SRAM bit cell indicate that the VMIN of a 4Mb or larger cache exceeds the maximum supply voltage (VMAX) for reliability. Thus, circuit techniques for cache VMIN reduction are transitioning from an energy-efficient solution to a requirement for cache functionality. Traditionally, error-correcting codes (ECC) such as single-error correction, double-error detection (SECDED) aim to protect the cache operation from radiation-induced soft errors. Moreover, during the qualification of a system-on-chip (SoC) processor, test engineers monitor the rate of correctable cache errors from SECDED for observing the on-die interactions between integrated components (e.g., CPU, GPU, DSP, etc.). This presentation highlights the opportunity to exploit ECC for reducing the cache VMINwhile simultaneously providing coverage for radiation-induced soft errors. Silicon test-chip measurements from a 7Mb data cache in a 20nm technology demonstrate a VMINreduction of 19% from SECDED. In addition, silicon measurements provide a salient insight in that only 0.12% of the cache words contain an error when operating at the cache VMINwith SECDED. Therefore, SECDED improves VMINby 19% while maintaining 99.88% coverage for radiation-induced soft errors. In applying SECDED for a lower cache VMIN, the rate of correctable errors exponentially increases, thus eliminating a useful metric for on-die observability. The presentation concludes by offering alternative solutions for on-die observability.

The minimum supply voltage (Vmin) is explored for sequential logic circuits by statistically simulating the impact of within-die process variations and gate-dielectric soft breakdown on data retention and hold time. As supply voltage (Vcc) scales, statistical circuit simulations demonstrate that hold time increases faster than circuit delay or cycle time, consequently the required number of min-delay buffers increases. For this reason, a new hold-time violation metric defines Vminas the Vccin which the hold time exceeds a target percentage of the cycle time. Simulation results in a 22nm tri-gate CMOS technology indicate a data-retention Vminof 0.61Vnorm and a hold-time Vminof 0.73Vnorm, where Vnormrepresents a normalized voltage for the process technology node. A key insight reveals that upsizing the first clock inverter in the sequential circuit reduces the hold-time Vminby 18% and the overall Vminby 16%.

The semiconductor industry is facing a critical research challenge: design future high performance and energy efficient systems while satisfying historical standards for reliability and lower costs. The primary cause of this challenge is device and circuit parameter variability, which results from the manufacturing process and system operation. As technology scales, the adverse impact of these variations on system-level metrics increases. In this paper, we describe an interdisciplinary effort toward robust and resilient designs that mitigate the effects of device and circuit parameter variations in order to enhance system performance, energy efficiency, and reliability. Collaboration between the technology, CAD, circuit, and system levels of the compute hierarchy can foster the development of cost-effective and efficient solutions.

This session brings together specialists from the DfT, DfY and DfR domains that will address key problems together with their solutions for the 14 nm node and beyond, dealing with extremely complex chips affected by high defect levels, unpredictable and heterogeneous timing behavior, circuit degradation over time, including extreme situations related with the ultimate CMOS nodes, where all processor nodes, routers and links of single-chip massively parallel tera-device processors could comprise timing faults (such as delay faults or clock skews); a large percentage of these parts are affected by catastrophic failures; all parts experience significant performance degradations over time; and new catastrophic failures occur at low MTBF.

Traditional processors are designed to guarantee error-free operation under worst-case (1) device & interconnect parameter variations resulting from less than ideal manufacturing process control; (2) static & erratic defects; (3) operating environments such as temperature excursions and voltage droops; (4) critical path activation and path delay degradations due to multiple inputs switching simultaneously in gates containing transistor stacks, or signal coupling from neighboring lines in interconnect paths; (5) speed degradation over the operating lifetime due to transistor aging under voltage, temperature & current stress; (6) early-life failures due to latent defect accelerations; and (7) soft error due to cosmic rays and alpha particle impacts. The voltage-frequency settings for all processors are set based on these infrequently encountered worst-case considerations, even though under typical conditions voltage can be pushed down further or frequency increased without causing errors for most of the processors, thus limiting both energy efficiency and performance in scaled CMOS technologies.

