|
| Notices |
Welcome to the DriverHeaven.net forums. You are currently viewing our boards as a guest which gives you limited access to view most discussions and access our other features. By joining our free community you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content and access many other special features. Registration is fast, simple and absolutely free so please, join our community today! If you have any problems with the registration process or your account login, please contact contact us. |
 |
Mar 20, 2004, 05:17 PM
|
#1
|
|
Burned
Join Date: May 2002
Posts: 29,662
|
P4 Hyperthreading off/on
Anyone ran benchies on their P4 with hyperthreading off and on? the P4 overclocks higher with hyperthreading off, but i found a massive drop in sandra arithmetic and multimedia tests. pcmark02 and 3dmark01 are fine.. just those specific sandra tests.
im sure this is well known, but its the first time ive ran with HT off......
|
|
|
Mar 20, 2004, 05:28 PM
|
#2
|
|
DH's Dormant Dragon
Join Date: May 2002
Location: IN Rem-Dormancy
Posts: 23,663
|
saposedly... game preformance will increase....
|
|
|
Mar 20, 2004, 05:30 PM
|
#3
|
|
DriverHeaven Extreme Member
Join Date: May 2003
Location: Virginia, USA
Posts: 3,567
|
Quote:
Originally posted by Judas
saposedly... game preformance will increase....
|
Game performance increase off or on?
|
|
|
Mar 20, 2004, 05:34 PM
|
#4
|
|
Burned
Join Date: May 2002
Posts: 29,662
|
well 3dmark01 should be the indicator of that to a certain extent and I saw no differences, except you can overclock 50-100mhz higher with HT off. unfortunately the sandra benchies show that the threading is obviously doing something to those tests........
arithmetic dropped from high 11k to high 9k and I normally get 30/42k in multimedia, that dropped quite a bit too. everything else I tried was the same (of course 70mhz higher).... interesting.
|
|
|
Mar 20, 2004, 07:26 PM
|
#5
|
|
DH's Dormant Dragon
Join Date: May 2002
Location: IN Rem-Dormancy
Posts: 23,663
|
nice to see Ut2004 results.... still very much interesting...
|
|
|
Mar 20, 2004, 07:48 PM
|
#6
|
|
DriverHeaven Extreme Member
Join Date: May 2003
Location: Virginia, USA
Posts: 3,567
|
I saw this bench somewhere when HT was first shown off and tested and what it showed was that either the game or application stayed the same or improved without a loss. So for a game or Benchmark to get better with it off doens't make any sense, especially since you'd think 2 cpu's would be better than one.
|
|
|
Mar 20, 2004, 08:41 PM
|
#7
|
|
DriverHeaven Extreme Member
Join Date: May 2002
Location: Ohio
Posts: 8,479
|
As far as I understand same games do use HT but for the life of me I'm not sure which ones. Besides what is the point of leaving it off? You can overclock a bit faster but lose some of the dual sex appeal. HT is one of the reasons I even went the route I did.
1. FSB
2. HT
Well 3..Video Encoding.
|
|
|
Mar 20, 2004, 08:47 PM
|
#8
|
|
DriverHeaven Extreme Member
Join Date: May 2003
Location: Virginia, USA
Posts: 3,567
|
Quote:
Originally posted by Necrosis
As far as I understand same games do use HT but for the life of me I'm not sure which ones. Besides what is the point of leaving it off? You can overclock a bit faster but lose some of the dual sex appeal. HT is one of the reasons I even went the route I did.
1. FSB
2. HT
Well 3..Video Encoding.
|
The reasons why I may go that route this summer.
|
|
|
Mar 20, 2004, 08:54 PM
|
#9
|
|
DriverHeaven Extreme Member
Join Date: May 2002
Location: Ohio
Posts: 8,479
|
Eventually I will be going back to AMD no doubt but heck I can still keep this around for specific work, and heck have something for folks to game at home /w me.
