MOC 50291 Developing High-performance Applications using Microsoft Windows HPC Server 2008- Seminare

• nach Absprache

Introduction to High-Performance Computing and HPC Server 2008

• Motivation for HPC
• Brief product history of CCS and HPCS
• Brief overview of HPC Server 2008 — components, job submission, scheduler
• Product differentiators
• Software development technologies:  parametric sweep, threads, OpenMP, MPI, SOA, etc.
• Measuring performance — linear speedup
• Predicting performance — Amdahl’s law

Multi-threading for Performance

• Multi-threading for responsiveness and performance
• The costs of multi-threading
• Structured, fork-join parallelism
• Multi-threading in C# using the .NET Thread class
• Multi-threading in VC++ using the Windows API
• Load balancing
• Scheduling multi-threaded apps on Windows HPC Server

The Dangers of Multithreading

• Race conditions
• Critical sections
• Starvation
• Livelock
• Deadlock
• Compiler and language implications
• Memory models
• Locks
• Interlocking
• Lock-free designs

The HPCS Job Scheduler

• Throughput vs. performance
• Nodes vs. sockets vs. cores
• Jobs vs. Tasks
• Job and task states
• Default scheduling policies
• The impact of job priorities and job preemption
• Job resources and dynamic growing / shrinking
• Submission and activation filters

Parallel Application Design

• Two sample design problems…
• Foster’s method
• Common problem decompositions
• Common communication patterns
• Computation vs. communication
• Design patterns:  master-worker, pipeline, map-reduce, SOA, parametric sweep, and more

Introduction to OpenMP

• What is OpenMP?
• Shared-memory programming
• Using OpenMP in Visual Studio with VC++
• Parallel regions
• Execution model
• Data parallelism
• Load balancing, static vs. dynamic scheduling
• Scheduling OpenMP apps on Windows HPC Server

Running and measuring performance on the cluster

• Barriers
• Critical sections
• Synchronization approaches
• Implementing common design patterns — conditional, task, master-worker, nested
• Data coherence and flushing
• Environment variables
• Common pitfalls

Introduction to the .NET Task Parallel Library

• What is the TPL?
• Moving from threads to tasks
• Using the TPL in Visual Studio with C#
• Execution model
• Parallel.For
• Data and task parallelism
• Synchronization approaches
• Concurrent data structures
• Scheduling TPL-based apps on Windows HPC Server

Interfacing with HPCS-based Clusters

• Cluster Manager
• Job Manager
• Job Description Files
• clusrun
• Console window
• PowerShell
• Scripts
• Programmatic access via HPCS API v2.0

Intro to SOA with HPC Server 2008

• Service-oriented architectures
• SOA and WCF
• Mapping SOA onto Jobs and the Job Scheduler
• Private vs. shared sessions
• Secure vs. insecure sessions

Create SOA-based Apps with HPC Server 2008

• Service-side programming
• Service configuration
• Client-side programming
• WCF configuration and tracing

General Performance Tuning of Parallel Applications

• Performance counters
• Heat map in Windows HPC Server 2008
• Customizing the heat map
• perfmon
• xperf (aka the Windows Performance Toolkit)
• SOA tuning
• What to look for…
• Other tools

Introduction to MPI

• Shared-memory vs. distributed-memory
• The essence of MPI programming — message-passing SPMD
• Microsoft MPI
• Using MSMPI in Visual Studio with VC++
• Execution model
• MPI Send and Receive
• mpiexec
• Scheduling MPI apps on Windows HPC Server

Data Parallelism and MPI’s Collective Operations

• Data parallelism in MPI
• A real world example
• Broadcast
• Scatter
• Gather
• Barriers
• Reductions
• Defining your own reduction operator
• Common pitfalls

MPI.NET

• Why MPI.NET?
• Using MPI.NET in Visual Studio with C#
• Type-safe Send and Receive
• Collective operations in MPI.NET
• Execution model
• Scheduling MPI.NET apps on Windows HPC Server

Using MPI — Debugging, Tracing, and Other Tools

• Local debugging with Visual Studio
• Remote debugging with Visual Studio
• General MPI tracing
• Tracing with ETW (Event Tracing for Windows)
• Trace visualization
• Other tools for MPI developers

Designing MPI Applications

• Hiding latency by overlapping computation and communication
• Avoiding deadlock
• Hybrid designs involving both MPI and OpenMP
• Buffering
• Error handling
• I/O and large datasets

MPI-2

• Groups
• Communicators
• Topologies
• Non-scalar data:  packing/unpacking, non-contiguous arrays, and user-defined datatypes
• MPI I/O
• Remote memory access
• [ Dynamic process creation is not supported in MSMPI ]

Excel-based HPC Apps

• Excel as a computation engine
• Performing Excel computations on Windows HPC Server 2008
• Using Excel Services
• Using Excel UDFs
• Future versions of Excel and HPC Server

Porting UNIX apps to Windows HPC Server 2008

• The most common porting issues
• 32-bit to 64-bit
• UNIX calls
• Manual porting of UNIX code
• Cygwin
• MinGW
• Microsoft SUA — Subsystem for UNIX-based Applications

Open Grid Forum HPC Basic Profile

• What is the OGF HPC Basic Profile?
• Platform-neutral job submission
• JSDL — Job Submission Description Language
• Enabling in Windows HPC Server 2008

Setup and Administration of Windows HPC Server 2008

• Hardware requirements
• Software requirements
• Initial decisions
• Headnode setup
• Compute node setup
• Broker node setup
• Developer machine setup
• Diagnostics
• Maintenance — including performance
• Troubleshooting