Originally posted by: Markfw900
The X2's in rares case do, but not if you install the X2 driver and sometimes also need the "hotfix".
In the 1 in a million cases where that doesn't fix it, set affinity, and it always works, there really is no problem.
Originally posted by: FireChicken
I have heard of stuttering problems with AMD and INtel dual core chips. If you are taking gaming wise. They were due to a SMP being written into the games and also dual core optimizations in the viedo drivers. These conflicted with each other and caused stuttering. This was fixed by editing your video registry and removing the dual core usage in the video drivers.
I got this from a quake 4 gudie but it might help u.
http://ucguides.savagehelp.com/Quake4/FAQ_DualCore.htm
http://hardforum.com/showthread.php?t=983781&highlight=hotfix
Originally posted by: Markbnj
Seems like most of these issues are caused by one simple problem. There is a timer/counter on the CPU that threads can read. Games and some other kinds of software often use this counter to determine how much time has passed since the program last got scheduled to do work. This lets games, for example, correctly calculate how much the world has changed between time slices. The problem on a dual core is that the thread may get scheduled on one core, then get suspended, then get scheduled on the other core. When it reads the timer the second time it will get a value from a totally different register, which almost certainly won't be the value expected. In some cases the value might suggest a much longer time period has elapsed, so you get a hitch or stutter. In other cases it might suggest that a negative amount of time has passed, and this can cause crashes. In still other cases the counters might be in close enough synch that it doesn't matter at all.
AMD has a "dual core optimizer" fix on their site that essentially synchs these counters at startup, so a thread can read from either and not have a problem. I don't know how Intel handles it.
Originally posted by: LW07
Originally posted by: Markbnj
Seems like most of these issues are caused by one simple problem. There is a timer/counter on the CPU that threads can read. Games and some other kinds of software often use this counter to determine how much time has passed since the program last got scheduled to do work. This lets games, for example, correctly calculate how much the world has changed between time slices. The problem on a dual core is that the thread may get scheduled on one core, then get suspended, then get scheduled on the other core. When it reads the timer the second time it will get a value from a totally different register, which almost certainly won't be the value expected. In some cases the value might suggest a much longer time period has elapsed, so you get a hitch or stutter. In other cases it might suggest that a negative amount of time has passed, and this can cause crashes. In still other cases the counters might be in close enough synch that it doesn't matter at all.
AMD has a "dual core optimizer" fix on their site that essentially synchs these counters at startup, so a thread can read from either and not have a problem. I don't know how Intel handles it.
So all I have to do is install the optimizer and I'll be good to go if I get the X2?
Originally posted by: Markbnj
Originally posted by: LW07
Originally posted by: Markbnj
Seems like most of these issues are caused by one simple problem. There is a timer/counter on the CPU that threads can read. Games and some other kinds of software often use this counter to determine how much time has passed since the program last got scheduled to do work. This lets games, for example, correctly calculate how much the world has changed between time slices. The problem on a dual core is that the thread may get scheduled on one core, then get suspended, then get scheduled on the other core. When it reads the timer the second time it will get a value from a totally different register, which almost certainly won't be the value expected. In some cases the value might suggest a much longer time period has elapsed, so you get a hitch or stutter. In other cases it might suggest that a negative amount of time has passed, and this can cause crashes. In still other cases the counters might be in close enough synch that it doesn't matter at all.
AMD has a "dual core optimizer" fix on their site that essentially synchs these counters at startup, so a thread can read from either and not have a problem. I don't know how Intel handles it.
So all I have to do is install the optimizer and I'll be good to go if I get the X2?
I would never make such a blanket statement. But if you have lockups due to that specific issue the optimizer will cure them.
I don't have the X2 or any dual core CPU yet(I'm still on my 3500+), but I'm asking this so I can refer to this when at around Christmas when I get the X2. Are the issues only confined to the first X2s, or are the X2s out today still having the problem?
And how would I tell what was causing the lockups?