Difference between revisions of "Scan"
AngelaChen (talk | contribs) |
|||
Line 74: | Line 74: | ||
=Code To Implement= | =Code To Implement= | ||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
{{CodeToImplement|StepEfficientParallelSumScanner|sumScan|scan.exercise}} | {{CodeToImplement|StepEfficientParallelSumScanner|sumScan|scan.exercise}} | ||
{{Parallel|int[] sumScan(int[] data)}} | {{Parallel|int[] sumScan(int[] data)}} | ||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
=Visualization= | =Visualization= |
Revision as of 19:36, 13 March 2023
Contents
Motivation
Scan, also known as parallel prefix, is a fundamental and useful operation in parallel programming. We will gain experience in building Hillis & Steele scan with an optional work efficient Blellock scan.
Further, the dependencies in scan:
make it seem to have little hope for parallelism. However, simple yet clever approaches can achieve critical path lengths.
While we will simply implement prefix sum, scan can be used for other associative operations.
Background
Hillis-Steele Prefix Sum
(Extra Credit) Work-efficient
Exclusive (Blelloch)
Scan Primitives for Vector Computers
Prefix Sums and Their Applications
Inclusive
Lecture
Code To Use
class PhasableIntArrays extends AbstractPhasable
PhasableIntArrays
One of the downsides to parallel scan requires memory. For our scans we add the additional requirement that we will not mutate the incoming array parameter. We could create log(n) arrays, one for each level but that would be wasteful. If we create two array buffers and switch which is the source to read from and which is the destination to write to at each power-of-two level of the scan, we should be good to go. This is what PhasableIntArrays
helps us with. By using its getDstForPhase(int phaseIndex)
and getSrcForPhase(int phaseIndex)
, we can get the source and destination array corresponding to the phase.
The table below shows which of the two buffer arrays, a or b, will be passed back for each phase index as the source or the destination array.
phase | 0 | 1 | 2 | 3 | 4 |
src | data | a | b | a | b |
dst | a | b | a | b | a |
bonus info: power of 2 | 1 | 2 | 4 | 8 | 16 |
NOTE: think about which array is the correct one to return from your scan method given how you write your code.
PowersOf2Iterable
This studio should come in handy here. Recall that we implemented a PowersOfTwoLessThan
class in this studio.
Example usage:
for(int v : new PowersOfTwoLessThan(71)) { System.out.println(v); }
Code To Implement
class: | StepEfficientParallelSumScanner.java | |
methods: | sumScan | |
package: | scan.exercise | |
source folder: | student/src/main/java |
method: int[] sumScan(int[] data)
(parallel implementation required)
Visualization
class: | ScanApp.java | VIZ |
package: | scan.viz | |
source folder: | student/src/main/java |
Testing Your Solution
Correctness
Required
class: | __ScanTestSuite.java | |
package: | scan.exercise | |
source folder: | testing/src/test/java |
Optional Work Efficient
class: | _WorkEfficientScanTestSuite.java | |
package: | scan.challenge | |
source folder: | testing/src/test/java |