Abstract
This paper examines whether lossy compression can be used effectively in physics simulations as a possible strategy to combat the expected data-movement bottleneck in future high performance computing architectures. We show that, for the codes and simulations we tested, compression levels of 3–5X can be applied without causing significant changes to important physical quantities. Rather than applying signal processing error metrics, we utilize physics-based metrics appropriate for each code to assess the impact of compression. We evaluate three different simulation codes: a Lagrangian shock-hydrodynamics code, an Eulerian higher-order hydrodynamics turbulence modeling code, and an Eulerian coupled laser-plasma interaction code. We compress relevant quantities after each time-step to approximate the effects of tightly coupled compression and study the compression rates to estimate memory and disk-bandwidth reduction. We find that the error characteristics of compression algorithms must be carefully considered in the context of the underlying physics being modeled.
Highlights
The computing power of large systems is increasing faster than their memory and disk bandwidth
This paper has examined the impact of lossy compression on three physics simulation codes – LULESH, Miranda, and pF3D
This is the first study of the effects of applying lossy compression to the physics state of simulations as a strategy for mitigating the data movement bottleneck expected on future systems
Summary
The computing power of large systems is increasing faster than their memory and disk bandwidth. To have a significant impact on the memory or disk bandwidth requirements of simulations, lossy compression methods are necessary. The APAX compression algorithm can achieve throughputs of 2.5 to 8 GB/s while using between 0.1 and 0.4mm in 28 mm CMOS and increasing the power consumption of a chip by less than one percent This rate is close enough to the memory bandwidth per core that memory compression might increase overall application performance. An important consideration is that memory compression has a larger impact on the accuracy of a simulation than disk compression because it occurs hundreds of times more frequently than checkpointing. The final section summarizes our results and indicates topics for future research
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
