International audienceWhile virtualization only introduces a small overhead on machines with few cores, this is not the case on larger ones. Most of the overhead on the latter machines is caused by the Non-Uniform Memory Access (NUMA) architecture they are using. In order to reduce this overhead, this paper shows how NUMA placement heuristics can be implemented inside Xen. With an evaluation of 29 applications on a 48-core machine, we show that the NUMA placement heuristics can multiply the performance of 9 applications by more than 2
An increasing number of new multicore systems use the Non-Uniform Memory Access architecture due to ...
The problem of placement of threads, or virtual cores, on physical cores in a multicore system has b...
The latency of memory access times is hence non-uniform, because it depends on where the request ori...
International audienceWhile virtualization only introduces a small overhead on machines with few cor...
While virtualization only introduces a negligible overhead on machines with few cores, this is not t...
International audienceThis paper addresses the problem of efficiently virtualizing NUMA architecture...
International audienceNowadays, NUMA architectures are common in compute-intensive systems. Achievin...
International audienceNowadays, virtualization is a central element in data centers as it allows sha...
Our work addresses the problem of placement of threads, or virtual cores, onto physical cores in a m...
International audienceThe ever-growing level of parallelism within the multi-core and multi-processo...
International audiencevNUMA is the most recent technology used by hypervisors to deal with Non Unifo...
International audienceDynamic task-parallel programming models are popular on shared-memory systems,...
A common approach to improve memory access in NUMA machines exploits operating system (OS) page prot...
As the number of cores increases Non-Uniform Memory Access (NUMA) is becoming increasingly prevalent...
An increasing number of new multicore systems use the Non-Uniform Memory Access architecture due to ...
An increasing number of new multicore systems use the Non-Uniform Memory Access architecture due to ...
The problem of placement of threads, or virtual cores, on physical cores in a multicore system has b...
The latency of memory access times is hence non-uniform, because it depends on where the request ori...
International audienceWhile virtualization only introduces a small overhead on machines with few cor...
While virtualization only introduces a negligible overhead on machines with few cores, this is not t...
International audienceThis paper addresses the problem of efficiently virtualizing NUMA architecture...
International audienceNowadays, NUMA architectures are common in compute-intensive systems. Achievin...
International audienceNowadays, virtualization is a central element in data centers as it allows sha...
Our work addresses the problem of placement of threads, or virtual cores, onto physical cores in a m...
International audienceThe ever-growing level of parallelism within the multi-core and multi-processo...
International audiencevNUMA is the most recent technology used by hypervisors to deal with Non Unifo...
International audienceDynamic task-parallel programming models are popular on shared-memory systems,...
A common approach to improve memory access in NUMA machines exploits operating system (OS) page prot...
As the number of cores increases Non-Uniform Memory Access (NUMA) is becoming increasingly prevalent...
An increasing number of new multicore systems use the Non-Uniform Memory Access architecture due to ...
An increasing number of new multicore systems use the Non-Uniform Memory Access architecture due to ...
The problem of placement of threads, or virtual cores, on physical cores in a multicore system has b...
The latency of memory access times is hence non-uniform, because it depends on where the request ori...