Presentation is loading. Please wait.

Presentation is loading. Please wait.

High Performance Messaging on Workstations

Similar presentations


Presentation on theme: "High Performance Messaging on Workstations"— Presentation transcript:

1 High Performance Messaging on Workstations
Illinois Fast Messages (FM) for Myrinet Presented By: Scott Thomas 11/12/2018 Cal Poly Network Performance Research Group

2 Cal Poly Network Performance Research Group
Introduction Workstations clusters Advantages Lower Cost Larger Software Base Greater Accessibility Disadvantages Protocol Overhead Buffer Management Operating System Overhead 11/12/2018 Cal Poly Network Performance Research Group

3 Cal Poly Network Performance Research Group
Background Previous methods: Mapping network device interface into user space directly via load/store operations Myrinet Network Features Uses LANai network coprocessor (5 MIPS) 3 DMA engines (in, out, host) Workstation Features Memory bus is fast, I/O Bus is slow! 11/12/2018 Cal Poly Network Performance Research Group

4 Cal Poly Network Performance Research Group
11/12/2018 Cal Poly Network Performance Research Group

5 Cal Poly Network Performance Research Group
Critical Issues Critical Issues Division of labor between the host and network coprocessor Assign as much functionality to the host as possible Management of the input/output bus Use DMA as much as possible Buffer management Minimal amount of queues 11/12/2018 Cal Poly Network Performance Research Group

6 The Fast Messages Approach
Illinois Fast Messages 1.0 methods FM_send_4() - Send a four word message FM_send() - Send a longer message FM_extract() - Process a received message 11/12/2018 Cal Poly Network Performance Research Group

7 Cal Poly Network Performance Research Group
FM Implementation FM 1.0 Host Program LANai Control Program (LCP) Message Transmission 1. Get data to the LANai 2. Put the data on the line 3. Remove data from the line at receiver 4. Transport data to host memory 11/12/2018 Cal Poly Network Performance Research Group

8 Network Coprocessor Program
Baseline vs. Streamed 11/12/2018 Cal Poly Network Performance Research Group

9 Cal Poly Network Performance Research Group
SBus Management Methods of Interaction all-DMA outgoing: host copies data into DMA, give LANai pointers, triggers send incoming: uses pointer to write data into host memory hybrid outgoing: host copies data into LANai’s memory, triggers send incoming: DMAs messages into host memory 11/12/2018 Cal Poly Network Performance Research Group

10 Cal Poly Network Performance Research Group
SBus Management 11/12/2018 Cal Poly Network Performance Research Group

11 Buffer Management / Flow Control
4 queues LANai send LANai receive host receive host reject Return-to-sender flow control Receiver returns packet to sender if no memory available 11/12/2018 Cal Poly Network Performance Research Group

12 Comparative Performance
Adding small features can cause huge performance hits. 11/12/2018 Cal Poly Network Performance Research Group

13 Comparative Performance
11/12/2018 Cal Poly Network Performance Research Group

14 Cal Poly Network Performance Research Group
Conclusion Improvements improved I/O bus performance for non-DMA operations faster network interface processor 11/12/2018 Cal Poly Network Performance Research Group

15 Cal Poly Network Performance Research Group
11/12/2018 Cal Poly Network Performance Research Group


Download ppt "High Performance Messaging on Workstations"

Similar presentations


Ads by Google