Publications by Saugata Ghose

@inproceedings{abc,
	abstract = {We are experiencing an explosive growth in the number of consumer devices, including smartphones, tablets, web-based computers such as Chromebooks, and wearable devices. For this class of devices, energy efficiency is a first-class concern due to the limited battery capacity and thermal power budget. We find that data movement is a major contributor to the total system energy and execution time in consumer devices. The energy and performance costs of moving data between the memory system and the compute units are significantly higher than the costs of computation. As a result, addressing data movement is crucial for consumer devices. In this work, we comprehensively analyze the energy and performance impact of data movement for several widely-used Google consumer workloads: (1) the Chrome web browser; (2) TensorFlow Mobile, Google{\textquoteright}s machine learning framework; (3) video playback, and (4) video capture, both of which are used in many video services such as YouTube and Google Hangouts. We find that processing-inmemory (PIM) can significantly reduce data movement for all of these workloads, by performing part of the computation close to memory. Each workload contains simple primitives and functions that contribute to a significant amount of the overall data movement. We investigate whether these primitives and functions are feasible to implement using PIM, given the limited area and power constraints of consumer devices. Our analysis shows that offloading these primitives to PIM logic, consisting of either simple cores or specialized accelerators, eliminates a large amount of data movement, and significantly reduces total system energy (by an average of 55.4\% across the workloads) and execution time (by an average of 54.2\%).},
	author = {Amirali Boroumand and Saugata Ghose and Youngsok Kim and Rachata Ausavarungnirun and Eric Shiu and Rahul Thakur and Dae-Hyun Kim and Aki Kuusela and Allan Knies and Parthasarathy Ranganathan and Onur Mutlu},
	booktitle = {Proceedings of the 23rd International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS)},
	title = {Google Workloads for Consumer Devices: Mitigating Data Movement Bottlenecks},
	venue = {Williamsburg, VA, USA},
	year = {2018}
}

@inproceedings{abc,
	abstract = {Graphics Processing Units (GPUs) exploit large amounts of threadlevel parallelism to provide high instruction throughput and to efficiently hide long-latency stalls. The resulting high throughput, along with continued programmability improvements, have made GPUs an essential computational resource in many domains. Applications from different domains can have vastly different compute and memory demands on the GPU. In a large-scale computing environment, to efficiently accommodate such wide-ranging demands without leaving GPU resources underutilized, multiple applications can share a single GPU, akin to how multiple applications execute concurrently on a CPU. Multi-application concurrency requires several support mechanisms in both hardware and software. One such key mechanism is virtual memory, which manages and protects the address space of each application. However, modern GPUs lack the extensive support for multi-application concurrency available in CPUs, and as a result suffer from high performance overheads when shared by multiple applications, as we demonstrate. We perform a detailed analysis of which multi-application concurrency support limitations hurt GPU performance the most. We find that the poor performance is largely a result of the virtual memory mechanisms employed in modern GPUs. In particular, poor address translation performance is a key obstacle to efficient GPU sharing. State-of-the-art address translation mechanisms, which were designed for single-application execution, experience significant inter-application interference when multiple applications spatially share the GPU. This contention leads to frequent misses in the shared translation lookaside buffer (TLB), where a single miss can induce long-latency stalls for hundreds of threads. As a result, the GPU often cannot schedule enough threads to successfully hide the stalls, which diminishes system throughput and becomes a first-order performance concern. Based on our analysis, we propose MASK, a new GPU framework that provides low-overhead virtual memory support for the concurrent execution of multiple applications. MASK consists of three novel address-translation-aware cache and memory management mechanisms that work together to largely reduce the overhead of address translation: (1) a token-based technique to reduce TLB contention, (2) a bypassing mechanism to improve the effectiveness of cached address translations, and (3) an application-aware memory scheduling scheme to reduce the interference between address translation and data requests. Our evaluations show that MASK restores much of the throughput lost to TLB contention. Relative to a state-of-the-art GPU TLB, MASK improves system throughput by 57.8\%, improves IPC throughput by 43.4\%, and reduces applicationlevel unfairness by 22.4\%. MASK{\textquoteright}s system throughput is within 23.2\% of an ideal GPU system with no address translation overhead.},
	author = {Rachata Ausavarungnirun and Vance Miller and Joshua Landgraf and Saugata Ghose and Jayneel Gandhi and Adwait Jog and Christopher Rossbach and Onur Mutlu},
	booktitle = {Proceedings of the 23rd International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS)},
	title = {MASK: Redesigning the GPU Memory Hierarchy to Support Multi-Application Concurrency},
	venue = {Williamsburg, VA, USA},
	year = {2018}
}