In this tutorial, a 45nm resilient microprocessor core with error-detection and recovery circuits demonstrates the opportunity for improving performance and energy efficiency by mitigating the impact of dynamic parameter variations. The design methodology describes the additional steps beyond a standard design flow for integrating error-detection and recovery circuits into a microprocessor core. Silicon measurements indicate that the resilient design enables a 41% throughput benefit at iso-energy or a 22% energy reduction at iso-throughput, as compared to a conventional design.

Conventional microprocessors require a clock frequency (F CLK ) guardband to ensure correct functionality during infrequent dynamic operating variations in supply voltage (V CC ), temperature, and transistor aging. Consequently, these inflexible designs cannot exploit opportunities for higher performance by increasing F CLK or lower energy by reducing V CC during favorable operating conditions. This presentation describes a 45nm resilient microprocessor with error-detection and recovery circuits to detect and correct timing errors from dynamic variations to mitigate the F CLK guardband, thus enabling higher performance or lower energy as compared to a conventional design. The microprocessor core supports two distinct error-detection designs and two separate error-recovery techniques, allowing a direct comparison of the relative trade-offs. Silicon measurements demonstrate that resilient circuits enable a 41% throughput gain at equal energy or a 22% energy reduction at equal throughput, as compared to a conventional design when executing a benchmark program with a 10% V CC droop. In addition, the resilient circuits guide an adaptive clock controller that tracks recovery cycles and adapts to persistent variations by changing F CLK . The combination of error-detection and recovery circuits with dynamic adaptation allows the microprocessor to adapt to the operating environment to deliver maximum efficiency. The presentation concludes by discussing the opportunity of applying resilient techniques to enhance the dynamic operating range (i.e., high-performance and low-power modes) for microprocessors.

Three circuit techniques for dynamic variation tolerance are presented: (i) Sensors with adaptive voltage and frequency circuits, (ii) Tunable replica circuits for timing-error prediction with error recovery, and (iii) Embedded error-detection sequential circuits with error recovery. These circuits mitigate the clock frequency guardbands for dynamic variations, thus improving microprocessor performance and energy-efficiency. These circuits are described with a focus on the different trade-offs in guardband reduction and design overhead. Opportunities for CAD to further enhance microprocessor performance and energy efficiency are offered.

Voltage and frequency margins necessary to ensure correct processor operation under dynamic voltage, temperature, and aging variations result in performance and power overheads. Resilient circuit techniques, including embedded error-detection sequentials and tunable replica circuits, allow these margins to be reduced or eliminated, resulting in reliable, energy-efficient operation.

A statistical performance simulator is developed to explore the impact of parameter variations on the maximum clock frequency (FMAX) and throughput distributions of multi-core processors in a future 22 nm technology. The simulator captures the effects of die-to-die (D2D) and within-die (WID) transistor and interconnect parameter variations on critical path delays in a die. A key component of the simulator is an analytical multi-core processor throughput model, which enables computationally efficient and accurate throughput calculations, as compared with cycle-accurate performance simulators, for single-threaded and highly parallel multi-threaded (MT) workloads. Based on microarchitecture designs from previous microprocessors, three multi-core processors with either small, medium, or large cores are projected for the 22 nm technology generation to investigate a range of design options. These three multi-core processors are optimized for maximum throughput within a constant die area. A traditional single-core processor is also scaled to the 22 nm technology to provide a baseline comparison.The salient contributions from this paper are: 1) product-level variation analysis for multi-core processors must focus on throughput, rather than just FMAX, and 2) multi-core processors are more variation tolerant than single-core processors due to the larger impact of memory latency and bandwidth on throughput.To elucidate these two points, statistical simulations indicate that multi-core and single-core processors with an equivalent total core area have similar FMAX distributions (mean degradation of 9% and standard deviation of 5%) for MT applications. In contrast to single-core processors, memory latency and bandwidth constraints significantly limit the throughput dependency on FMAX in multi-core processors, thus reducing the throughput mean degradation and standard deviation by ~50% for the small and medium core designs and by ~30% for the large core design. This improvement in the throughput distribution indicates that multi-core processors could significantly reduce the product design and process development complexities due to parameter variations as compared to single-core processors, enabling faster time to market for high-performance microprocessor products.

