Filters








5,361 Hits in 5.0 sec

Scaling the issue window with look-ahead latency prediction

Yongxiang Liu, Anahita Shayesteh, Gokhan Memik, Glenn Reinman
2004 Proceedings of the 18th annual international conference on Supercomputing - ICS '04  
Scaling issue window size can certainly provide more ILP; however, future processor scaling demands threaten to limit the size of the issue window.  ...  Using this knowledge, the instructions are placed into a sorting structure, which allows instructions with shorter waiting times enter the issue window ahead of those instructions with longer waiting times  ...  Issue Window Scaling Performance Assisting Speculative Schedulers with Look-ahead Load Latency Prediction Speculative schedulers, such as Cyclone, schedule load dependents ahead of time by assuming the  ... 
doi:10.1145/1006209.1006240 dblp:conf/ics/LiuSMR04 fatcat:jsr5sz3u6vh6bbyri3spn5niiy

Look-Ahead Architecture Adaptation to Reduce Processor Power Consumption

Zhichun Zhu, Xiaodong Zhang
2005 IEEE Micro  
Acknowledgments We thank the anonymous referees for their constructive comments.  ...  A critical issue of any look-ahead technique, however, is to find a reliable slackness indicator.  ...  Look-ahead techniques, on the other hand, use indicators whose changes can be detected 12 LOOK-AHEAD ARCHITECTURE ADAPTATION IEEE MICRO An increasing number of researchers are investigating ways to reduce  ... 
doi:10.1109/mm.2005.70 fatcat:czf6h7n655hu3fb3j7c57vuk7m

(ECOC 2020) Extending P4 In-Band Telemetry to the User Equipment for Latency and Localization-aware Autonomous Networking with AI Forecasting

Davide Scano, Francesco Paolucci, Koteswararao Kondepu, Andrea Sgambelluri, Luca Valcarenghi, Filippo Cugini
2021 Journal of Optical Communications and Networking  
The proposal also includes an Artifical Intelligence (AI)-assisted forecast system able to predict latency and geolocation in advance and trigger faster edge steering.  ...  However, the synchronization between latency and location information, typically provided by different platforms, is hard to achieve with current monitoring systems.  ...  The JU receives support from the European Union's H2020 research and innovation programme and from Italy Ministry of Education, University and Research (MIUR).  ... 
doi:10.1364/jocn.425891 fatcat:j53zdgsq25dtbm5eyeprtnn2yi

A simple latency tolerant processor

Satyanarayana Nekkalapu, Haitham Akkary, Komal Jothi, Renjith Retnamma, Xiaoyu Song
2008 2008 IEEE International Conference on Computer Design  
The non-blocking property of this architecture provides tolerance to hundreds of cycles of cache miss latency on a simple in-order issue core, thus allowing many more such cores to be integrated on the  ...  How can a designer build a processor that provides a core with good single-thread performance in the presence of long latency cache misses, while enabling as many of these cores to be placed on the same  ...  The architecture ability to tolerate long memory latencies can be measured by the fraction of look-ahead execution after a miss that is independent of the miss data.  ... 
doi:10.1109/iccd.2008.4751889 dblp:conf/iccd/NekkalapuAJRS08 fatcat:z6z3vdg4nreipjtu3xzgvexp5m

Scaling Social Media Applications Into Geo-Distributed Clouds

Yu Wu, Chuan Wu, Bo Li, Linquan Zhang, Zongpeng Li, Francis C. M. Lau
2015 IEEE/ACM Transactions on Networking  
in a simple but effective epidemic model; 2) one-shot optimal content migration and request distribution based on efficient optimization algorithms to address the predicted demand; and 3) a -step look-ahead  ...  Our key contribution is an online content migration and request distribution algorithm with the following features: 1) future demand prediction by novelly characterizing social influences among the users  ...  Impact of Look-Ahead Window Size We next investigate the performance of our online algorithm when different look-ahead window sizes are employed, i.e., in Algorithm 1.  ... 
doi:10.1109/tnet.2014.2308254 fatcat:rlvxp6ez4rcardfyld67hx2f4q

