R9.1 MFD refresh/ redraw delay

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic

   I have experienced and many other pilots on COPA forums have reported a terrible slow down in the MFD refresh rate with R9.1.  This gives serious pause.  Is this a function of software bloat as experienced by many computer users when program have outgrown their hardware base?  Is the memory buffer assigned to map or weather downloads sufficient? Are you trying to incorporate too many "cool" features such that the system stagnates?  Is this speed issue related to the distance to altitude designation we just "lost" in R9.1?  Is the problem a function of bus speed, or, videocard limitations?  How can such a slow system ever handle three dimensional rewrites in SVT?  The problem has been hanging there--is Avidyne going to address this potentially serious concern? 

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic
Harvey Members Profile Send Private Message Find Members Posts Add to Buddy List Release 9 Insiders Joined: 18 May 2009 Location: Tampa, FL. Status: Offline Points: 18	Post Options Post Reply Quote Harvey Report Post Thanks(0) Quote Reply Topic: R9.1 MFD refresh/ redraw delay Posted: 08 Jul 2010 at 7:17pm
	I have experienced and many other pilots on COPA forums have reported a terrible slow down in the MFD refresh rate with R9.1. This gives serious pause. Is this a function of software bloat as experienced by many computer users when program have outgrown their hardware base? Is the memory buffer assigned to map or weather downloads sufficient? Are you trying to incorporate too many "cool" features such that the system stagnates? Is this speed issue related to the distance to altitude designation we just "lost" in R9.1? Is the problem a function of bus speed, or, videocard limitations? How can such a slow system ever handle three dimensional rewrites in SVT? The problem has been hanging there--is Avidyne going to address this potentially serious concern?

AviMike Members Profile Send Private Message Find Members Posts Add to Buddy List Admin Group Joined: 24 May 2010 Location: Melbourne, FL Status: Offline Points: 33	Post Options Post Reply Quote AviMike Report Post Thanks(0) Quote Reply Posted: 03 Aug 2010 at 8:58am
	The map performance slowdown is from our newest function, Procedure Preview. The Preview turns terrain on and off as the flight plan is being edited. Loading terrain from the compact flash is a bottleneck in Release 9.1. Engineering has been working on it for the last few months, even before R9.1. The demand is much higher when Synthetic Vision is requiring higher resolution terrain at higher update rates. There are a couple of significant improvements that you’ll see in the next releases. The first improvement is the speed at which data is passed from the Compact Flash to the graphics processing unit (GPU). Due to this improvement the bus is no longer the bottleneck, it becomes the access speed of the compact flash, which is quite fast. This more direct terrain transfer also reduces the CPU workload required to pass data so we gain back CPU power as well. The second improvement is the speed at which we draw water (oceans and inland water). We changed it from a large number of polygons to being a layer in the terrain. This also frees up the CPU workload and gives more time to other map features than to drawing water. This will be implemented as part of the Synthetic Vision release (i.e. R9.3). There are some other minor tweaks but the these two improvements will make the map back to what it was prior to R9.1, and hopefully faster. If there is anything that we find our prior to a release that will improve the map speed then I’ll let you know.