NDLI: FLEP: Enabling Flexible and Efficient Preemption on GPUs

Please wait, while we are loading the content...

Big Data Analytics and Intelligence at Alibaba Cloud

Determining Application-specific Peak Power and Energy Requirements for Ultra-low Power Processors

REDSPY: Exploring Value Locality in Software

Failure-Atomic Slotted Paging for Persistent Memory

ProRace: Practical Data Race Detection for Production Use

Black-box Concurrent Data Structures for NUMA Architectures

Optimizing CNNs on Multicores for Scalability, Performance and Goodput

Improving Datacenter Efficiency

DudeTM: Building Durable Transactions with Decoupling for Persistent Memory

Graspan: A Single-machine Disk-based Graph System for Interprocedural Static Analyses of Large-scale Systems Code

Efficient Address Translation for Architectures with Multiple Page Sizes

FLEP: Enabling Flexible and Efficient Preemption on GPUs

Identifying Security Critical Properties for the Dynamic Verification of a Processor

Bolt: I Know What You Did Last Summer... In The Cloud

Automated Synthesis of Comprehensive Memory Model Litmus Test Suites

Hardware-Software Co-design to Mitigate DRAM Refresh Overheads: A Case for Refresh-Aware Process Scheduling

Enabling Lightweight Transactions with Precision Time

Prophet: Precise QoS Prediction on Non-Preemptive Accelerators to Improve Utilization in Warehouse-Scale Computers

Translation-Triggered Prefetching

What Scalable Programs Need from Transactional Memory

Crossing Guard: Mediating Host-Accelerator Coherence Interactions

CoRAL: Confined Recovery in Distributed Asynchronous Graph Processing

Locality Transformations for Nested Recursive Iteration Spaces

ReFlex: Remote Flash ≈ Local Flash

SC-DCNN: Highly-Scalable Deep Convolutional Neural Network using Stochastic Computing

Page Fault Support for Network Controllers

SaberLDA: Sparsity-Aware Learning of Topic Models on GPUs

Verification of a Practical Hardware Security Architecture Through Static Information Flow Analysis

Neurosurgeon: Collaborative Intelligence Between the Cloud and Mobile Edge

DCatch: Automatically Detecting Distributed Concurrency Bugs in Cloud Systems

Kill the Program Counter: Reconstructing Program Behavior in the Processor Cache Hierarchy

IncBricks: Toward In-Network Computation with an In-Network Cache

Mallacc: Accelerating Memory Allocation

Typed Architectures: Architectural Support for Lightweight Scripting

TriCheck: Memory Model Verification at the Trisection of Software, Hardware, and ISA

An Architecture Supporting Formal and Compositional Binary Analysis

KickStarter: Fast and Accurate Computations on Streaming Graphs via Trimmed Approximations