Performance improvement with circuit-level speculation

Tong Liu, Shih-Lien Lu
2000 Proceedings of the 33rd annual ACM/IEEE international symposium on Microarchitecture - MICRO 33  
Most of the time the partial implementation gives the correct result as if the function is implemented fully but with fewer gates delay allowing a higher pipeline frequency.  ...  Simulation results show that this method provides some performance improvement for a wide-issue superscalar if these stages are finely pipelined.  ...  The prediction rate of an N-bit adder with k bits carry chain is: P(N, k) = 1 1 ) 2 1 ( 2 − − + − k N k For example, a 64-bit approximation adder with 8-bit (8 = 64 ) look-ahead gives correct result 95%  ... 
doi:10.1145/360128.360166 fatcat:tijmgzwlxjhcnknq7oesjjrzfu

Multithreading decoupled architectures for complexity-effective general purpose computing

Michael Sung, Ronny Krashinsky, Krste Asanović
2001 SIGARCH Computer Architecture News  
This work investigates the possibility of using multithreading to overcome the loss of decoupling dependencies that represent the cause of this main limitation in decoupled architectures.  ...  Decoupled architectures have not traditionally been used in the context of general purpose computing because of their inability to tolerate control-intensive code that exists across a wide range of applications  ...  Large structures such as issue windows that require associative dependency-checking are the limiting factors in scaling the issue rate of superscalar machines [1] .  ... 
doi:10.1145/563647.563658 fatcat:fjmdpove5ravhclvctbfurz6im

Dimensioning V2N Services in 5G Networks through Forecast-based Scaling

Jorge Martin-Perez, Koteswararao Kondepu, Danny De Vleeschauwer, Venkatarami Reddy, Carlos Guimaraes, Andrea Sgambelluri, Luca Valcarenghi, Chrysa Papagianni, Carlos J. Bernardos
2022 IEEE Access  
Using the the most accurate forecasting technique, we propose n-max algorithm as a forecast-based scaling algorithm for vertical scaling of edge resources, comparing its benefits against state-of-the-art  ...  With the increasing adoption of intelligent transportation systems and the upcoming era of autonomous vehicles, vehicular services (such as remote driving, cooperative awareness, and hazard warning) will  ...  that gave the lowest RMSE for n minutes look-ahead predictions (see Table 4 ).  ... 
doi:10.1109/access.2022.3142346 fatcat:ywsqlayzajdubchfpykwyn4lda

LACOPE: Latency-Constrained Pitch Estimation for Speech Enhancement

Hendrik Schröter, Tobias Rosenkranz, Alberto N. Escalante-B., Andreas K. Maier
2021 Conference of the International Speech Communication Association  
That is, a recurrent neural network compensates delay introduced by the feature computation by predicting the pitch for a desired point, allowing a trade-off between pitch accuracy and latency.  ...  Our pitch estimation performance is on par with SOTA algorithms like PYIN or CREPE for spoken speech in all noise conditions while introducing minimal latency.  ...  Fig. 5 shows the pitch accuracy measured in percentage of frames with a pitch difference below 50 cents. Here, the tradeoff between look-ahead and pitch accuracy becomes clear.  ... 
doi:10.21437/interspeech.2021-633 dblp:conf/interspeech/SchroterREM21 fatcat:vrdqduzccbeehkjwtx4epppmzi

Real-time prediction of respiratory motion based on local regression methods

D Ruan, J A Fessler, J M Balter
2007 Physics in Medicine and Biology  
The proposed method reduced the prediction error for all imaging rates and latency lengths, particularly for long prediction lengths.  ...  One of the challenges in using these techniques is real-time tracking and predicting target motion, which is necessary to accommodate system latencies.  ...  The authors gratefully acknowledge Dr Gregory Sharp, Dr Paul Keall, Dr Amit Sawant, Dr Raghu Venkat and Vikram Srivasta for valuable discussions, as well as the anonymous reviewers for their helpful comments  ... 
doi:10.1088/0031-9155/52/23/024 pmid:18029998 fatcat:l6xdpwa3mbgdzex5nplmrstake

