In this paper, we propose a system for efficient use of shared memory between CPU and GPU. The system, called Fusion Architecture, assures consistency of the shared memory and minimizes cache misses that frequently occurs on Heterogeneous System Architecture or Unified Virtual Memory based systems. It also maximizes the performance for memory intensive jobs by efficient allocation of GPU cores. To test between architectures on various scenarios, we introduce the Fusion Architecture Analyzer, which compares OpenMP, OpenCL, CUDA, and the proposed architecture in terms of memory overhead and process time. As a result, Proposed fusion architectures show that the Fusion Architecture runs benchmarks 55% faster and reduces memory overheads by 220% in average.

The emerging Spin-Transfer Torque RAM (STT-RAM) is a promising component that can be used to improve the efficiency as a result of its high storage density and low leakage power. However, the state-of-the-art STT-RAM is not ready to replace SRAM technology due to the negative effect of its write operations. The write operations require longer latency and more power than the same operations in SRAM. Therefore, a hybrid cache with SRAM and STT-RAM technologies is proposed to obtain the benefits of STT-RAM while minimizing its negative effects by using SRAM. To efficiently use of the hybrid cache, it is important to place write intensive data onto the cache. Such data should be placed on SRAM to minimize the negative effect. Thus, we propose a technique that optimizes placement of data in main memory. It drives the proper combination of advantages and disadvantages for SRAM and STT-RAM in the hybrid cache. As a result of the proposed technique, write intensive data are loaded to SRAM and read intensive data are loaded to STT-RAM. In addition, our technique also optimizes temporal locality to minimize conflict misses. Therefore, it improves performance and energy consumption of the hybrid cache architecture in a certain range.

In wireless sensor networks, increasing the sensing rate of each node to improve the data accuracy usually incurs a decrease of network lifetime. In this study, an energy-adaptive data compression scheme is proposed to efficiently control the sensing rate in an energy-harvesting wireless sensor network (WSN). In the proposed scheme, by utilizing the surplus energy effectively for the data compression, each node can increase the sensing rate without any rise of blackout time. Simulation result verifies that the proposed scheme gathers more amount of sensory data per unit time with lower number of blackout nodes than the other compression schemes for WSN.

Nonvolatile memory (NVM) is being considered as an alternative of traditional memory devices such as SRAM and DRAM, which suffer from various limitations due to the technology scaling of modern integrated circuits. Although NVMs have advantages including nonvolatility, low leakage current, and high density, their inferior write performance in terms of energy and endurance becomes a major challenge to the successful design of NVM-based memory systems. In order to overcome the aforementioned drawback of the NVM, extensive research is required to develop energy- and endurance-aware optimization techniques for NVM-based memory systems. However, researchers have experienced difficulty in finding a suitable simulation tool to prototype and evaluate new NVM optimization schemes because existing simulation tools do not consider the feature of NVM devices. In this article, we introduce a NVM-based cache simulator to support rapid prototyping and evaluation of NVM-based caches, as well as energy- and endurance-aware NVM cache optimization schemes. We demonstrate that the proposed NVM cache simulator can easily prototype PRAM cache and PRAM+STT-RAM hybrid cache as well as evaluate various write traffic reduction schemes and wear leveling schemes.

Microcontrollers (MCUs) for endpoint smart sensor devices of internet-of-thing (IoT) are being implemented as system-on-chip (SoC) with on-chip instruction flash memory, in which user firmware is embedded. MCUs directly fetch binary code-based instructions through bit-line sense amplifier (S/A) integrated with on-chip flash memory. The S/A compares bit cell current with reference current to identify which data are programmed. The S/A in reading '0' (erased) cell data consumes a large sink current, which is greater than off-current for '1' (programmed) cell data. The main motivation of our approach is to reduce the number of accesses of erased cells by binary code level transformation. This paper proposes a built-in write/read path architecture using binary code inversion method based on hot-spot region detection of instruction code access to reduce sensing current in S/A. From the profiling result of instruction access patterns, hot-spot region of an original compiled binary code is conditionally inverted with the proposed bit-inversion techniques. The de-inversion hardware only consumes small logic current instead of analog sink current in S/A and it is integrated with the conventional S/A to restore original binary instructions. The proposed techniques are applied to the fully-custom designed MCU with ARM Cortex-M0$^{TM}$ using 0.18um Magnachip Flash-embedded CMOS process and the benefits in terms of power consumption reduction are evaluated for Dhrystone$^{TM}$ benchmark. The profiling environment of instruction code executions is implemented by extending commercial ARM KEIL$^{TM}$ MDK (MCU Development Kit) with our custom-designed access analyzer.

Density of Phase-Change Memory (PCM) devices has been doubled through the employment of multi-level cell (MLC) technology. However, this doubled-capacity comes in the expense of severe performance degradation, as compared to the conventional single-level cell (SLC) PCM. This negative effect on the performance of the MLC PCM detracts from the potential benefits of the MLC PCM. This paper introduces an efficient way of minimizing the performance degradation while maximizing the capacity benefits of the MLC PCM. To this end, we propose a location-aware hybrid management of SLC and MLC in compressed PCM main memory systems. Our trace-driven simulations using real application workloads demonstrate that the proposed technique enhances the performance and energy consumption by 45.1% and 46.5%, respectively, on the average, over the conventional technique that only uses a MLC PCM.

In general, PCM (Phase Change Memory) is unsuitable as a main memory because it has limitations: high read/write latency and low endurance. However, the DRAM&PCM hybrid memory with the same level is one of the effective structures for a next generation main memory because it can utilize an advantage of both DRAM and PCM. Therefore, it needs an effective page management method for exploiting each memory characteristics dynamically and adaptively. So we aim reducing an access time and write count of PCM by using an effective page replacement. According to our simulation, the proposed algorithm for the DRAM&PCM hybrid can reduce the PCM access count by around 60% and the PCM write count by 42% given the same PCM size, compared with Clock-DWF algorithm.

NAND flash memory has unique characteristics like as 'out-place-update' and limited lifetime compared with traditional storage systems. According to out-of-place update scheme, a number of invalid (or called dead) pages can be generated. In this case, garbage collection is needed to reclaim invalid pages. Because garbage collection results in not only erase operations but also copy operations of valid (or called live) pages to other blocks, many garbage collection techniques have proposed to reduce the overhead and to increase the lifetime of NAND Flash systems. This techniques sometimes select victim blocks including cold data for the wear leveling. However, most of them overlook the cost of selecting victim blocks including cold data. In this paper, we propose a garbage collection technique named CAPi (Cost Age with Proportion of invalid pages). Considering the additional overhead of what to select victim blocks including cold data, CAPi improves the response time in garbage collection and increase the lifetime in memory systems. Additionally, the proposed scheme also improves the efficiency of garbage collection by separating cold data from hot data in valid pages. In experimental evaluation, we showed that CAPi yields up to, at maximum, 73% improvement in lifetime compared with existing garbage collections.