@inproceedings{abc,
	abstract = {NAND flash memory density continues to scale to keep up with the increasing storage demands of data-intensive applications. Unfortunately, as a result of this scaling, the lifetime of NAND flash memory has been decreasing. Each cell in NAND flash memory can endure only a limited number of writes, due to the damage caused by each program and erase operation on the cell. This damage can be partially repaired on its own during the idle time between program or erase operations (known as the dwell time), via a phenomenon known as the self-recovery effect. Prior works study the self-recovery effect for planar (i.e., 2D) NAND flash memory, and propose to exploit it to improve flash lifetime, by applying high temperature to accelerate selfrecovery. However, these findings may not be directly applicable to 3D NAND flash memory, due to significant changes in the design and manufacturing process that are required to enable practical 3D stacking for NAND flash memory. In this paper, we perform the first detailed experimental characterization of the effects of self-recovery and temperature on real, state-of-the-art 3D NAND flash memory devices. We show that these effects influence two major factors of NAND flash memory reliability: (1) retention loss speed (i.e., the speed at which a flash cell leaks charge), and (2) program variation (i.e., the difference in programming speed across flash cells). We find that self-recovery and temperature affect 3D NAND flash memory quite differently than they affect planar NAND flash memory, rendering prior models of self-recovery and temperature ineffective for 3D NAND flash memory. Using our characterization results, we develop a new model for 3D NAND flash memory reliability, which predicts how retention, wearout, self-recovery, and temperature affect raw bit error rates and cell threshold voltages. We show that our model is accurate, with an error of only 4.9\%. Based on our experimental findings and our model, we propose HeatWatch, a new mechanism to improve 3D NAND flash memory reliability. The key idea of HeatWatch is to optimize the read reference voltage, i.e., the voltage applied to the cell during a read operation, by adapting it to the dwell time of the workload and the current operating temperature. HeatWatch (1) efficiently tracks flash memory temperature and dwell time online, (2) sends this information to our reliability model to predict the current voltages of flash cells, and (3) predicts the optimal read reference voltage based on the current cell voltages. Our detailed experimental evaluations show that HeatWatch improves flash lifetime by 3.85{\texttimes} over a baseline that uses a fixed read reference voltage, averaged across 28 real storage workload traces, and comes within 0.9\% of the lifetime of an ideal read reference voltage selection mechanism.},
	author = {Yixin Luo and Saugata Ghose and Yu Cai and Erich F. Haratsch and Onur Mutlu},
	booktitle = {Proceedings of the 24th International Symposium on High-Performance Computer Architecture (HPCA)},
	title = {HeatWatch: Improving 3D NAND Flash Memory Device Reliability by Exploiting Self-Recovery and Temperature-Awareness},
	venue = {Vienna, Austria},
	year = {2018}
}

@inproceedings{abc,
	abstract = {Solid-state drives (SSDs) are used in a wide array of computer systems today, including in datacenters and enterprise servers. As the I/O demands of these systems continue to increase, manufacturers are evolving SSD architectures to keep up with this demand. For example, manufacturers have introduced new high-bandwidth interfaces to replace the conventional SATA host-interface protocol. These new interfaces, such as the NVMe protocol, are designed specifically to enable the high amounts of concurrent I/O bandwidth that SSDs are capable of delivering.