We explore the power benefits of changing a microprocessor path histogram through circuit sizing based on statistical timing analysis and optimization (STAO) versus a deterministic timing approach that uses statistical design to establish a global guardband followed by conventional optimization (SDGG). Using an analytical modeling approach, we quantify the differences in total power between the two approaches while maintaining an equivalent performance distribution. For a relative 1σ random WID stage delay variation of 5% and representative microprocessor critical paths, the analysis indicates that the STAO approach enables ~2% power reduction over the SDGG approach. To achieve a 4% and 6% power reduction through the STAO approach, the process variation needs to increase by a factor of 2x and 4x, respectively.

A statistical performance simulator is developed to explore the impact of die-to-die (D2D) and within-die (WID) parameter variations on the distributions of maximum clock frequency (FMAX) and throughput for multi-core processors in a future 22nm technology.allThe simulator integrates a compact analytical throughput model, which captures the key dependencies of multi-core processors, into a statistical simulation framework that models the effects of D2D and WID parameter variations on critical path delays across a die. The salient contributions from this paper are: (1) Product-level variation analysis for multi-core processors must focus on throughput, rather than just FMAX, and (2) Multi-core processors are inherently more variation tolerant than single-core processors due to the larger impact of memory latency and bandwidth on overall throughput. To elucidate these two points, multi-core and single-core processors have a similar chip-level FMAX distribution (mean degradation of 9% and standard deviation of 5%) for multi-threaded applications. In contrast to single-core processors, memory latency and bandwidth constraints significantly limit the throughput dependency on FMAX in multi-core processors, thus reducing the throughput mean degradation and standard deviation by 50%. Since single-threaded applications running on a multi-core processor can execute on the fastest core, mean FMAX and throughput gains of 4% are achieved from the nominal design target.

Insertion of time-borrowing (TB) flip-flops in multi-cycle repeater-based on-chip interconnects enables significant improvements in mean performance and energy by averaging systematic and random within-die (WID) delay variations across multiple interconnect segments. A statistically-based analytical model is derived to design a TB N-cycle interconnect with optimal delay variation tolerance. The model elucidates the dependency of the transparency window required to achieve data delay averaging on the delay variation mismatch between interconnect segments. Statistical circuit simulations and analyses in a 65nm process technology demonstrate that TB multi-cycle interconnects enable a 4-6% mean maximum clock frequency (FMAX) improvement and a corresponding 10% average energy savings over optimally designed multi-cycle interconnects with conventional master-slave flip-flops. The maximum mean FMAX benefit ranges from 4.0-7.5%, corresponding to approximately a bin-split shift in the FMAX distribution. For 1.41X larger WID delay variations, the maximum mean FMAX gain rises to 5-10%.

Die-to-die and within-die variations impact the frequency and power of fabricated dies, affecting functionality, performance, and revenue. Variation-tolerant circuits and post-silicon tuning techniques are important for minimizing the impacts of these variations. This paper describes several circuit techniques that can be employed to ensure efficient circuit operation in the presence of ever-increasing variations.

