560 Hits in 6.7 sec

Base-delta-immediate compression

Gennady Pekhimenko, Vivek Seshadri, Onur Mutlu, Phillip B. Gibbons, Michael A. Kozuch, Todd C. Mowry
2012 Proceedings of the 21st international conference on Parallel architectures and compilation techniques - PACT '12  
In this paper, we introduce a new compression algorithm called Base-Delta-Immediate (B∆I) compression, a practical technique for compressing data in on-chip caches.  ...  Cache compression is a promising technique to increase on-chip cache capacity and to decrease on-chip and off-chip bandwidth usage.  ...  We thank Greg Ganger, Kayvon Fatahalian and Michael Papamichael for their feedback on this paper's writing.  ... 
doi:10.1145/2370816.2370870 dblp:conf/IEEEpact/PekhimenkoSMGKM12 fatcat:lakldkz74bhsfetstbd5kycrd4

SCP: Synergistic cache compression and prefetching

Bhargavraj Patel, Nikos Hardavellas, Gokhan Memik
2015 2015 33rd IEEE International Conference on Computer Design (ICCD)  
Utilizing the cache compression hardware to compress the storage arrays for a STEMS streaming engine, in addition to the data cache, allows the streaming engine to operate entirely on-chip using space  ...  to implement the storage components of streaming prefetchers on the compressed cache.  ...  Without loss of generality, we assume Spatio-Temporal Memory Streaming (STeMS) [5] as the underlying stream prefetcher and Base Delta Immediate (BDI) [2] to compress the cache data.  ... 
doi:10.1109/iccd.2015.7357098 dblp:conf/iccd/PatelHM15 fatcat:km72ggsgmnfb7opioskryvhsei

Practical Data Compression for Modern Memory Hierarchies [article]

Gennady Pekhimenko
2016 arXiv   pre-print
In this thesis, we describe a new, practical approach to integrating hardware-based data compression within the memory hierarchy, including on-chip caches, main memory, and both on-chip and off-chip interconnects  ...  We use this observation to develop a new cache insertion policy for compressed caches, the Size-based Insertion Policy (SIP), which uses the size of a compressed block as one of the metrics to predict  ...  [10] , and (ii) Base-Delta-Immediate Compression [185] .  ... 
arXiv:1609.02067v1 fatcat:i4z7m2ydtjgwvlwmglno26nb54

Applying Data Compression Techniques on Systolic Neural Network Accelerator [article]

Navid Mirnouri
2016 arXiv   pre-print
Another technique is Data compression which is used in memory systems in order to save capacity and bandwidth.  ...  Approximate Computing is one promising technique that uses a trade off between precision and efficiency of computing.  ...  Base-delta-immediate is a compression technique for increasing on-chip cache's capacity. Data in cache lines are not wide in range.  ... 
arXiv:1701.03734v1 fatcat:nc7alulx3bbtxczw62wfekumie

Dictionary sharing: An efficient cache compression scheme for compressed caches

Biswabandan Panda, Andre Seznec
2016 2016 49th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO)  
The effectiveness of a compressed cache depends on three features: i) the compression scheme, ii) the compaction scheme, and iii) the cache layout of the compressed cache.  ...  In this paper, we propose dictionary sharing (DISH), a dictionary based cache compression scheme that reduces this wastage.  ...  ACKNOWLEDGMENT The authors would like to thank the anonymous reviewers and the members of the ALF team for their suggestions. The authors also thank Somayeh Sardashti and David Wood for their help.  ... 
doi:10.1109/micro.2016.7783704 dblp:conf/micro/PandaS16 fatcat:hapntyzezvckhfwicudouc2e74

Adaptive object code compression

John Gilbert, David M. Abrahamson
2006 Proceedings of the 2006 international conference on Compilers, architecture and synthesis for embedded systems - CASES '06  
This tree is used as a basis for dynamically building and maintaining an LZW dictionary for compression/decompression of individual instruction cache lines.  ...  We introduce a novel data structure, the compulsory miss tree, that is used to identify a partial order in which compulsory misses will have occurred in an instruction cache whenever a cache miss occurs  ...  Acknowledgments We are grateful to Edsko de Vries for helpful discussions and for critical feedback on early drafts of this paper.  ... 
doi:10.1145/1176760.1176795 dblp:conf/cases/GilbertA06 fatcat:q463lpnyendhxefutavjvcdnbi


Albin Eldstål-Ahrens, Ioannis Sourdis
2020 ACM Transactions on Architecture and Code Optimization (TACO)  
The first variant has a shared last-level cache (LLC) on the processor-die, which is modified to store both compressed and uncompressed data.  ...  The second has a 3D-stacked DRAM cache with larger cache lines that match the granularity of the compressed memory blocks and stores only uncompressed data.  ...  These designs employ compression methods such as Base-Delta-Immediate (BDI) [24] on single cache blocks, or DISH [18] , which allow a small number of adjacent blocks to share a dictionary.  ... 
doi:10.1145/3424668 fatcat:kcbtlsy5pvemniapnvaxqbgeqa

Lane Compression

Yousun Ko, Alex Chadwick, Daniel Bates, Robert Mullins
2021 ACM Transactions on Embedded Computing Systems  
This article presents Lane Compression, a lightweight lossless compression technique for machine learning that is based on a detailed study of the statistical properties of machine learning data.  ...  Hardware implementations are described and the scheme's simplicity makes it suitable for compressing both on-chip and off-chip traffic.  ...  Works has focused on compressing off-chip memory traffic [5, 6, 24, 31, 43, 46] , data between computing units and cache [6, 8, 15, 17, 18, 52] , or over the on-chip network traffic for a tiled architecture  ... 
doi:10.1145/3431815 fatcat:zoimkqqn2be6xiasj3kt3eiage