A scalable front-end architecture for fast instruction delivery

Glenn Reinman, Todd Austin, Brad Calder
1999 SIGARCH Computer Architecture News  
Delivering the performance necessary to meet future processor execution targets requires that the performance of the instruction delivery mechanism scale with the execution core.  ...  To counter these challenges, we present a fetch architecture that permits a faster cycle time than previous designs and scales better with future process technologies.  ...  Acknowledgments We would like to thank the anonymous reviewers for providing useful comments on this paper. This work was funded in part by NSF CAREER grant No. CCR-9733278, NSF grant No.  ... 
doi:10.1145/307338.300999 fatcat:tmv7hpbcavet5mkxacwhgd2pba

Learning to Count Words in Fluent Speech enables Online Speech Recognition [article]

George Sterpu, Christian Saam, Naomi Harte
2020 arXiv   pre-print
Practical usage is however limited to cases where full utterance latency is acceptable.  ...  Experiments performed on the LRS2, LibriSpeech, and Aishell-1 datasets of English and Mandarin speech show that the online system performs comparable with the offline one when having a dynamic algorithmic  ...  Because of latency considerations, instead of the original full connectivity in Figure 1b, we use the type displayed in Figure 1d , with controlled look-back e LB and look-ahead e LA frames.  ... 
arXiv:2006.04928v3 fatcat:7nfd3jghvbehzciyeb34i7hnoy

Nexus: a novel weighted-graph-based prefetching algorithm for metadata servers in petabyte-scale storage systems

Peng Gu, Yifeng Zhu, Hong Jiang, Jun Wang
2006 Sixth IEEE International Symposium on Cluster Computing and the Grid (CCGRID'06)  
of metadata operations can be reduced by up to 67%, compared with LRU and an existing state of the art prefetching algorithm.  ...  Extensive trace-driven simulations show that by adopting our new prefetching algorithm, the hit rate for metadata access on the client site can be increased by up to 13%, while the average response time  ...  the usefulness of the relationships captured by the look-ahead window.  ... 
doi:10.1109/ccgrid.2006.73 dblp:conf/ccgrid/GuZJW06 fatcat:iwub545uqvg6dmvshq7ic4pkr4

High-Performance Throughput Computing

S. Chaudhry, P. Caprioli, S. Yip, M. Tremblay
2005 IEEE Micro  
Run-ahead execution can effectively cover L1 latency. Covering the longer latencies of remote caches and memory came at a hardware cost and some potential slowdown.  ...  The left bar of each set is for a moderately speculative, out-of-order core; the right bar is for a core with an aggressive issue policy. Figure 7 . Window termination conditions.  ... 
doi:10.1109/mm.2005.49 fatcat:wq3ukuhpg5gubkia7bdjwz2nfu

Scaling social media applications into geo-distributed clouds

Yu Wu, Chuan Wu, Bo Li, Linquan Zhang, Zongpeng Li, Francis C.M. Lau
2012 2012 Proceedings IEEE INFOCOM  
look-ahead mechanism to adjust the one-shot optimization results towards the offline optimum.  ...  Our key contribution is an online content migration and request distribution algorithm with the following features: (1) future demand prediction by novelly characterizing social influences among the users  ...  Impact of Look-ahead Window Size We also investigate the performance improvement when different look-ahead window sizes are employed in our online algorithm.  ... 
doi:10.1109/infcom.2012.6195813 dblp:conf/infocom/WuWLZLL12 fatcat:rxluzbp5tnhdnkw5rnr4f6vy2a
« Previous Showing results 1 — 15 out of 5,361 results