Matches in SemOpenAlex for { <https://semopenalex.org/work/W2888832568> ?p ?o ?g. }
- W2888832568 endingPage "44" @default.
- W2888832568 startingPage "27" @default.
- W2888832568 abstract "Contemporary discrete GPUs support rich memory management features such as virtual memory and demand paging. These features simplify GPU programming by providing a virtual address space abstraction similar to CPUs and eliminating manual memory management, but they introduce high performance overheads during (1) address translation and (2) page faults. A GPU relies on high degrees of thread-level parallelism (TLP) to hide memory latency. Address translation can undermine TLP, as a single miss in the translation lookaside buffer (TLB) invokes an expensive serialized page table walk that often stalls multiple threads. Demand paging can also undermine TLP, as multiple threads often stall while they wait for an expensive data transfer over the system I/O (e.g., PCIe) bus when the GPU demands a page. In modern GPUs, we face a trade-off on how the page size used for memory management affects address translation and demand paging. The address translation overhead is lower when we employ a larger page size (e.g., 2MB large pages, compared with conventional 4KB base pages), which increases TLB coverage and thus reduces TLB misses. Conversely, the demand paging overhead is lower when we employ a smaller page size, which decreases the system I/O bus transfer latency. Support for multiple page sizes can help relax the page size trade-off so that address translation and demand paging optimizations work together synergistically. However, existing page coalescing (i.e., merging base pages into a large page) and splintering (i.e., splitting a large page into base pages) policies require costly base page migrations that undermine the benefits multiple page sizes provide. In this paper, we observe that GPGPU applications present an opportunity to support multiple page sizes without costly data migration, as the applications perform most of their memory allocation en masse (i.e., they allocate a large number of base pages at once).We show that this en masse allocation allows us to create intelligent memory allocation policies which ensure that base pages that are contiguous in virtual memory are allocated to contiguous physical memory pages. As a result, coalescing and splintering operations no longer need to migrate base pages." @default.
- W2888832568 created "2018-09-07" @default.
- W2888832568 creator A5003307688 @default.
- W2888832568 creator A5005696647 @default.
- W2888832568 creator A5028280505 @default.
- W2888832568 creator A5036666743 @default.
- W2888832568 creator A5050695684 @default.
- W2888832568 creator A5057505785 @default.
- W2888832568 creator A5084751735 @default.
- W2888832568 date "2018-08-28" @default.
- W2888832568 modified "2023-10-16" @default.
- W2888832568 title "Mosaic" @default.
- W2888832568 cites W1973573211 @default.
- W2888832568 cites W1983235612 @default.
- W2888832568 cites W2004807638 @default.
- W2888832568 cites W2020733012 @default.
- W2888832568 cites W2047060659 @default.
- W2888832568 cites W2047240985 @default.
- W2888832568 cites W2049403384 @default.
- W2888832568 cites W2056816375 @default.
- W2888832568 cites W2057569249 @default.
- W2888832568 cites W2061812855 @default.
- W2888832568 cites W2071767440 @default.
- W2888832568 cites W2079248286 @default.
- W2888832568 cites W2080592089 @default.
- W2888832568 cites W2089336473 @default.
- W2888832568 cites W2090584832 @default.
- W2888832568 cites W2093828978 @default.
- W2888832568 cites W2098040113 @default.
- W2888832568 cites W2098274770 @default.
- W2888832568 cites W2100926301 @default.
- W2888832568 cites W2103707825 @default.
- W2888832568 cites W2107333973 @default.
- W2888832568 cites W2115172404 @default.
- W2888832568 cites W2125551452 @default.
- W2888832568 cites W2126987530 @default.
- W2888832568 cites W2129381159 @default.
- W2888832568 cites W2130604611 @default.
- W2888832568 cites W2135438094 @default.
- W2888832568 cites W2136684653 @default.
- W2888832568 cites W2138163628 @default.
- W2888832568 cites W2142119745 @default.
- W2888832568 cites W2142444503 @default.
- W2888832568 cites W2142875853 @default.
- W2888832568 cites W2143609451 @default.
- W2888832568 cites W2146241244 @default.
- W2888832568 cites W2146757372 @default.
- W2888832568 cites W2149234156 @default.
- W2888832568 cites W2150722965 @default.
- W2888832568 cites W2153806645 @default.
- W2888832568 cites W2153882937 @default.
- W2888832568 cites W2155503253 @default.
- W2888832568 cites W2156703074 @default.
- W2888832568 cites W2159908132 @default.
- W2888832568 cites W2162293132 @default.
- W2888832568 cites W2163365446 @default.
- W2888832568 cites W2163687928 @default.
- W2888832568 cites W2166263440 @default.
- W2888832568 cites W2167362358 @default.
- W2888832568 cites W2169880332 @default.
- W2888832568 cites W2170257519 @default.
- W2888832568 cites W2208481289 @default.
- W2888832568 cites W2238307037 @default.
- W2888832568 cites W2238595726 @default.
- W2888832568 cites W2293640271 @default.
- W2888832568 cites W2323625620 @default.
- W2888832568 cites W2412101011 @default.
- W2888832568 cites W2522765117 @default.
- W2888832568 cites W2536535850 @default.
- W2888832568 cites W2604399885 @default.
- W2888832568 cites W2621371400 @default.
- W2888832568 cites W2625231790 @default.
- W2888832568 cites W2764065518 @default.
- W2888832568 cites W2793599434 @default.
- W2888832568 cites W2793655593 @default.
- W2888832568 cites W4235500388 @default.
- W2888832568 cites W4236382111 @default.
- W2888832568 cites W4240476789 @default.
- W2888832568 cites W4243035950 @default.
- W2888832568 cites W4250481301 @default.
- W2888832568 cites W70437322 @default.
- W2888832568 doi "https://doi.org/10.1145/3273982.3273986" @default.
- W2888832568 hasPublicationYear "2018" @default.
- W2888832568 type Work @default.
- W2888832568 sameAs 2888832568 @default.
- W2888832568 citedByCount "7" @default.
- W2888832568 countsByYear W28888325682019 @default.
- W2888832568 countsByYear W28888325682020 @default.
- W2888832568 countsByYear W28888325682021 @default.
- W2888832568 crossrefType "journal-article" @default.
- W2888832568 hasAuthorship W2888832568A5003307688 @default.
- W2888832568 hasAuthorship W2888832568A5005696647 @default.
- W2888832568 hasAuthorship W2888832568A5028280505 @default.
- W2888832568 hasAuthorship W2888832568A5036666743 @default.
- W2888832568 hasAuthorship W2888832568A5050695684 @default.
- W2888832568 hasAuthorship W2888832568A5057505785 @default.
- W2888832568 hasAuthorship W2888832568A5084751735 @default.
- W2888832568 hasConcept C111919701 @default.