If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
XP on newer platforms
Does 32 bit XP have the ability to actually use newer dual core and
quad core machines efficiently? |
Ads |
#2
|
|||
|
|||
XP on newer platforms
|
#3
|
|||
|
|||
XP on newer platforms
On Tue, 05 Jul 2016 18:35:58 -0400, Paul wrote:
wrote: Does 32 bit XP have the ability to actually use newer dual core and quad core machines efficiently? These are some benchmarks I ran a while back. You can see the WinXPx32 used as a baseline, does pretty well, when compared to the three other OSes on the lines just below it. SuperPI 7ZIP 1T 2T 4T (on 2C 2T E8400 CPU) 1.5XS 9.34 (WinXPx32) 3:19.546 21:32 12:09 11:16 Win7SP1x64 3:18.791 21:01 11:48 10:46 Win81x64 3:19.656 20:59 11:59 10:44 Win10x64 3:21.813 21:08 12:10 10:52 1T 12T 24T (on 6C 12T 4930K CPU) Win7SP1x64 1:59.028 15:22 2:09 2:00 Win81x64 1:59.197 15:29 2:09 1:58 Win10x64 2:00.044 15:38 2:10 2:01 It's hard to remember now, but I may have had a problem getting WinXP installed on the second machine. I expect with enough effort, it could be done. WinXP home supports one socket. WinXP Pro supports two sockets. You can have as many cores as you want on a socket (at least, until you run out of bits to handle them in Task Manager - the affinity dialog supports 32 cores). So from that perspective, the OS should have installed. There can be subtle difference in the scheduler behavior. The newer OSes understand the metrics (cost) of moving a task to a different core a bit better. In particular, you might get an extra 5-7 percent on one of those oddball multicore AMD processors. But short of dropping cores entirely, I don't think WinXP would be all that bad. And you can tell from the table, that Windows 7 isn't a bad alternative. As alternatives go. One reason for the Win10 results to be off, could be Win10 "doing its laundry" while I'm trying to benchmark it :-) If it had a "maintenance OFF" button, I'd have used it. I try not to cheat on these benchmarks, by compensating for the bad table manners of any OS (don't stack the deck by waiting for maintenance to finish or whatever). If an OS has bad table manners, it should suffer for it. Paul Thanks Paul. I was looking at some off lease 2 core duo 3.0mz machines and I just wanted to be sure they would run OK with XP. Some have a w7 p license, some don't |
#4
|
|||
|
|||
XP on newer platforms
|
#5
|
|||
|
|||
XP on newer platforms
|
#6
|
|||
|
|||
XP on newer platforms
|
#7
|
|||
|
|||
XP on newer platforms
|
Thread Tools | |
Display Modes | |
|
|