Energy aware lossless data compression

Kenneth Barr, Krste Asanović
2003 Proceedings of the 1st international conference on Mobile systems, applications and services - MobiSys '03  
One such energy-aware suggestion is asymmetric compression, the use of one compression algorithm on the transmit side and a different algorithm for the receive path.  ...  If the energy required to compress data is less than the energy required to send it, there is a net energy savings and an increase in battery life for portable computers.  ...  ACKNOWLEDGMENTS Thanks to John Ankcorn, Christopher Batten, Jamey Hicks, Ronny Krashinsky, Hari Balakrishnan, and the anonymous reviewers for their comments and assistance.  ... 
doi:10.1145/1066116.1066123 fatcat:unyoxuspt5g37abrgzbm2juyge

Energy-aware lossless data compression

Kenneth C. Barr, Krste Asanović
2006 ACM Transactions on Computer Systems  
One such energy-aware suggestion is asymmetric compression, the use of one compression algorithm on the transmit side and a different algorithm for the receive path.  ...  If the energy required to compress data is less than the energy required to send it, there is a net energy savings and an increase in battery life for portable computers.  ...  ACKNOWLEDGMENTS Thanks to John Ankcorn, Christopher Batten, Jamey Hicks, Ronny Krashinsky, Hari Balakrishnan, and the anonymous reviewers for their comments and assistance.  ... 
doi:10.1145/1151690.1151692 fatcat:x22sj3wugbbqzfcgpksq2ndkai

L2C2: Last-Level Compressed-Cache NVM and a Procedure to Forecast Performance and Lifetime [article]

Carlos Escuin, Pablo Ibañez, Teresa Monreal, Jose M. Llaberia, Victor Viñals
2022 arXiv   pre-print
We also introduce L2C2, a LLC design intended for implementation in NV memory technology that combines fault tolerance, compression, and internal write wear leveling for the first time.  ...  replacement, wear-leveling, compression, etc.) on the temporal evolution of the indices of interest, such as the effective capacity of the NV-LLC or the system IPC.  ...  The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.  ... 
arXiv:2204.09504v3 fatcat:gfnfd3qlmve2nd7poeeyqcho6m

E^2MC: Entropy Encoding Based Memory Compression for GPUs

Sohan Lal, Jan Lucas, Ben Juurlink
2017 2017 IEEE International Parallel and Distributed Processing Symposium (IPDPS)  
We propose an entropy encoding based memory compression (E 2 MC) technique for GPUs, which is based on the well-known Huffman encoding.  ...  We study the feasibility of entropy encoding for GPUs and show that it achieves higher compression ratios than state-of-the-art GPU compression techniques.  ...  Base-Delta-Immediate (BDI) compression [9] exploits value similarity. While these techniques can decompress with few cycles, their compression ratio is low, typically only 1.5×.  ... 
doi:10.1109/ipdps.2017.101 dblp:conf/ipps/LalLJ17 fatcat:hkaetvxnazaffbplwlfwjxt734

Efficient Similarity-aware Compression to Reduce Bit-writes in Non-Volatile Main Memory for Image-based Applications [article]

Zhangyu Chen, Yu Hua, Pengfei Zuo, Yuanyuan Sun, Yuncheng Guo
2019 arXiv   pre-print
Although compressing data in write accesses to NVMs on-the-fly reduces the bit-writes in NVMs, existing precise or approximate compression schemes show limited performance improvements for data of bitmaps  ...  By exploiting the pixel-level similarity, we propose SimCom, an efficient similarity-aware compression scheme in hardware layer, to compress data for each write access on-the-fly.  ...  Base Delta Immediate (BDI) [30] .  ... 
arXiv:1905.02487v1 fatcat:xh7rktbgebhs7evjaahymetooi

An approximation to the greedy algorithm for differential compression

R. C. Agarwal, K. Gupta, S. Jain, S. Amalapurapu
2006 IBM Journal of Research and Development  
Our algorithm, hsadelta (hash suffix array delta), finds the best matches for every offset of the version file, with respect to a certain granularity and above a certain length threshold.  ...  The algorithm has two variations depending on how we choose the block size.  ...  Acknowledgments The authors thank Randal Burns, Laurent Chavet, Fred Douglis, Dana Shapira, and Andrew Tridgell for providing their code and/or data sets.  ... 
doi:10.1147/rd.501.0149 fatcat:iuhuu4df3zffbdxi6utb5xyecq

Exact and Error-bounded Approximate Color Buffer Compression and Decompression [article]

Jim Rasmusson, Jon Hasselgren, Tomas Akenine-Moeller
2007 Proceedings of the ACM SIGGRAPH/EUROGRAPHICS conference on Graphics hardware - HWWS '04  
After that, we introduce a new scheme based on an exactly reversible color transform, simple prediction, and Golomb-Rice encoding.  ...  Our results indicate superior compression ratios compared to existing algorithms, and we believe that approximate compression can be important for mobile GPUs.  ...  Thanks to the anonymous reviewers for their helpful comments.  ... 
doi:10.2312/eggh/eggh07/041-048 fatcat:t35fdfth6jgmljabql733fhzqi
« Previous Showing results 1 — 15 out of 560 results