Filters








484 Hits in 5.8 sec

Rethinking memory system design for data-intensive computing

Onur Mutlu
2015 2015 International Conference on Embedded Computer Systems: Architectures, Modeling, and Simulation (SAMOS)  
 Summary The DRAM Scaling Problem  DRAM stores charge in a capacitor (charge-based memory)  Capacitor must be large enough for reliable sensing  Access transistor should be large enough for low leakage  ...  DRAM Subarray Operation (load one byte) Row Buffer (4 Kbits) Data Bus 8 bits DRAM array 4 Kbits Step 1: Activate row Transfer row Step 2: Read Transfer byte onto bus RowClone: In-DRAM  ... 
doi:10.1109/samos.2015.7363650 dblp:conf/samos/Mutlu15 fatcat:fc5a6u4sinhotaibqt6bsull2a

Increasing the Profit of Cloud Providers through DRAM Operation at Reduced Margins

Christos Kalogirou, Christos D. Antonopoulos, Nikolaos Bellas, Spyros Lalis, Lev Mukhanov, Georgios Karakonstantis
2020 2020 20th IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing (CCGRID)  
Moreover, energy and cost gains are higher in nodes with larger memory capacity, which shows that our approach will scale favorably in future data-dominated data centers.  ...  In our evaluation framework rather than using workload unaware DRAM error models as in existing studies [9] , [11] , [12] , [33] , we extend our recent workload aware DRAM error model with power values  ... 
doi:10.1109/ccgrid49817.2020.00-38 dblp:conf/ccgrid/KalogirouABLMK20 fatcat:4okvnxw2ofbifams54v365vruq

Large Scale Studies of Memory, Storage, and Network Failures in a Modern Data Center [article]

Justin Meza
2019 arXiv   pre-print
Faulty devices make operating large scale data centers challenging because the workloads running in modern data centers consist of interdependent programs distributed across many servers, so failures that  ...  The workloads running in the modern data centers of large scale Internet service providers (such as Amazon, Baidu, Facebook, Google, and Microsoft) support billions of users and span globally distributed  ...  Related Research in DRAM Failures in Modern Data Centers Many prior works have examined DRAM failures at a small scale or in a controlled environment.  ... 
arXiv:1901.03401v1 fatcat:xizly26q6zebdovjosjpyw2o5q

Power Management Techniques for Data Centers: A Survey [article]

Sparsh Mittal
2014 arXiv   pre-print
This, however, has resulted in significant increase in the power consumption of the data centers. For this reason, managing power consumption of data centers has become essential.  ...  In this paper, we highlight the need of achieving energy efficiency in data centers and survey several recent architectural techniques designed for power management of data centers.  ...  The large power consumption and high concentration of nodes in data centers leads to increased node failures.  ... 
arXiv:1404.6681v2 fatcat:iuiczutd5rdwraxops6bzmoxdm

An energy-efficient and error-resilient server ecosystem exceeding conservative scaling limits

Georgios Karakonstantis, Konstantinos Tovletoglou, Lev Mukhanov, Hans Vandierendonck, Dimitrios S. Nikolopoulos, Peter Lawthers, Panos Koutsovasilis, Manolis Maroudas, Christos D. Antonopoulos, Christos Kalogirou, Nikos Bellas, Spyros Lalis (+14 others)
2018 2018 Design, Automation & Test in Europe Conference & Exhibition (DATE)  
Our characterization results on a 64-bit ARMv8 micro-server in 28nm process reveal large voltage margins in terms of Vmin variation among the 8 cores of the CPU chip, among 3 different sigma chips, and  ...  Consequently, there is a need for more energy efficient servers to empower traditional centralized Cloud data-centers as well as emerging decentralized data-centers at the Edges of the Cloud.  ...  Therefore, the predictor, apart from predicting the safe Vmin, it can also assist task scheduling in conjunction to frequency scaling according to the current workload on the system to further improve  ... 
doi:10.23919/date.2018.8342175 dblp:conf/date/KarakonstantisT18 fatcat:q7nxvds5mnhn7nv2pouokoayu4

Reliability and Performance Trade-off Study of Heterogeneous Memories

Manish Gupta, David Roberts, Mitesh Meswani, Vilas Sridharan, Dean Tullsen, Rajesh Gupta
2016 Proceedings of the Second International Symposium on Memory Systems - MEMSYS '16  
In this paper, authors use the real-world DRAM failure data to conduct a trade-off study on reliability and performance of Heterogeneous Memory Architectures (HMA).  ...  Unfortunately, reliability of such arrangements has not been studied carefully thus making them less attractive for data centers and mission-critical systems.  ...  As heterogeneous memory penetrates the commercial data centers, it will open the doors for large-scale studies similar to the DRAM field study, and real-world fault and error rates can be incorporated  ... 
doi:10.1145/2989081.2989113 dblp:conf/memsys/GuptaRMSTG16 fatcat:275k25jrpfaqpdpuvi3af3mug4

Research Problems and Opportunities in Memory Systems

2014 Supercomputing Frontiers and Innovations  
memory systems), 3) providing predictable performance and QoS to applications sharing the memory system (i.e., QoS-aware memory systems).  ...  We also briefly describe our ongoing related work in combating scaling challenges of NAND flash memory.  ...  Some of the research reported here was also partially supported by GSRC, Intel URO Memory Hierarchy Program, Intel Science and Technology Center on Cloud Computing, NIH, NSF, and SRC.  ... 
doi:10.14529/jsfi140302 fatcat:2zfa7zk3qjgohdsgxmkkqaamuu

