A key question for the microprocessor research and design community is whether scaling multicores will provide the performance and value needed to scale down many more technology generations. To provide a quantitative answer to this question, a comprehensive study that projects the speedup potential of future multicores and examines the underutilization of integration capacity-dark silicon-is timely and crucial.

/pdf/dark-silicon-and-the-end-of-multicore-scaling-fxij8qeoty.pdf

Dark Silicon and the End of Multicore Scaling

International Technology Roadmap for Semiconductors 2003の要求清浄度について － シリコンウエハ表面と雰囲気環境に要求される清浄度, 分析方法の現状について －

Since 2005, processor designers have increased core counts to exploit Moore's Law scaling, rather than focusing on single-core performance. The failure of Dennard scaling, to which the shift to multicore parts is partially a response, may soon limit multicore scaling just as single-core scaling has been curtailed. This paper models multicore scaling limits by combining device scaling, single-core scaling, and multicore scaling to measure the speedup potential for a set of parallel workloads for the next five technology generations. For device scaling, we use both the ITRS projections and a set of more conservative device scaling parameters. To model single-core scaling, we combine measurements from over 150 processors to derive Pareto-optimal frontiers for area/performance and power/performance. Finally, to model multicore scaling, we build a detailed performance model of upper-bound performance and lower-bound core power. The multicore designs we study include single-threaded CPU-like and massively threaded GPU-like multicore chip organizations with symmetric, asymmetric, dynamic, and composed topologies. The study shows that regardless of chip organization and topology, multicore scaling is power limited to a degree not widely appreciated by the computing community. Even at 22 nm (just one year from now), 21% of a fixed-size chip must be powered off, and at 8 nm, this number grows to more than 50%. Through 2024, only 7.9x average speedup is possible across commonly used parallel workloads, leaving a nearly 24-fold gap from a target of doubled performance per generation.

/pdf/dark-silicon-and-the-end-of-multicore-scaling-goghgnng6o.pdf

Dark silicon and the end of multicore scaling

Processing-in-memory (PIM) is a promising solution to address the "memory wall" challenges for future computer systems. Prior proposed PIM architectures put additional computation logic in or near memory. The emerging metal-oxide resistive random access memory (ReRAM) has showed its potential to be used for main memory. Moreover, with its crossbar array structure, ReRAM can perform matrix-vector multiplication efficiently, and has been widely studied to accelerate neural network (NN) applications. In this work, we propose a novel PIM architecture, called PRIME, to accelerate NN applications in ReRAM based main memory. In PRIME, a portion of ReRAM crossbar arrays can be configured as accelerators for NN applications or as normal memory for a larger memory space. We provide microarchitecture and circuit designs to enable the morphable functions with an insignificant area overhead. We also design a software/hardware interface for software developers to implement various NNs on PRIME. Benefiting from both the PIM architecture and the efficiency of using ReRAM for NN computation, PRIME distinguishes itself from prior work on NN acceleration, with significant performance improvement and energy saving. Our experimental results show that, compared with a state-of-the-art neural processing unit design, PRIME improves the performance by ~2360× and the energy consumption by ~895×, across the evaluated machine learning benchmarks.

PRIME: a novel processing-in-memory architecture for neural network computation in ReRAM-based main memory

This paper investigates the problem of partitioning a shared cache between multiple concurrently executing applications. The commonly used LRU policy implicitly partitions a shared cache on a demand basis, giving more cache resources to the application that has a high demand and fewer cache resources to the application that has a low demand. However, a higher demand for cache resources does not always correlate with a higher performance from additional cache resources. It is beneficial for performance to invest cache resources in the application that benefits more from the cache resources rather than in the application that has more demand for the cache resources. This paper proposes utility-based cache partitioning (UCP), a low-overhead, runtime mechanism that partitions a shared cache between multiple applications depending on the reduction in cache misses that each application is likely to obtain for a given amount of cache resources. The proposed mechanism monitors each application at runtime using a novel, cost-effective, hardware circuit that requires less than 2kB of storage. The information collected by the monitoring circuits is used by a partitioning algorithm to decide the amount of cache resources allocated to each application. Our evaluation, with 20 multiprogrammed workloads, shows that UCP improves performance of a dual-core system by up to 23% and on average 11% over LRU-based cache partitioning.

/pdf/utility-based-cache-partitioning-a-low-overhead-high-4t50s5xud2.pdf

Utility-Based Cache Partitioning: A Low-Overhead, High-Performance, Runtime Mechanism to Partition Shared Caches

Next generation byte addressable nonvolatile memories (NVMs) such as PCM, Memristor, and 3D X-Point are attractive solutions for mobile and other end-user devices, as they offer memory scalability as well as fast persistent storage. However, NVM's limitations of slow writes and high write energy are magnified for applications that require atomic, consistent, isolated and durable (ACID) persistence. For maintaining ACID persistence guarantees, applications not only need to do extra writes to NVM but also need to execute a significant number of additional CPU instructions for performing NVM writes in a transactional manner. Our analysis shows that maintaining persistence with ACID guarantees increases CPU energy up to 7.3x and NVM energy up to 5.1x compared to a baseline with no ACID guarantees. For computing platforms such as mobile devices, where energy consumption is a critical factor, it is important that the energy cost of persistence is reduced.To address the energy overheads of persistence with ACID guarantees, we develop novel energy-aware persistence (EAP) principles that identify data durability (logging) as the dominant factor in energy increase. Next, for low energy states, we formulate energy efficient durability techniques that include a mechanism to switch between performance and energy efficient logging modes, support for NVM group commit, and a memory management method that reduces energy by trading capacity via less frequent garbage collection. For critical energy states, we propose a relaxed durability mechanism -- ACI-RD -- that relaxes data logging without affecting the correctness of an application. Finally, we evaluate EAP's principles with real applications and benchmarks. Our experimental results demonstrate up to 2x reduction in CPU and 2.4x reduction in NVM energy usage compared to the traditional ACID persistence.