|
|
|
Mar 21, 2004, 01:26 AM
|
#10
|
|
DH's Latest Mac Convert
Join Date: Jun 2003
Location: Basement of the first floor
Posts: 15,630
|
the program would have to be optimised to use 2 CPU's to make use of HT technology
|
|
|
Mar 21, 2004, 04:07 AM
|
#11
|
|
Burned
Join Date: May 2002
Posts: 29,662
|
|
|
|
Mar 21, 2004, 04:25 AM
|
#12
|
|
confutatis maledictis
Join Date: May 2002
Location: somewhere dark
Posts: 5,952
|
|
|
|
Mar 21, 2004, 04:56 AM
|
#13
|
|
Burned
Join Date: May 2002
Posts: 29,662
|
some nice linkages indeed. Its been interesting to find out you can overclock the processor higher with HT off, but I dont see any benefits really from what I have done, except 3dmark which scores higher with HT off (big deal who cares about 3dmark01 anymore anyway). sandra scores lower, real world performance well its hard to tell I cant see much difference because the machine does everything virtually instantly with it on or off. I guess for multitasking HT on is a benefit as it breaks tasks up between the virtual processors.
|
|
|
Mar 21, 2004, 05:18 AM
|
#14
|
|
confutatis maledictis
Join Date: May 2002
Location: somewhere dark
Posts: 5,952
|
might as well leave it on then, for those multi-tasking situations. Have you tried any rendering with it on or off?
|
|
|
Mar 21, 2004, 05:33 AM
|
#15
|
|
Burned
Join Date: May 2002
Posts: 29,662
|
yes certain applications use it, quite substantial differences in some cases, even some photoshop filters are 30% faster, ive ran tests this morning. its staying on for sure.
|
|
|
Mar 21, 2004, 07:15 PM
|
#16
|
|
Delete Me
Join Date: Mar 2004
Posts: 14,676
|
HT is amazing...why would anyone turn it off...and as near as I can tell OCing differences are miniscule as I have not yet found them (i'm about 500 MHz over stock on aircooling with HT enabled, and I can hit 750 over stock, btu temperatures begin to get iffy when i load it up). I use gordian Knot and VirtualDub a lot, and well, HT makes those processes MUCH faster(perhaps 10-20 minutes difference) as for gaming, if your CPU si fast enough and new enough to even have HT as an option, i highly doubt it is to any advantage (69fps is no better then 60fps) to disable it for those few extra clock cycles
|
|
|
Mar 25, 2004, 11:09 PM
|
#17
|
|
Banned
Join Date: Jul 2002
Location: Gurnee Illinois
Posts: 4,677
|
Quote:
Originally posted by Judas
saposedly... game preformance will increase....
|
Almost no games benefit from it. HT only allows most programs to just use 50% of CPU power. But some things, such as F@H benefit from HT. A dedicated folding rig increases output by roughly 20% with HT enabled.
Quote:
Originally posted by Necrosis
As far as I understand same games do use HT but for the life of me I'm not sure which ones. Besides what is the point of leaving it off? You can overclock a bit faster but lose some of the dual sex appeal. HT is one of the reasons I even went the route I did.
1. FSB
2. HT
Well 3..Video Encoding.
|
NO, doesn't help on video encoding. Load up divx and encode a video, only allow 50% of CPU usage.
|
|
|
Mar 25, 2004, 11:17 PM
|
#18
|
|
confutatis maledictis
Join Date: May 2002
Location: somewhere dark
Posts: 5,952
|
Quote:
Originally posted by DriveEuro
NO, doesn't help on video encoding. Load up divx and encode a video, only allow 50% of CPU usage.
|
he wasn't saying HT helped in that . . . he was listing reasons he bought a P4
|
|
|
Mar 25, 2004, 11:24 PM
|
#19
|
|
Banned
Join Date: Jul 2002
Location: Gurnee Illinois
Posts: 4,677
|
Quote:
Originally posted by Vampyromaniac
he wasn't saying HT helped in that . . . he was listing reasons he bought a P4
|
Um, well HT doesn't help in video encoding... Unless some encoders have recently changed their software....
|
|
|
Mar 25, 2004, 11:29 PM
|
#20
|
|
confutatis maledictis
Join Date: May 2002
Location: somewhere dark
Posts: 5,952
|
Quote:
Originally posted by DriveEuro
Um, well HT doesn't help in video encoding...
|
Again . . . nobody said that it did.
|
|
|
Mar 25, 2004, 11:38 PM
|
#21
|
|
Styleless Wonder
Join Date: Jun 2002
Location: Ottawa, Ontario
Posts: 6,049
|
HT on if I had it
Be sure to turn on HT options in games which allow it. A lot of the time it's in the config file. I know Q3 based games have that option. For sure RTCW does.
|
|
|
|
|
|