$n$

, and the runtime of each subtask has an exponential tail, coded computation can speed up distributed matrix multiplication by a factor of

$\\log n$

. For data shuffling, we use codes to reduce communication bottlenecks, exploiting the excess in storage. We show that when a constant fraction

$\\alpha $

of the data matrix can be cached at each worker, and

$n$

is the number of workers, coded shuffling reduces the communication cost by a factor of

$\\left({\\alpha + \\frac {1}{n}}\\right)\\gamma (n)$

compared with uncoded shuffling, where

$\\gamma (n)$

is the ratio of the cost of unicasting

$n$

messages to

$n$

users to multicasting a common message (of the same size) to

$n$

users. For instance,

$\\gamma (n) \\simeq n$

if multicasting a message to

$n$

users is as cheap as unicasting a message to one user. We also provide experimental results, corroborating our theoretical gains of the coded algorithms.","publishedDate","2017-08-04","publishedYear",2017,"venue","IEEE Transactions on Information Theory",866,"authors","keywords",[156,157,158],"pdfUrl","https://arxiv.org/pdf/1512.02673","topVenue","Stochastic Gradient Optimization Techniques","Advanced Data Storage Technologies","Caching and Content Delivery",{"_58":160,"_60":161},[],[162],{"_63":163,"_65":164,"_67":165,"_69":166},1780509701965,{"_72":168,"_74":75,"_76":163,"_77":-5,"_78":79,"_80":79,"_81":79,"_82":-5,"_83":-5,"_84":85,"_33":86,"_87":88},[71,167],"[\"portal\",\"industries\"]","industries",[169,170,171,172,173,174,175,176,177,178,179,180],{"_15":242,"_27":243,"_29":-5,"_30":244,"_31":245,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":246,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":243,"_57":245},{"_15":238,"_27":239,"_29":-5,"_30":-5,"_31":240,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":241,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":239,"_57":240},{"_15":232,"_27":233,"_29":-5,"_30":234,"_31":235,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":236,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":233,"_57":235},{"_15":228,"_27":229,"_29":-5,"_30":-5,"_31":230,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":231,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":229,"_57":230},{"_15":224,"_27":225,"_29":-5,"_30":-5,"_31":226,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":227,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":225,"_57":226},{"_15":218,"_27":219,"_29":-5,"_30":220,"_31":221,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":222,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":219,"_57":221},{"_15":213,"_27":214,"_29":-5,"_30":-5,"_31":215,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":216,"_193":79,"_194":217,"_195":79,"_196":-5,"_56":214,"_57":215},{"_15":209,"_27":210,"_29":-5,"_30":-5,"_31":211,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":212,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":210,"_57":211},{"_15":205,"_27":206,"_29":-5,"_30":-5,"_31":207,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":208,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":206,"_57":207},{"_15":201,"_27":202,"_29":-5,"_30":-5,"_31":203,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":204,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":202,"_57":203},{"_15":197,"_27":198,"_29":-5,"_30":-5,"_31":199,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":200,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":198,"_57":199},{"_15":181,"_27":182,"_29":-5,"_30":-5,"_31":183,"_32":-5,"_33":34,"_35":-5,"_37":-5,"_38":184,"_40":-5,"_42":-5,"_185":186,"_187":-5,"_188":75,"_189":190,"_191":192,"_193":79,"_194":85,"_195":79,"_196":-5,"_56":182,"_57":183},"tax_ind_smartphone","智能手机","Smartphone","manual","taxonomyType","INDUSTRY","code","level","parentId","tax_ind_root","path","科技行业/智能手机","sortOrder","isLeaf","entityCount","children","tax_ind_robot","机器人","Robotics","科技行业/机器人","tax_ind_wearable","智能穿戴","Smart Wearables","科技行业/智能穿戴","tax_ind_energy","新能源与智能制造","New Energy & Smart Manufacturing","科技行业/新能源与智能制造","tax_ind_biotech","生物科技","Biotech","科技行业/生物科技","tax_ind_quantum","量子计算","Quantum Computing","科技行业/量子计算",true,"tax_ind_blockchain","区块链",[223],"Blockchain","科技行业/区块链","Web3","tax_ind_enterprise","企业软件","Enterprise Software","科技行业/企业软件","tax_ind_security","网络安全","Cybersecurity","科技行业/网络安全","tax_ind_semi","半导体",[237],"Semiconductor","科技行业/半导体","芯片","tax_ind_cloud","云计算","Cloud Computing","科技行业/云计算","tax_ind_ai","人工智能",[247,248],"Artificial Intelligence","科技行业/人工智能","AI","AI 技术","env",{"_251":23},"SITE_BASE_URL"]

Ramtin Pedarsani 人物

相关论文查看全部 (1)

相关公司查看全部 (1)

相关事件

相关专利

相关文章