While modern SSDs with sophisticated features such as the NVMe protocol are already on the market, existing SSD simulation tools have fallen behind, as they do not capture these new features. We find that state-of-the-art SSD simulators have three shortcomings that prevent them from accurately modeling the performance of real off-the-shelf SSDs. First, these simulators do not model critical features of new protocols (e.g., NVMe), such as their use of multiple application-level queues for requests and the elimination of OS intervention for I/O request processing. Second, these simulators often do not accurately capture the impact of advanced SSD maintenance algorithms (e.g., garbage collection), as they do not properly or quickly emulate steady-state conditions that can significantly change the behavior of these algorithms in real SSDs. Third, these simulators do not capture the full end-to-end latency of I/O requests, which can incorrectly skew the results reported for SSDs that make use of emerging non-volatile memory technologies. By not accurately modeling these three features, existing simulators report results that deviate significantly from real SSD performance.

In this work, we introduce a new simulator, called MQSim, that accurately models the performance of both modern SSDs and conventional SATA-based SSDs. MQSim faithfully models new high-bandwidth protocol implementations, steady-state SSD conditions, and the full end-to-end latency of requests in modern SSDs. We validate MQSim, showing that it reports performance results that are only 6\%-18\% apart from the measured actual performance of four real state-of-the-art SSDs. We show that by modeling critical features of modern SSDs, MQSim uncovers several real and important issues that were not captured by existing simulators, such as the performance impact of inter-flow interference. We have released MQSim as an open-source tool, and we hope that it can enable researchers to explore directions in new and different areas.},
	author = {Arash Tavakkol and Juan Gomez-Luna and Mohammad Sadrosadati and Saugata Ghose and Onur Mutlu},
	booktitle = {Proceedings of the 16th USENIX Conference on File and Storage Technologies},
	title = {MQsim: a framework for enabling realistic studies of modern multi-queue SSD devices},
	venue = {Oakland, CA, USA},
	year = {2018}
}

@inproceedings{abc,
	abstract = {Contemporary discrete GPUs support rich memory management features such as virtual memory and demand paging. These features simplify GPU programming by providing a virtual address space abstraction similar to CPUs and eliminating manual memory management, but they introduce high performance overheads during (1) address translation and (2) page faults. A GPU relies on high degrees of thread-level parallelism (TLP) to hide memory latency. Address translation can undermine TLP, as a single miss in the translation lookaside buffer (TLB) invokes an expensive serialized page table walk that often stalls multiple threads. Demand paging can also undermine TLP, as multiple threads often stall while they wait for an expensive data transfer over the system I/O (e.g., PCIe) bus when the GPU demands a page.

In modern GPUs, we face a trade-off on how the page size used for memory management affects address translation and demand paging. The address translation overhead is lower when we employ a larger page size (e.g., 2MB large pages, compared with conventional 4KB base pages), which increases TLB coverage and thus reduces TLB misses. Conversely, the demand paging overhead is lower when we employ a smaller page size, which decreases the system I/O bus transfer latency. Support for multiple page sizes can help relax the page size trade-off so that address translation and demand paging optimizations work together synergistically. However, existing page coalescing (i.e., merging base pages into a large page) and splintering (i.e., splitting a large page into base pages) policies require costly base page migrations that undermine the benefits multiple page sizes provide. In this paper, we observe that GPGPU applications present an opportunity to support multiple page sizes without costly data migration, as the applications perform most of their memory allocation en masse (i.e., they allocate a large number of base pages at once). We show that this en masse allocation allows us to create intelligent memory allocation policies which ensure that base pages that are contiguous in virtual memory are allocated to contiguous physical memory pages. As a result, coalescing and splintering operations no longer need to migrate base pages.

