Generalized cross subspace alignment codes for coded distributed batch matrix multiplication
The goal of coded distributed batch matrix multiplication is to efficiently multiply L instances
of λ x κ matrices, A=(A 1,···, AL), with L instances of κ x μ matrices B=(B 1,···, BL), by
distributing the computation across S servers, such that the response from any R servers (R
is called the recovery threshold) is sufficient to compute the L matrix products, AB=(A 1 B 1,
A 2 B 2,···, ALBL). Existing solutions either compute each AlBl one at a time by partitioning
individual matrices and coding across these partitions, or rely only on batch processing, ie …
of λ x κ matrices, A=(A 1,···, AL), with L instances of κ x μ matrices B=(B 1,···, BL), by
distributing the computation across S servers, such that the response from any R servers (R
is called the recovery threshold) is sufficient to compute the L matrix products, AB=(A 1 B 1,
A 2 B 2,···, ALBL). Existing solutions either compute each AlBl one at a time by partitioning
individual matrices and coding across these partitions, or rely only on batch processing, ie …
[PDF][PDF] Generalized Cross Subspace Alignment Codes for Coded Distributed Batch Matrix
The goal of coded distributed batch matrix multiplication is to efficiently multiply L instances
of λ× κ matrices, A=(A1,···, AL), with L instances of κ× µ matrices B=(B1,···, BL), by distributing
the computation across S servers, such that the response from any R servers (R is called the
recovery threshold) is sufficient to compute the L matrix products, AB=(A1B1, A2B2,···,
ALBL). Existing solutions either compute each AlBl one at a time by partitioning individual
matrices and coding across these partitions, or rely only on batch processing, ie, coding …
of λ× κ matrices, A=(A1,···, AL), with L instances of κ× µ matrices B=(B1,···, BL), by distributing
the computation across S servers, such that the response from any R servers (R is called the
recovery threshold) is sufficient to compute the L matrix products, AB=(A1B1, A2B2,···,
ALBL). Existing solutions either compute each AlBl one at a time by partitioning individual
matrices and coding across these partitions, or rely only on batch processing, ie, coding …
以上显示的是最相近的搜索结果。 查看全部搜索结果