MapReduce for the Cell B. E. Architecture

1
MapReduce for the Cell B. E. Architecture
Marc de Kruijf University of Wisconsin-Madison Ad
vised by Professor Sankaralingam
2
MapReduce

• A model for parallel programming
• Proposed by Google
• Large scale distributed systems 1,000 node
  clusters
• Applications
• Distributed sort
• Distributed grep
• Indexing
• Simple, high-level interface
• Runtime handles
• parallelization, scheduling, synchronization, and
  communication

3
Cell B. E. Architecture

• A heterogeneous computing platform
• 1 PPE, 8 SPEs
• Programming is hard
• Multi-threading is explicit
• SPE local memories are software-managed
• The Cell is like a cluster-on-a-chip

4
Motivation

• MapReduce
• Scalable parallel model
• Simple interface

• Cell B. E.
• Complex parallel architecture
• Hard to program

MapReduce for the Cell B.E. Architecture
5
Overview

• Motivation
• MapReduce
• Cell B.E. Architecture
• MapReduce Example
• Design
• Evaluation
• Workload Characterization
• Application Performance
• Conclusions and Future Work

6
MapReduce Example

• Counting word occurrences in a set of documents

7
Overview

• Motivation
• MapReduce
• Cell B.E. Architecture
• MapReduce Example
• Design
• Evaluation
• Workload Characterization
• Application Performance
• Conclusions and Future Work

8
Design
Flow of Execution Five stages Map, Partition,
Quick-sort, Merge-sort, Reduce
9
Design
Flow of Execution Five stages Map, Partition,
Quick-sort, Merge-sort, Reduce 1. Map streams
key/value pairs
10
Design

• Flow of Execution
• Five stages Map, Partition, Quick-sort,
  Merge-sort, Reduce
• 1. Map streams key/value pairs
• Key grouping implemented as
• 2. Partition hash and distribute
• 3. Quick-sort
• 4. Merge-sort

two-phase external sort
11
Design

• Flow of Execution
• Five stages Map, Partition, Quick-sort,
  Merge-sort, Reduce
• 1. Map streams key/value pairs
• Key grouping implemented as
• 2. Partition hash and distribute
• 3. Quick-sort
• 4. Merge-sort

two-phase external sort
12
Design

• Flow of Execution
• Five stages Map, Partition, Quick-sort,
  Merge-sort, Reduce
• 1. Map streams key/value pairs
• Key grouping implemented as
• 2. Partition hash and distribute
• 3. Quick-sort
• 4. Merge-sort

two-phase external sort
13
Design

• Flow of Execution
• Five stages Map, Partition, Quick-sort,
  Merge-sort, Reduce
• 1. Map streams key/value pairs
• Key grouping implemented as
• 2. Partition hash and distribute
• 3. Quick-sort
• 4. Merge-sort
• 5. Reduce reduces
• key/list-of-values pairs to
• key/value pairs.

two-phase external sort
14
Overview

• Motivation
• MapReduce
• Cell B.E. Architecture
• MapReduce Example
• Design
• Evaluation
• Workload Characterization
• Application Performance
• Conclusions and Future Work

15
Evaluation Methodology

• MapReduce Model Characterization
• Synthetic micro-benchmark with six parameters
• Run on a 3.2 GHz Cell Blade
• Measured effect of each parameter on execution
  time
• Application Performance Comparison
• Six full applications
• MapReduce versions run on 3.2 GHz Cell Blade
• Single-threaded versions run on 2.4 GHz Core 2
  Duo
• Evaluation
• Measured speedup comparing execution times
• Measured overheads on the Cell monitoring SPE
  idle time
• Measured ideal speedup assuming no Cell overheads

16
MapReduce Model Characterization

• Model Characteristics

Effect on Execution Time
Characteristic Description
Map intensity Execution cycles per input byte to Map
Reduce intensity Execution cycles per input byte to Reduce
Map fan-out Ratio of input size to output size in Map
Reduce fan-in Number of values per key in Reduce
Partitions Number of partitions
Input size Input size in bytes
17
Application Performance

• Applications
• histogram counts bitmap RGB occurrences
• kmeans clustering algorithm
• linearReg least-squares linear regression
• wordCount word count
• NAS_EP EP benchmark from NAS suite
• distSort distributed sort

18
Speedup Over Core 2 Duo
19
Runtime Overheads
20
Overview

• Motivation
• MapReduce
• Cell B.E. Architecture
• MapReduce Example
• Design
• Evaluation
• Workload Characterization
• Application Performance
• Conclusions and Future Work

21
Conclusions and Future Work

• Conclusions
• Programmability benefits
• High-performance on computationally intensive
  workloads
• Not applicable to all application types
• Future Work
• Additional performance tuning
• Extend for clusters of Cell processors
• Hierarchical MapReduce

22
Questions?
23
Backup Slides
24
MapReduce API

• void MapReduce_exec(MapReduce Specification
  specification)
• The exec function initializes the MapReduce
  runtime and executes MapReduce according to the
  user specification.
• void MapReduce_emitIntermediate(void key, void
  value)
• void MapReduce_emit(void value)
• These two functions are called by the
  user-defined Map and Reduce functions,
  respectively. These functions take references to
  pointers as arguments, and modify the referenced
  pointer to point to pre-allocated storage. It is
  then the responsibility of the application to
  provision this storage.

25
Optimizations

• Priority work queue
• Distributes load
• Avoids serialization
• Pipelined execution maximizes concurrency
• Double-buffering
• Application support
• Map only
• Map with sorted output
• Chaining invocations

26
Optimizations

• Priority work queue
• Distributes load
• Avoids serialization
• Pipelined execution maximizes concurrency
• Double-buffering
• Application support
• Map only
• Map with sorted output
• Chaining invocations

27
Optimizations

• Balanced merge (n / log(n) better bandwidth
  utilization as n ? 8)
• Map and Reduce output regions pre-allocated.
• optimal memory alignment
• bulk memory transfers
• no user memory management
• no dynamic allocation overhead