Locality-Aware CTA Clustering for Modern GPUs

Approximate Storage of Compressed and Encrypted Videos

3DGates: An Instruction-Level Energy Analysis and Optimization of 3D Printers

Towards "Full Containerization" in Containerized Network Function Virtualization

Moonwalk: NRE Optimization in ASIC Clouds

CHERI JNI: Sinking the Java Security Model into the C

Thermostat: Application-transparent Page Management for Two-tiered Main Memory

Towards Practical Default-On Multi-Core Record/Replay

TETRIS: Scalable and Efficient Neural Network Acceleration with 3D Memory

AMNESIAC: Amnesic Automatic Computer

An Analysis of Persistent Memory Use with WHISPER

AsyncClock: Scalable Inference of Asynchronous Event Causality

Browsix: Bridging the Gap Between Unix and the Browser

Sound Loop Superoptimization for Google Native Client

Exploiting Intra-Request Slack to Improve SSD Performance

Dynamic Resource Management for Efficient Utilization of Multitasking GPUs

GRIFFIN: Guarding Control Flows Using Intel Processor Trace

Breaking the Boundaries in Heterogeneous-ISA Datacenters

Pallas: Semantic-Aware Checking for Finding Deep Bugs in Fast Path

History-Based Arbitration for Fairness in Processor-Interconnect of NUMA Servers

Voltage Regulator Efficiency Aware Power Management

FLEP: Enabling Flexible and Efficient Preemption on GPUs

Content Provider	ACM Digital Library
Author	Jiang, Changjun Liu, Xu Wu, Bo Zhou, Xiaobo
Abstract	GPUs are widely adopted in HPC and cloud computing platforms to accelerate general-purpose workloads. However, modern GPUs do not support flexible preemption, leading to performance and priority inversion problems in multi-tasking environments. In this paper, we propose and develop FLEP, the first software system that enables flexible kernel preemption and kernel scheduling on commodity GPUs. The FLEP compilation engine transforms the GPU program into preemptable forms, which can be interrupted during execution and yield all or part of the streaming multi-processors (SMs) in the GPU. The FLEP runtime engine intercepts all kernel invocations and determines which kernels and how those kernels should be preempted and scheduled. Experimental results on two-kernel co-runs demonstrate up to 24.2X speedup for high-priority kernels and up to 27X improvement on normalized average turnaround time for kernels with the same priority. FLEP reduces the preemption latency by up to 41% compared to yielding the whole GPU when the waiting kernels only need several SMs. With all the benefits, FLEP only introduces 2.5% runtime overhead, which is substantially lower than the kernel slicing approach.
Starting Page	483
Ending Page	496
Page Count	14
File Format	PDF
ISBN	9781450344654
DOI	10.1145/3037697.3037742
Language	English
Publisher	Association for Computing Machinery (ACM)
Publisher Date	2017-04-04
Publisher Place	New York
Access Restriction	Subscribed
Subject Keyword	Preemption Multi-tasking Gpgpu Kernel scheduling
Content Type	Text
Resource Type	Article

Sl.	Authority	Responsibilities	Communication Details
1	Ministry of Education (GoI), Department of Higher Education	Sanctioning Authority	https://www.education.gov.in/ict-initiatives
2	Indian Institute of Technology Kharagpur	Host Institute of the Project: The host institute of the project is responsible for providing infrastructure support and hosting the project	https://www.iitkgp.ac.in
3	National Digital Library of India Office, Indian Institute of Technology Kharagpur	The administrative and infrastructural headquarters of the project	Dr. B. Sutradhar bsutra@ndl.gov.in
4	Project PI / Joint PI	Principal Investigator and Joint Principal Investigators of the project	Dr. B. Sutradhar bsutra@ndl.gov.in Prof. Saswat Chakrabarti will be added soon
5	Website/Portal (Helpdesk)	Queries regarding NDLI and its services	support@ndl.gov.in
6	Contents and Copyright Issues	Queries related to content curation and copyright issues	content@ndl.gov.in
7	National Digital Library of India Club (NDLI Club)	Queries related to NDLI Club formation, support, user awareness program, seminar/symposium, collaboration, social media, promotion, and outreach	clubsupport@ndl.gov.in
8	Digital Preservation Centre (DPC)	Assistance with digitizing and archiving copyright-free printed books	dpc@ndl.gov.in
9	IDR Setup or Support	Queries related to establishment and support of Institutional Digital Repository (IDR) and IDR workshops	idr@ndl.gov.in

FLEP:Enabling Flexible and Efficient Preemption on GPUs (CARN, Vol. 45, Issue. 1)

FLEP:Enabling Flexible and Efficient Preemption on GPUs (SIGP, Vol. 52, Issue. 4)

FLEP:Enabling Flexible and Efficient Preemption on GPUs (OPSR, Vol. 51, Issue. 2)

A translation system for enabling data mining applications on GPUs

Towards Multi-tenant GPGPU : Event-driven Programming Model for System-wide Scheduling on Shared GPUs

A compiler and runtime system for enabling data mining applications on gpus

Enabling and Exploiting Flexible Task Assignment on GPU through SM-Centric Program Transformations

Efficient Implementation of Evaluating Multivariate Quadratic System with GPUs

Warped-slicer: efficient intra-SM slicing through dynamic resource partitioning for GPU multiprogramming

FLEP: Enabling Flexible and Efficient Preemption on GPUs

Similar Documents

FLEP:Enabling Flexible and Efficient Preemption on GPUs (CARN, Vol. 45, Issue. 1)

FLEP:Enabling Flexible and Efficient Preemption on GPUs (SIGP, Vol. 52, Issue. 4)

FLEP:Enabling Flexible and Efficient Preemption on GPUs (OPSR, Vol. 51, Issue. 2)

A translation system for enabling data mining applications on GPUs

Towards Multi-tenant GPGPU : Event-driven Programming Model for System-wide Scheduling on Shared GPUs

A compiler and runtime system for enabling data mining applications on gpus

Enabling and Exploiting Flexible Task Assignment on GPU through SM-Centric Program Transformations

Efficient Implementation of Evaluating Multivariate Quadratic System with GPUs

Warped-slicer: efficient intra-SM slicing through dynamic resource partitioning for GPU multiprogramming

FLEP: Enabling Flexible and Efficient Preemption on GPUs