Matches in SemOpenAlex for { <https://semopenalex.org/work/W3183550499> ?p ?o ?g. }
Showing items 1 to 91 of
91
with 100 items per page.
- W3183550499 abstract "Modern computing platforms tend to deploy multiple GPUs (2, 4, or more) on a single node to boost system performance, with each GPU having a large capacity of global memory and streaming multiprocessors (SMs). GPUs are an expensive resource, and boosting utilization of GPUs without causing performance degradation of individual workloads is an important and challenging problem. Although services like MPS support simultaneous execution of multiple co-operative kernels on a single device, they do not solve the above problem for uncooperative kernels, MPS being oblivious to the resource needs of each kernel. We propose a fully automated compiler-assisted scheduling framework. The compiler constructs GPU tasks by identifying kernel launches and their related GPU operations (e.g. memory allocations). For each GPU task, a probe is instrumented in the host-side code right before its launch point. At runtime, the probe conveys the information about the task's resource requirements (e.g. memory and compute cores) to a scheduler, such that the scheduler can place the task on an appropriate device based on the task's resource requirements and devices' load in a memory-safe, resource-aware manner. To demonstrate its advantages, we prototyped a throughput-oriented scheduler based on the framework, and evaluated it with the Rodinia benchmark suite and the Darknet neural network framework on NVIDIA GPUs. The results show that the proposed solution outperforms existing state-of-the-art solutions by leveraging its knowledge about applications' multiple resource requirements, which include memory as well as SMs. It improves throughput by up to 2.5x for Rodinia benchmarks, and up to 2.7x for Darknet neural networks. In addition, it improves job turnaround time by up to 4.9x, and limits individual kernel performance degradation to at most 2.5%." @default.
- W3183550499 created "2021-08-02" @default.
- W3183550499 creator A5049228142 @default.
- W3183550499 creator A5056248574 @default.
- W3183550499 creator A5061235810 @default.
- W3183550499 date "2021-07-18" @default.
- W3183550499 modified "2023-10-17" @default.
- W3183550499 title "Effective GPU Sharing Under Compiler Guidance" @default.
- W3183550499 cites W1506474685 @default.
- W3183550499 cites W1596936080 @default.
- W3183550499 cites W1651324627 @default.
- W3183550499 cites W1713367009 @default.
- W3183550499 cites W1969701927 @default.
- W3183550499 cites W2005574683 @default.
- W3183550499 cites W2068681864 @default.
- W3183550499 cites W2078994750 @default.
- W3183550499 cites W2080592089 @default.
- W3183550499 cites W2097643185 @default.
- W3183550499 cites W2117539524 @default.
- W3183550499 cites W2125551452 @default.
- W3183550499 cites W2141992894 @default.
- W3183550499 cites W2153375074 @default.
- W3183550499 cites W2186615578 @default.
- W3183550499 cites W2194775991 @default.
- W3183550499 cites W2323909431 @default.
- W3183550499 cites W2604787577 @default.
- W3183550499 cites W2767239597 @default.
- W3183550499 cites W2886481383 @default.
- W3183550499 cites W2899071864 @default.
- W3183550499 cites W2903278032 @default.
- W3183550499 cites W2962835968 @default.
- W3183550499 cites W2964330525 @default.
- W3183550499 cites W2970971581 @default.
- W3183550499 cites W3098136731 @default.
- W3183550499 doi "https://doi.org/10.48550/arxiv.2107.08538" @default.
- W3183550499 hasPublicationYear "2021" @default.
- W3183550499 type Work @default.
- W3183550499 sameAs 3183550499 @default.
- W3183550499 citedByCount "0" @default.
- W3183550499 crossrefType "posted-content" @default.
- W3183550499 hasAuthorship W3183550499A5049228142 @default.
- W3183550499 hasAuthorship W3183550499A5056248574 @default.
- W3183550499 hasAuthorship W3183550499A5061235810 @default.
- W3183550499 hasBestOaLocation W31835504991 @default.
- W3183550499 hasConcept C111919701 @default.
- W3183550499 hasConcept C114614502 @default.
- W3183550499 hasConcept C118524514 @default.
- W3183550499 hasConcept C120314980 @default.
- W3183550499 hasConcept C13280743 @default.
- W3183550499 hasConcept C162324750 @default.
- W3183550499 hasConcept C169590947 @default.
- W3183550499 hasConcept C173608175 @default.
- W3183550499 hasConcept C185798385 @default.
- W3183550499 hasConcept C205649164 @default.
- W3183550499 hasConcept C206729178 @default.
- W3183550499 hasConcept C21547014 @default.
- W3183550499 hasConcept C33923547 @default.
- W3183550499 hasConcept C41008148 @default.
- W3183550499 hasConcept C74193536 @default.
- W3183550499 hasConceptScore W3183550499C111919701 @default.
- W3183550499 hasConceptScore W3183550499C114614502 @default.
- W3183550499 hasConceptScore W3183550499C118524514 @default.
- W3183550499 hasConceptScore W3183550499C120314980 @default.
- W3183550499 hasConceptScore W3183550499C13280743 @default.
- W3183550499 hasConceptScore W3183550499C162324750 @default.
- W3183550499 hasConceptScore W3183550499C169590947 @default.
- W3183550499 hasConceptScore W3183550499C173608175 @default.
- W3183550499 hasConceptScore W3183550499C185798385 @default.
- W3183550499 hasConceptScore W3183550499C205649164 @default.
- W3183550499 hasConceptScore W3183550499C206729178 @default.
- W3183550499 hasConceptScore W3183550499C21547014 @default.
- W3183550499 hasConceptScore W3183550499C33923547 @default.
- W3183550499 hasConceptScore W3183550499C41008148 @default.
- W3183550499 hasConceptScore W3183550499C74193536 @default.
- W3183550499 hasLocation W31835504991 @default.
- W3183550499 hasOpenAccess W3183550499 @default.
- W3183550499 hasPrimaryLocation W31835504991 @default.
- W3183550499 hasRelatedWork W1541585229 @default.
- W3183550499 hasRelatedWork W1583465708 @default.
- W3183550499 hasRelatedWork W1601646354 @default.
- W3183550499 hasRelatedWork W1604898313 @default.
- W3183550499 hasRelatedWork W1853049011 @default.
- W3183550499 hasRelatedWork W2078700326 @default.
- W3183550499 hasRelatedWork W2147654880 @default.
- W3183550499 hasRelatedWork W4235959758 @default.
- W3183550499 hasRelatedWork W4245265375 @default.
- W3183550499 hasRelatedWork W2479014312 @default.
- W3183550499 isParatext "false" @default.
- W3183550499 isRetracted "false" @default.
- W3183550499 magId "3183550499" @default.
- W3183550499 workType "article" @default.