Energy Aware Persistence: Reducing Energy Overheads of Memory-based Persistence in NVMs

Row-Hammer (RH) is a DRAM data-disturbance failure that occurs when a row is activated frequently, which causes bit-flips in nearby rows. Row-Hammer is a significant security threat as an attacker can exploit the bit-flips to do privilege escalation and leak confidential data. While several solutions aim to mitigate RH, such solutions depend on the RH threshold and adversarial access patterns. Solutions developed for a given threshold become ineffective for newer devices with lower thresholds, and new attack patterns continue to break existing mitigations. Currently, there is no guaranteed solution for RH, which means that the system remains vulnerable to security threats even in the presence of RH mitigation.In this paper, we contend that simply relying on RH mitigation is insufficient to provide security in the presence of reducing threshold and motivated attackers. We propose SafeGuard, which equips the system with low-cost integrity protection as a defense against potential attacks that break the RH mitigation. As SafeGuard can detect arbitrary failures, it converts the problem of RH bit-flips from a security threat (silent consumption of corrupted data) to a reliability concern (detectable uncorrectable errors caused by integrity violation). We develop SafeGuard for systems that employ ECC modules and show that SafeGuard can provide strong detection to both SECDED (46-bit MAC per cache-line) and Chipkill (32-bit MAC per cache-line) while retaining the correction capability of conventional designs. SafeGuard avoids incurring any storage overheads in DRAM by simply reorganizing the ECC code to operate at a cache-line granularity (64 bytes) instead of a word granularity (8 bytes). Our evaluations show that SafeGuard has a negligible impact on both the system performance (0.7%) and the system reliability due to naturally occurring errors while still providing a strong defense against the security risk of RH by detecting arbitrary bit-flips.

SafeGuard: Reducing the Security Risk from Row-Hammer via Low-Cost Integrity Protection

A computer cache memory organization called Probabilistic Set Associative Cache (PAC) has the hardware complexity and latency of a direct-mapped cache but functions as a set-associative cache for a fraction of the time, thus yielding better than direct mapped cache hit rates. The organization is considered a (1+P)—way set associative cache, where the chosen parameter called Override Probability P determines the average associativity, for example, for P=0.1, effectively it operates as if a 1.1-way set associative cache.

Probabilistic associative cache

Quantum computers with hundreds of qubits will be available soon. Unfortunately, high device error-rates pose a significant challenge in using these near-term quantum systems to power real-world applications. Executing a program on existing quantum systems generates both correct and incorrect outcomes, but often, the output distribution is too noisy to distinguish between them. In this paper, we show that erroneous outcomes are not arbitrary but exhibit a well-defined structure when represented in the Hamming space. Our experiments on IBM and Google quantum computers show that the most frequent erroneous outcomes are more likely to be close in the Hamming space to the correct outcome. We exploit this behavior to improve the ability to infer the correct outcome. We propose Hamming Reconstruction (HAMMER), a post-processing technique that leverages the observation of Hamming behavior to reconstruct the noisy output distribution, such that the resulting distribution has higher fidelity. We evaluate HAMMER using experimental data from Google and IBM quantum computers with more than 500 unique quantum circuits and obtain an average improvement of 1.37x in the quality of solution. On Google’s publicly available QAOA datasets, we show that HAMMER sharpens the gradients on the cost function landscape.

HAMMER: boosting fidelity of noisy Quantum circuits by exploiting Hamming behavior of erroneous outcomes

Conventionally, systems have relied on technology scaling to provide smaller cells, which helps in increasing the capacity of on-chip and off-chip structures. Unfortunately, scaling technology to smaller nodes causes increased susceptibility to faults. We study the problem of efficiently tolerating transient failures using scalable Spin-Transfer Torque RAM (STTRAM) as an example. At smaller feature sizes, the energy required to flip a STTRAM cell reduces, which makes these cells more susceptible to random failures caused by thermal noise. Such failures can be tolerated by periodic scrubbing and provisioning each line with Error Correction Code (ECC). However, to tolerate the desired bit-error-rate, the cache needs ECC-6 (six bit error correction) per line, incurring impractical storage overheads. Ideally, we want to tolerate these faults without relying on multi-bit ECC. We propose SuDoku, a design that provisions each line with ECC-1 and a strong error detection code, and relies on a region-based RAID-4 to perform correction of multi-bit errors. Unfortunately, simply having such a RAID-4 based architecture is ineffective at tolerating a high-rate of transient faults and provides an MTTF in the order of only a few seconds. We describe a novel data resurrection scheme that can repair multiple faulty lines in a RAID-4 region to increase the MTTF to several hours. We propose an extension of SuDoku, which hashes a given line into two regions of RAID-4 to significantly enhance reliability and increase the MTTF to trillions of hours. Our evaluations show that SuDoku provides 874x higher reliability than ECC-6, incurs 30% less storage than ECC-6, and performs within 0.1% of an ideal fault-free baseline.

Moinuddin K. Qureshi

Papers

Energy Aware Persistence: Reducing Energy Overheads of Memory-based Persistence in NVMs

SafeGuard: Reducing the Security Risk from Row-Hammer via Low-Cost Integrity Protection

Probabilistic associative cache

HAMMER: boosting fidelity of noisy Quantum circuits by exploiting Hamming behavior of erroneous outcomes

SuDoku: Tolerating High-Rate of Transient Failures for Enabling Scalable STTRAM