Temperature management in data centers

Nosayba El-Sayed, Ioan A. Stefanovici, George Amvrosiadis, Andy A. Hwang, Bianca Schroeder
2012 Proceedings of the 12th ACM SIGMETRICS/PERFORMANCE joint international conference on Measurement and Modeling of Computer Systems - SIGMETRICS '12  
A large fraction of the consumed energy is spent on data center cooling, which has motivated a large body of work on temperature management in data centers.  ...  This paper provides a multi-faceted study of temperature management in data centers.  ...  In particular, we thank Wolf-Dietrich Weber for his efforts in making the data available for us and Chris Mayfield who provided scripts to extract some of the data.  ... 
doi:10.1145/2254756.2254778 dblp:conf/sigmetrics/El-SayedSAHS12 fatcat:rbn6t6sbw5b3pi67lbp3gm7sxi

Temperature management in data centers

Nosayba El-Sayed, Ioan A. Stefanovici, George Amvrosiadis, Andy A. Hwang, Bianca Schroeder
2012 Performance Evaluation Review  
A large fraction of the consumed energy is spent on data center cooling, which has motivated a large body of work on temperature management in data centers.  ...  This paper provides a multi-faceted study of temperature management in data centers.  ...  In particular, we thank Wolf-Dietrich Weber for his efforts in making the data available for us and Chris Mayfield who provided scripts to extract some of the data.  ... 
doi:10.1145/2318857.2254778 fatcat:rjbjhx5r6fcqnoqtz4gx6sp2ue

Understanding Reduced-Voltage Operation in Modern DRAM Devices

Kevin K. Chang, Onur Mutlu, A. Giray Yağlıkçı, Saugata Ghose, Aditya Agrawal, Niladrish Chatterjee, Abhijith Kashyap, Donghyuk Lee, Mike O'Connor, Hasan Hassan
2017 Proceedings of the ACM on Measurement and Analysis of Computing Systems  
Aggressive supply voltage reduction requires a thorough understanding of the effect voltage scaling has on DRAM access latency and DRAM reliability.  ...  Based on our observations, we propose a new DRAM energy reduction mechanism, called Voltron.  ...  This research was supported in part by the ISTC-CC, SRC, and NSF (grants 1212962 and 1320531). Kevin Chang is supported in part by an SRCEA/Intel Fellowship.  ... 
doi:10.1145/3084447 dblp:journals/pomacs/ChangYGACKLOHM17 fatcat:oaonzbkwpjhj5jabnmklgdgmn4

Hardware/software techniques for DRAM thermal management

Song Liu, Brian Leung, Alexander Neckar, Seda Ogrenci Memik, Gokhan Memik, Nikos Hardavellas
2011 2011 IEEE 17th International Symposium on High Performance Computer Architecture  
Another important property of DRAM temperature is the large variation in DRAM chip temperatures.  ...  Our experiments show that in a high performance memory system, our schemes reduce the peak DRAM chip temperature by as much as 8.39°C over 10 workloads (5.36°C on average).  ...  This work is in part supported by NSF Grant CCF-0916746, CCF-0747201 and CNS-0720691.  ... 
doi:10.1109/hpca.2011.5749756 dblp:conf/hpca/LiuLNMMH11 fatcat:pdawaugstfhodgwapxsspbztme

A Modern Primer on Processing in Memory [article]

Onur Mutlu, Saugata Ghose, Juan Gómez-Luna, Rachata Ausavarungnirun
2020 arXiv   pre-print
data movement, especially off-chip to on-chip, is very expensive in terms of bandwidth, energy and latency, much more so than computation.  ...  are increasingly data-intensive, and memory bandwidth and energy do not scale well, (2) energy consumption is a key limiter in almost all computing platforms, especially server and mobile systems, (3)  ...  Figure 13 illustrates the key idea of using the pattern of predictable access latency failures in a DRAM subarray to generate a unique DRAM device identifier.  ... 
arXiv:2012.03112v1 fatcat:hq2i2xzq4nbszenq7rqmjzcjci

Main Memory Scaling: Challenges and Solution Directions [chapter]

Onur Mutlu
2015 More than Moore Technologies for Next Generation Computer Design  
predictable performance and QoS to applications sharing the memory system (i.e., QoS-aware memory systems).  ...  Creation of new killer applications and usage models for computers likely depends on how well the memory system can support the efficient storage and manipulation of data in such data-intensive applications  ...  Part of the structure of this chapter is based on an evolving set of talks I have delivered at various venues on Scaling the Memory System in the Many-Core Era and Rethinking Memory System Design for Data-Intensive  ... 
doi:10.1007/978-1-4939-2163-8_6 fatcat:okw4kxakuja43kac65zy5c35ye

A Large-Scale Study of Flash Memory Failures in the Field

Justin Meza, Qiang Wu, Sanjev Kumar, Onur Mutlu
2015 Proceedings of the 2015 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems - SIGMETRICS '15  
In a data center environment, flash-based SSD failures can lead Keywords to downtime and, in the worst case, data loss.  ...  But failures are also noticeable for dense allocations of data on the SSD and a large DRAM buffer usage data mappings (e.g.  ... 
doi:10.1145/2745844.2745848 dblp:conf/sigmetrics/MezaWKM15 fatcat:blzbw5rorvb33ku6jsy52bc4ji

A Large-Scale Study of Flash Memory Failures in the Field

Justin Meza, Qiang Wu, Sanjev Kumar, Onur Mutlu
2015 Performance Evaluation Review  
In a data center environment, flash-based SSD failures can lead Keywords to downtime and, in the worst case, data loss.  ...  But failures are also noticeable for dense allocations of data on the SSD and a large DRAM buffer usage data mappings (e.g.  ... 
doi:10.1145/2796314.2745848 fatcat:ri6k55vsqzdr5nigtr7mksf2ey
« Previous Showing results 1 — 15 out of 484 results