We introduce Mosaic, a GPU memory manager that provides application-transparent support for multiple page sizes. Mosaic uses base pages to transfer data over the system I/O bus, and allocates physical memory in a way that (1) preserves base page contiguity and (2) ensures that a large page frame contains pages from only a single memory protection domain. We take advantage of this allocation strategy to design a novel in-place page size selection mechanism that avoids data migration. This mechanism allows the TLB to use large pages, reducing address translation overhead. During data transfer, this mechanism enables the GPU to transfer only the base pages that are needed by the application over the system I/O bus, keeping demand paging overhead low. Our evaluations show that Mosaic reduces address translation overheads while efficiently achieving the benefits of demand paging, compared to a contemporary GPU that uses only a 4KB page size. Relative to a state-of-the-art GPU memory manager, Mosaic improves the performance of homogeneous and heterogeneous multi-application workloads by 55.5\% and 29.7\% on average, respectively, coming within 6.8\% and 15.4\% of the performance of an ideal TLB where all TLB requests are hits.},
	author = {Rachata Ausavarungnirun and Joshua Landgraf and Vance Miller and Saugata Ghose and Jayneel Gandhi and Christopher Rossbach and },
	booktitle = {Proceedings of the 50th Annual IEEE/ACM International Symposium on Microarchitecture},
	title = {Mosaic: a GPU memory manager with application-transparent support for multiple page sizes},
	venue = {Cambridge, MA, USA},
	year = {2017}
}

@inproceedings{abc,
	abstract = {While the memory footprints of cloud and HPC applications continue to increase, fundamental issues with DRAM scaling are likely to prevent traditional main memory systems, composed of monolithic DRAM, from greatly growing in capacity. Hybrid memory systems can mitigate the scaling limitations of monolithic DRAM by pairing together multiple memory technologies (e.g., different types of DRAM, or DRAM and non-volatile memory) at the same level of the memory hierarchy. The goal of a hybrid main memory is to combine the different advantages of the multiple memory types in a cost-effective manner while avoiding the disadvantages of each technology. Memory pages are placed in and migrated between the different memories within a hybrid memory system, based on the properties of each page. It is important to make intelligent page management (i.e., placement and migration) decisions, as they can significantly affect system performance.In this paper, we propose utility-based hybrid memory management (UH-MEM), a new page management mechanism for various hybrid memories, that systematically estimates the utility (i.e., the system performance benefit) of migrating a page between different memory types, and uses this information to guide data placement. UH-MEM operates in two steps. First, it estimates how much a single application would benefit from migrating one of its pages to a different type of memory, by comprehensively considering access frequency, row buffer locality, and memory-level parallelism. Second, it translates the estimated benefit of a single application to an estimate of the overall system performance benefit from such a migration.We evaluate the effectiveness of UH-MEM with various types of hybrid memories, and show that it significantly improves system performance on each of these hybrid memories. For a memory system with DRAM and non-volatile memory, UH-MEM improves performance by 14\% on average (and up to 26\%) compared to the best of three evaluated state-of-the-art mechanisms across a large number of data-intensive workloads.},
	author = {Yang Li and Saugata Ghose and Jongmoo Choi and Jin Sun and Hui Wang and Onur Mutlu},
	booktitle = {Proceedins of the 2017 IEEE International Conference on Cluster Computing (CLUSTER)},
	title = {Utility-Based Hybrid Memory Management},
	venue = {Honolulu, HI, USA},
	year = {2017}
}

@inproceedings{abc,
	author = {Xi-Yue Xiang and Wentao Shi and Saugata Ghose and Lu Peng and Onur Mutlu and Nian-Feng Tzeng},
	booktitle = {Proceedings of the International Conference on Supercomputing, ICS 2017, Chicago, IL, USA},
	title = {Carpool: a bufferless on-chip network supporting adaptive multicast and hotspot alleviation.},
	url = {http://doi.acm.org/10.1145/3079079.3079090},
	year = {2017}
}

@inproceedings{abc,
	author = {Kevin K. Chang and Abdullah Giray Yaglik{\c c}i and Saugata Ghose and Aditya Agrawal and Niladrish Chatterjee and Abhijith Kashyap and Donghyuk Lee and Mike O{\textquoteright}Connor and Hasan Hassan and Onur Mutlu},
	booktitle = {Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems, Urbana-Champaign, IL, USA},
	title = {Understanding Reduced-Voltage Operation in Modern DRAM Devices: Experimental Characterization, Analysis, and Mechanisms.},
	url = {http://doi.acm.org/10.1145/3078505.3078590},
	year = {2017}
}