This paper explores the effectiveness of the simultaneous application of pipelining and parallel processing as a total power (static plus dynamic) reduction technique in digital systems. Previous studies have been limited to either pipelining or parallel processing, but both techniques can be used together to reduce supply voltage at a fixed throughput point. According to our first-order analyses, there exist optimal combinations of pipelining depth and parallel processing width to minimize total power consumption. We show that the leakage power from both subthreshold and gate-oxide tunneling plays a significant role in determining the optimal combination of pipelining depth and parallel processing width. Our experiments are conducted with timing information derived from a 65nm technology and fanout-of-four (FO4) inverter chains. The experiments show that the optimal combinations of both pipelining and parallel processing - 8 /spl sim/ 12 /spl times/ FO4 logic depth pipelining with 2 /spl sim/ 3-wide parallel processing - can reduce the total power by as much as 40% compared to an optimal system using only pipelining or parallel processing alone. We extend our study to show how process parameter variations - an increasingly important factor in nanometer technologies - affects these results. Our analyses reveal that the variations shift the optimal points to shallower pipelining and narrower parallel processing - 12 /spl times/ FO4 logic depth with 2-wide parallel processing - at a fixed yield point.

Fluctuations in intrinsic linear Vt, free of impact of parasitics, are measured for large arrays of NMOS and PMOS devices on a testchip in a 150nm logic technology. Local intrinsic σVT, free of extrinsic process, length and width variations, is random, and worsens with reverse body bias. Although the traditional area-dependent component is dominant, a significant component of the fluctuations in small devices depends only on device width or length.

A multilevel interconnect architecture design methodology that optimizes the interconnect cross-sectional dimensions of each metal layer is introduced that reduces logic macrocell area, cycle time, power consumption or number of metal layers. The predictive capability of this methodology, which is based on a stochastic wiring distribution, provides insight into defining the process technology parameters for current and future generations of microprocessors and application-specific integrated circuits (ASICs). Using this methodology on an ASIC logic macrocell case study for the 100 nm technology generation, the optimized n-tier multilevel interconnect architecture reduces macrocell area by 32%, cycle time by 16% or number of wiring tracks required on the topmost tier by 62% compared to a conventional design where pitches are doubled for every successive pair of levels. A new repeater insertion methodology is also described that further enhances gigascale integration (GSI) system performance. By using repeaters, a further reduction of 70% in macrocell area, 18% in cycle time, 25% in number of metal levels or 44% in power dissipation is achieved, when compared to an n-tier design without repeaters. The key distinguishing feature of the methodology is its comprehensive framework that simultaneously solves two distinct problems-optimal wire sizing and wiring layer assignment-using independent constraints on maximum repeater area for efficient design space exploration to optimize the area, power, frequency, and metal levels of a GSI logic megacell.

Minimum power CMOS ASIC macrocells are designed by minimizing the macrocell area using a new methodology to optimally insert repeaters for n-tier multilevel interconnect architectures. The minimum macrocell area and power dissipation are projected for the 100, 70 and 50 nm technology generations and compared with a n-tier design without using repeaters. Repeater insertion and a novel interconnect geometry scaling technique decrease the power dissipation by 58-68% corresponding to a macrocell area reduction of 70-78% for the global clock frequency designs of these three technology generations.

A circuit design methodology minimizing total power drain of a static complementary metal-oxide-semiconductor (CMOS) random logic network for a prescribed performance, operating temperature range, and short channel threshold voltage rolloff is investigated. Physical, continuous, smooth, and compact "transregional" MOSFET drain current models that consider high-field effects in scaled devices and permit tradeoffs between saturation drive current and subthreshold leakage current are employed to model CMOS circuit performance and power dissipation at low voltages. Transregional models are used in conjunction with physical short channel MOSFET threshold voltage rolloff models and stochastic interconnect distributions to project optimal supply voltages, threshold voltages, and device channel widths minimizing total power dissipated by CMOS logic circuits for each National Technology Roadmap for Semiconductors (NTRS) technology generation. Optimum supply voltage, corresponding to minimum total power dissipation, is projected to scale to 510 mV for the 50-nm 10-GHz CMOS generation in the year 2012. Techniques exploiting datapath parallelism to further scale the supply voltage are shown to offer decreasing reductions in power dissipation with technology scaling.