aivika-distributed 0.7.4 → 0.7.4.1

raw patch · 2 files changed

+15/−6 lines, 2 filesPVP ok

version bump matches the API change (PVP)

API changes (from Hackage documentation)

Files

• CHANGELOG.md +5/−0
• aivika-distributed.cabal +10/−6

@@ -1,4 +1,9 @@ +Version 0.7.4.1+-----++* Updated the estimaton of speed in the description after recent changes in the sequential module.+ Version 0.7.4 ----- 

@@ -1,5 +1,5 @@ name:            aivika-distributed-version:         0.7.4+version:         0.7.4.1 synopsis:        Parallel distributed discrete event simulation module for the Aivika library description:     This package extends the aivika-transformers [1] package and allows running parallel distributed simulations.@@ -17,7 +17,7 @@     is that this is the optimistic distributed simulation with possible rollbacks. It is assumed that optimistic methods      tend to better support the parallelism inherited in the models.      .-    You may test the distributed simulation using your own laptop only, although the package is still destined to be +    You can test the distributed simulation using your own laptop only, although the package is still destined to be      used with a multi-core computer, or computers connected in the distributed cluster.     .     There are additional packages that allow you to run the distributed simulation experiments by @@ -26,13 +26,17 @@     Please consult the AivikaSoft [3] website for more details.     .     Regarding the speed of simulation, the rough estimation is as follows. The distributed simulation module is slower up to-    6-9 times in comparison with the sequential aivika [2] simulation library using the equivalent sequential models. +    15-40 times in comparison with the sequential aivika [2] simulation library using the equivalent sequential models.+    The estimation has dramatically changed after started using another more fast pseudo-random number generator by default,+    which made the sequential module even more fast. The lower estimation is likely to correspond to complex models. +    The upper estimation will probably correspond to quite simple event-oriented and process-oriented models, where +    the sequential module can be exceptionally fast. +    .     Note that you can run up to 7 parallel logical processes on a single 8-core processor computer and run the Time Server      process too. On a 36-core processor, you can launch up to 35 logical processes simultaneously.     . -    So, this estimation seems to be quite good. At the same time, the message passing between the logical processes can -    dramatically decrease the speed of distributed simulation, especially if they cause rollbacks. Thus, much depends on -    the distributed model itself.+    At the same time, the message passing between the logical processes can dramatically decrease the speed of distributed +    simulation, especially if they cause rollbacks. Thus, much depends on the distributed model itself.     .     Finally, you can use the following test model [4] as an example.     .