@inproceedings{abc,
	author = {Donghyuk Lee and Samira Manabi Khan and Lavanya Subramanian and Saugata Ghose and Rachata Ausavarungnirun and Gennady Pekhimenko and Vivek Seshadri and Onur Mutlu},
	booktitle = {Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems, Urbana-Champaign, IL, USA},
	title = {Design-Induced Latency Variation in Modern DRAM Chips: Characterization, Analysis, and Latency Reduction Mechanisms.},
	url = {http://doi.acm.org/10.1145/3078505.3078533},
	year = {2017}
}

@inproceedings{abc,
	abstract = {Digital forensic investigators often need to extract data from a seized device that contains NAND flash memory. Many such devices are physically damaged, preventing investigators from using automated techniques to extract the data stored within the device. Instead, investigators turn to chip-off analysis, where they use a thermal-based procedure to physically remove the NAND flash memory chip from the device, and access the chip directly to extract the raw data stored on the chip.

We perform an analysis of the errors introduced into multi-level cell (MLC) NAND flash memory chips after the device has been seized. We make two major observations. First, between the time that a device is seized and the time digital forensic investigators perform data extraction, a large number of errors can be introduced as a result of charge leakage from the cells of the NAND flash memory (known as data retention errors). Second, when thermal-based chip removal is performed, the number of errors in the data stored within NAND flash memory can increase by two or more orders of magnitude, as the high temperature applied to the chip greatly accelerates charge leakage. We demonstrate that the chip-off analysis based forensic data recovery procedure is quite destructive, and can often render most of the data within NAND flash memory uncorrectable, and, thus, unrecoverable.

To mitigate the errors introduced during the forensic recovery process, we explore a new hardware- based approach. We exploit a fine-grained read reference voltage control mechanism implemented in modern NAND flash memory chips, called read-retry, which can compensate for the charge leakage that occurs due to (1) retention loss and (2) thermal-based chip removal. The read-retry mechanism successfully reduces the number of errors, such that the original data can be fully recovered in our tested chips as long as the chips were not heavily used prior to seizure. We conclude that the read-retry mechanism should be adopted as part of the forensic data recovery process. },
	author = {Aya Fukami and Saugata Ghose and Yixin Luo and Yu Cai and Onur Mutlu},
	booktitle = {Proceedings of the 2017 Digital Forensics Conference},
	title = {Improving the Reliability of Chip-Off Forensic Analysis of NAND Flash Memory Devices},
	venue = {{\"U}berlingen, Germany},
	year = {2017}
}

@inproceedings{abc,
	author = {Hasan Hassan and Nandita Vijaykumar and Samira Manabi Khan and Saugata Ghose and Kevin K. Chang and Gennady Pekhimenko and Donghyuk Lee and Oguz Ergin and Onur Mutlu},
	booktitle = {2017 IEEE International Symposium on High Performance Computer Architecture, HPCA 2017, Austin, TX, USA},
	title = {SoftMC: A Flexible and Practical Open-Source Infrastructure for Enabling Experimental DRAM Studies.},
	url = {https://doi.org/10.1109/HPCA.2017.62},
	year = {2017}
}

@inproceedings{abc,
	author = {Yu Cai and Saugata Ghose and Yixin Luo and Ken Mai and Onur Mutlu and Erich F. Haratsch},
	booktitle = {2017 IEEE International Symposium on High Performance Computer Architecture, HPCA 2017, Austin, TX, USA},
	title = {Vulnerabilities in MLC NAND Flash Memory Programming: Experimental Analysis, Exploits, and Mitigation Techniques.},
	url = {https://doi.org/10.1109/HPCA.2017.61},
	year = {2017}
}

@article{abc,
	author = {Yixin Luo and Saugata Ghose and Tianshi Li and Sriram Govindan and Bikash Sharma and Bryan Kelly and Amirali Boroumand and Onur Mutlu},
	journal = {CoRR},
	title = {Using ECC DRAM to Adaptively Increase Memory Capacity.},
	url = {http://arxiv.org/abs/1706.08870},
	year = {2017}
}

