Bug ID: JDK-8101271 GUIMark2.Vector: performance of Prism HW pipeline is worse than Prism SW pipeline

Type: Enhancement
Component: javafx
Sub-Component: graphics
Affected Version: fx2.0

Priority: P3
Status: Resolved
Resolution: Cannot Reproduce

Submitted: 2011-01-12
Updated: 2015-06-16
Resolved: 2013-05-30

GUIMark2.Vector benchmark run with Prism HW pipeline (-Djavafx.toolkit=glass)
shows worse results comparing to Prism SW pipeline (-Djavafx.toolkit=glass -Dprism.order=j2d).

Here are the results for fx2.0-b10:
 Prism/sw   18.57
 Prism/hw    7.96


Aurora link:
 http://aurora.russia.sun.com/performance/faces/ChessBoard.xhtml?reportName=FX2&parameters=%5Brefrelease%5D2.0%5Brefbuild%5D%3D+-1%5Brefjdkrelease%5D1.6.0_23%5Brelease%5D%28pr.product.productRelease+IS+NOT+NULL%29%5Bbuild%5D%28pr.product.build+%3D+%2755%27%29%5Bjdkrelease%5D%28jdk.product.productRelease+%3D+%271.6.0_23%27%29&splitting=%5BX+axis%5DproductConf%5BComplement%5DjdkRelease%2C+jdkBuild%5BY+axis%5Dbenchmark%2C+benchmarkConf%2C+productRelease%2C+productBuild%5BZ+axis%5Dos%2C+hwclass%2C+jdk&mixReference=false&viewStyle=chessboard

The similar results for latest presidio-graphics-scrum hudson build 
(http://jfx.sfbay.sun.com/hudson/view/Presidio/job/presidio-graphics-scrum/1091/)

Looks like this one is done. Not sure where along the line it was fixed, so marking as "Cannot Reproduce"
30-05-2013
Results for fx2.1-b06: winxp-mid-range: 20.41 (hw) vs 26.7 (sw) win7-high-end: 20.86 (hw) vs 26.7 (sw) Running in fullspeed mode the difference increases up to 10 fps.
21-12-2011
Current Aurora builds as of today show: Prism/sw: 20.84 Prism/hw: 19.29 (19.46 in graphics scrum) The distance between Ductus and OpenPisces has shrunk some more.
25-04-2011
Lowering it back to Major. Kirill, please work with the team/engineer if you need to up the priority and provide the proper justification. All I see is this bug moved to critical on April 1st with no explanation. BTW, Jim is looking at the issue this is dependent on.
08-04-2011
The latest b21 results are: Prism/sw 20.84 Prism/hw 18.81 The diff was decreased from 10 fps to 2 fps which is very good progress. Note, 8 fps improvement is not only due to performance improvement in FX itself. 4 fps are coming from Guimark2.Vector test changes, see changeset here: http://jfxsrc.us.oracle.com/javafx/presidio/scrum/graphics/tests/rev/de3667267599 See Guimark2.Vector results for builds b19, b20 and b21 (the latest 2 used new version of the test) here: http://aurora.russia.sun.com/performance/faces/ChessBoard.xhtml?reportName=FX2&parameters=%5Brefrelease%5D2.0%5Brefbuild%5D%3D+%2795%27%5Brefjdkrelease%5D1.6.0_23%5Brelease%5D%28pr.product.productRelease+IS+NOT+NULL%29%5Bbuild%5D%28pr.product.build+%3D+%2795%27%29OR%28pr.product.build+%3D+%27101%27%29OR%28pr.product.build+%3D+%27105%27%29%5Bjdkrelease%5D%28jdk.product.productRelease+IS+NOT+NULL%29&splitting=%5BX+axis%5DfxConf%2C+jdkRelease%5BComplement%5Dbenchmark%2C+jdkBuild%2C+jdk%5BZ+axis%5Dos%2C+hwclass%2C+benchmarkSuite%5BY+axis%5DbenchmarkName%2C+benchmarkConf%2C+fxRelease%2C+fxBuild&reference=%5BOthers%5DfxRelease%2C+fxBuild%2C+jdkRelease%2C+jdkBuild%2C+jdk%2C+benchmarkSuite%2C+benchmarkName%5BReference+Set%5Dbenchmark%2C+os%2C+benchmarkConf%2C+fxConf%2C+hwclass&mixReference=false&flags=&significance=empty&hideDataConfiguration=false&calculateSummary=false&showSummaryExpanded=false&showSummaryContents=true&showComplementAttributes=false&compactTables=true&viewStyle=chessboard&filter=
05-04-2011
This may be related to the work Jim is doing for RT-8722
21-01-2011