Task #4172 (closed)
Opened 14 years ago
Closed 14 years ago
Make sure that JPEG-2000 decoding works in parallel
| Reported by: | mlinkert | Owned by: | mlinkert |
|---|---|---|---|
| Priority: | critical | Milestone: | OMERO-Beta4.3 |
| Component: | Bio-Formats | Version: | n.a. |
| Keywords: | n.a. | Cc: | callan, dzmacdonald |
| Resources: | n.a. | Referenced By: | n.a. |
| References: | n.a. | Remaining Time: | 0.0d |
| Sprint: | 2011-02-24 (6) |
Description
We would like to be able to decode JPEG-2000 tiles (as stored in histology formats) in parallel. So, we'll need to test if that works as-is, or if there are bugs that need fixing in the jai-imageio component.
Attachments (1)
Change History (7)
comment:1 Changed 14 years ago by jmoore
comment:2 Changed 14 years ago by mlinkert
- Remaining Time set to 1
- Sprint set to 2011-02-10 (5)
comment:3 Changed 14 years ago by jburel
- Sprint changed from 2011-02-10 (5) to 2011-02-24 (6)
Moved from sprint 2011-02-10 (5)
comment:4 Changed 14 years ago by mlinkert
- Remaining Time changed from 1 to 0
- Resolution set to fixed
- Status changed from new to closed
See attached test case, which checks the MD5s of tiles read in parallel against the MD5s of tiles read sequentially. All of the tested SVS files pass, so closing this ticket.
comment:5 Changed 14 years ago by jmoore
- Cc dzmacdonald added
Any observations on the speed up, Melissa?
Changed 14 years ago by mlinkert-x
comment:6 Changed 14 years ago by mlinkert
For me, sequential reading takes about twice as long as parallel reading on average. ParallelTileTest?.java has been updated to print out the number of milliseconds for sequential and parallel reading, so if you like you can see how well it performs.
imported from bio-formats:#600