@article{abc,
	author = {Amirali Boroumand and Saugata Ghose and Minesh Patel and Hasan Hassan and Brandon Lucia and Nastaran Hajinazar and Kevin Hsieh and Krishna T. Malladi and Hongzhong Zheng and Onur Mutlu},
	journal = {CoRR},
	title = {LazyPIM: Efficient Support for Cache Coherence in Processing-in-Memory Architectures.},
	url = {http://arxiv.org/abs/1706.03162},
	year = {2017}
}

@inproceedings{abc,
	author = {Kevin K. Chang and Abdullah Giray Yaglik{\c c}i and Saugata Ghose and Aditya Agrawal and Niladrish Chatterjee and Abhijith Kashyap and Donghyuk Lee and Mike O{\textquoteright}Connor and Hasan Hassan and Onur Mutlu},
	booktitle = {CoRR},
	title = {Understanding Reduced-Voltage Operation in Modern DRAM Chips: Characterization, Analysis, and Mechanisms.},
	url = {http://arxiv.org/abs/1705.10292},
	year = {2017}
}

@article{abc,
	author = {Kevin K. Chang and A. Giray Yaalik{\c c}i and Saugata Ghose and Aditya Agrawal and Niladrish Chatterjee and Abhijith Kashyap and Donghyuk Lee and Mike O{\textquoteright}Connor and Hasan Hassan and Onur Mutlu},
	journal = {POMACS},
	title = {Understanding Reduced-Voltage Operation in Modern DRAM Devices: Experimental Characterization, Analysis, and Mechanisms.},
	url = {http://doi.acm.org/10.1145/3084447},
	year = {2017}
}

@article{abc,
	author = {Donghyuk Lee and Samira Manabi Khan and Lavanya Subramanian and Saugata Ghose and Rachata Ausavarungnirun and Gennady Pekhimenko and Vivek Seshadri and Onur Mutlu},
	journal = {POMACS},
	title = {Design-Induced Latency Variation in Modern DRAM Chips: Characterization, Analysis, and Latency Reduction Mechanisms.},
	url = {http://doi.acm.org/10.1145/3084464},
	year = {2017}
}

@inproceedings{abc,
	author = {Amirali Boroumand and Saugata Ghose and Minesh Patel and Hasan Hassan and Brandon Lucia and Kevin Hsieh and Krishna T. Malladi and Hongzhong Zheng and Onur Mutlu},
	booktitle = {Computer Architecture Letters},
	title = {LazyPIM: An Efficient Cache Coherence Mechanism for Processing-in-Memory.},
	url = {https://doi.org/10.1109/LCA.2016.2577557},
	year = {2017}
}

@article{abc,
	author = {Yu Cai and Saugata Ghose and Erich F. Haratsch and Yixin Luo and Onur Mutlu},
	journal = {CoRR},
	title = {Error Characterization, Mitigation, and Recovery in Flash Memory Based Solid-State Drives.},
	url = {http://arxiv.org/abs/1706.08642},
	year = {2017}
}

@inproceedings{abc,
	author = {Nandita Vijaykumar and Kevin Hsieh and Gennady Pekhimenko and Samira Manabi Khan and Ashish Shrestha and Saugata Ghose and Adwait Jog and Phillip B. Gibbons and Onur Mutlu},
	booktitle = {49th Annual IEEE/ACM International Symposium on Microarchitecture, MICRO 2016, Taipei, Taiwan},
	title = {Zorua: A holistic approach to resource virtualization in GPUs.},
	url = {http://dx.doi.org/10.1109/MICRO.2016.7783718},
	year = {2016}
}

@inproceedings{abc,
	author = {Kevin Hsieh and Samira Manabi Khan and Nandita Vijaykumar and Kevin K. Chang and Amirali Boroumand and Saugata Ghose and Onur Mutlu},
	booktitle = {34th IEEE International Conference on Computer Design, ICCD 2016, Scottsdale, AZ, USA},
	title = {Accelerating pointer chasing in 3D-stacked memory: Challenges, mechanisms, evaluation.},
	url = {http://dx.doi.org/10.1109/ICCD.2016.7753257},
	year = {2016}
}

@inproceedings{abc,
	author = {Xi-Yue Xiang and Saugata Ghose and Onur Mutlu and Nian-Feng Tzeng},
	booktitle = {34th IEEE International Conference on Computer Design, ICCD 2016, Scottsdale, AZ, USA},
	title = {A model for Application Slowdown Estimation in on-chip networks and its use for improving system fairness and performance.},
	url = {http://dx.doi.org/10.1109/ICCD.2016.7753327},
	year = {2016}
}

@inproceedings{abc,
	author = {Kevin K. Chang and Abhijith Kashyap and Hasan Hassan and Saugata Ghose and Kevin Hsieh and Donghyuk Lee and Tianshi Li and Gennady Pekhimenko and Samira Manabi Khan and Onur Mutlu},
	booktitle = {Proceedings of the 2016 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Science, Antibes Juan-Les-Pins, France},
	title = {Understanding Latency Variation in Modern DRAM Chips: Experimental Characterization, Analysis, and Optimization.},
	url = {http://doi.acm.org/10.1145/2896377.2901453},
	year = {2016}
}

@inproceedings{abc,
	author = {Yang Li and Di Wang and Saugata Ghose and Jie Liu and Sriram Govindan and Sean James and Eric Peterson and John Siegler and Rachata Ausavarungnirun and Onur Mutlu},
	booktitle = {2016 IEEE International Symposium on High Performance Computer Architecture, HPCA 2016, Barcelona, Spain},
	title = {SizeCap: Efficiently handling power surges in fuel cell powered data centers.},
	url = {http://dx.doi.org/10.1109/HPCA.2016.7446085},
	year = {2016}
}

@inproceedings{abc,
	author = {Kevin K. Chang and Prashant J. Nair and Donghyuk Lee and Saugata Ghose and Moinuddin K. Qureshi and Onur Mutlu},
	booktitle = {2016 IEEE International Symposium on High Performance Computer Architecture, HPCA 2016, Barcelona, Spain},
	title = {Low-Cost Inter-Linked Subarrays (LISA): Enabling fast inter-subarray data movement in DRAM.},
	url = {http://dx.doi.org/10.1109/HPCA.2016.7446095},
	year = {2016}
}

@inproceedings{abc,
	author = {Donghyuk Lee and Saugata Ghose and Gennady Pekhimenko and Samira Manabi Khan and Onur Mutlu},
	booktitle = {TACO},
	title = {Simultaneous Multi-Layer Access: Improving 3D-Stacked Memory Bandwidth at Low Cost.},
	url = {http://doi.acm.org/10.1145/2832911},
	year = {2016}
}

@article{abc,
	abstract = {Modern Graphics Processing Units (GPUs) are well provisioned to support the concurrent execution of thousands of threads. Unfortunately, different bottlenecks during execution and heterogeneous application requirements create imbalances in utilization of resources in the cores. For example, when a GPU is bottlenecked by the available off-chip memory bandwidth, its computational resources are often overwhelmingly idle, waiting for data from memory to arrive. 
This work describes the Core-Assisted Bottleneck Acceleration (CABA) framework that employs idle on-chip resources to alleviate different bottlenecks in GPU execution. CABA provides flexible mechanisms to automatically generate "assist warps" that execute on GPU cores to perform specific tasks that can improve GPU performance and efficiency. 
CABA enables the use of idle computational units and pipelines to alleviate the memory bandwidth bottleneck, e.g., by using assist warps to perform data compression to transfer less data from memory. Conversely, the same framework can be employed to handle cases where the GPU is bottlenecked by the available computational units, in which case the memory pipelines are idle and can be used by CABA to speed up computation, e.g., by performing memoization using assist warps. 
We provide a comprehensive design and evaluation of CABA to perform effective and flexible data compression in the GPU memory hierarchy to alleviate the memory bandwidth bottleneck. Our extensive evaluations show that CABA, when used to implement data compression, provides an average performance improvement of 41.7\% (as high as 2.6X) across a variety of memory-bandwidth-sensitive GPGPU applications.},
	author = {Nandita Vijaykumar and Gennady Pekhimenko and Adwait Jog and Saugata Ghose and Abhishek Bhowmick and Rachata Ausavarungnirun and Chita R. Das and Mahmut T. Kandemir and Todd C. Mowry and Onur Mutlu},
	journal = {CoRR},
	title = {A Framework for Accelerating Bottlenecks in GPU Execution with Assist Warps.},
	url = {http://arxiv.org/abs/1602.01348},
	year = {2016}
}

@inproceedings{abc,
	author = {Yixin Luo and Saugata Ghose and Yu Cai and Erich F. Haratsch and Onur Mutlu},
	booktitle = {IEEE Journal on Selected Areas in Communications},
	title = {Enabling Accurate and Practical Online Flash Channel Modeling for Modern MLC NAND Flash Memory.},
	url = {http://dx.doi.org/10.1109/JSAC.2016.2603608},
	year = {2016}
}

@article{abc,
	author = {Donghyuk Lee and Samira Manabi Khan and Lavanya Subramanian and Rachata Ausavarungnirun and Gennady Pekhimenko and Vivek Seshadri and Saugata Ghose and Onur Mutlu},
	journal = {CoRR},
	title = {Reducing DRAM Latency by Exploiting Design-Induced Latency Variation in Modern DRAM Chips.},
	url = {http://arxiv.org/abs/1610.09604},
	year = {2016}
}

@inproceedings{abc,
	author = {Rachata Ausavarungnirun and Saugata Ghose and Onur Kayiran and Gabriel H. Loh and Chita R. Das and Mahmut T. Kandemir and Onur Mutlu},
	booktitle = {2015 International Conference on Parallel Architecture and Compilation, PACT 2015, San Francisco, CA, USA},
	title = {Exploiting Inter-Warp Heterogeneity to Improve GPGPU Performance.},
	url = {http://dx.doi.org/10.1109/PACT.2015.38},
	year = {2015}
}

@inproceedings{abc,
	author = {Yu Cai and Yixin Luo and Saugata Ghose and Onur Mutlu},
	booktitle = {45th Annual IEEE/IFIP International Conference on Dependable Systems and Networks, DSN 2015, Rio de Janeiro, Brazil},
	title = {Read Disturb Errors in MLC NAND Flash Memory: Characterization, Mitigation, and Recovery.},
	url = {http://dx.doi.org/10.1109/DSN.2015.49},
	year = {2015}
}

@inproceedings{abc,
	author = {Yixin Luo and Yu Cai and Saugata Ghose and Jongmoo Choi and Onur Mutlu},
	booktitle = {IEEE 31st Symposium on Mass Storage Systems and Technologies, MSST 2015, Santa Clara, CA, USA},
	title = {WARM: Improving NAND flash memory lifetime with write-hotness aware retention management.},
	url = {http://dx.doi.org/10.1109/MSST.2015.7208284},
	year = {2015}
}

@article{abc,
	author = {Donghyuk Lee and Gennady Pekhimenko and Samira Manabi Khan and Saugata Ghose and Onur Mutlu},
	journal = {CoRR},
	title = {Simultaneous Multi Layer Access: A High Bandwidth and Low Cost 3D-Stacked Memory Interface.},
	url = {http://arxiv.org/abs/1506.03160},
	year = {2015}
}

@article{abc,
	author = {Yang Li and Jongmoo Choi and Jin Sun and Saugata Ghose and Hui Wang and Justin Meza and Jinglei Ren and Onur Mutlu},
	journal = {CoRR},
	title = {Managing Hybrid Main Memories with a Page-Utility Driven Performance Model.},
	url = {http://arxiv.org/abs/1507.03303},
	year = {2015}
}