AMD Processors
Decrease font size
Increase font size
Topic Title: In game play & 1090T x6 Hexa core. Boost & 6 core usage tests
Topic Summary:
Created On: 05/04/2010 12:49 AM
Status: Read Only
Linear : Threading : Single : Branch
<< 1 2 3 Previous Next Last unread
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 05/04/2010 02:32 AM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

ok baseline posted.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)
 05/04/2010 02:45 AM
User is offline View Users Profile Print this message

Author Icon
QB the Slayer
Case Modder

Posts: 1304
Joined: 01/23/2010

Excellent... now can you see why I say you really should only be looking at averages since there is no time line... even with nothing running except the OS and background tasks/resources (I know it is lean but they are still there) you have CPU use from 1.7-6.4% in this one single test (true tests would be done in at least 3's but that is uber time consuming)

The only way you can claim a game uses a core is if the average is above this range... I would even go so far as to say a 15% OS threshold. A game uses a core if it is above 15%.

Now this is all relative as well, since you may run a game that barely uses any core, and those numbers will have to be looked at with a different light... and maybe even thrown out since with the limited amount of info we may not be able to determine the amount of core use in such a situation.

QB

-------------------------

The MONSTER HTPC

CPU: AMD Phenom II X4 965 C2 (140W).||.Cooler: Corsair H80i
MB: Gigabyte 990FXA-UD7.||.RAM: 8 GB Mushkin Blackline DDR3 2000MHz (7-10-8-27-1T)
Case: CoolerMaster HAF 932.||.PSU: Corsair HX750
GPU:HIS IceQ 5 Radeon HD 5770 Turbo 1GB.||.Audio: Creative X-Fi Titanium Fatal1ty Pro w/ Logitech Z-5300e (5.1, 280W-RMS)
Drive: 2xKingston SSD 40GB RAID0.||.Storage: 6TB (4x500GB Caviar Black RAID0, 2TB Hitachi & 2TB Caviar Green)
 05/04/2010 02:49 AM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

I am going to say that if a core has an average of 15% or higher then it is used to some degree in game.

looking at base line and max and in game max needs to be taken into account.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)
 05/04/2010 03:02 AM
User is offline View Users Profile Print this message

Author Icon
slaveondope
Case Modder

Posts: 1592
Joined: 01/25/2010

How much would you say your avg. fps gains are in bc2?
Thats the game Ive been wondering about mostly.

-------------------------
More Advanced Micro Devices than Intel
 05/04/2010 03:05 AM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

gross 12 FPS

Average 9 FPS

smoothness eliminated all stutter or drop in FPS below 60 FPS.

Smoothness is all that matters to me, I was already over 100 FPS to 130 FPS but some intense parts could drop into 35 FPS before and now they are all over 60 FPS.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)
 05/04/2010 03:07 AM
User is offline View Users Profile Print this message

Author Icon
slaveondope
Case Modder

Posts: 1592
Joined: 01/25/2010

ok so like when you have those big explosion seens with alot of particles/physics it holds out alot better then

-------------------------
More Advanced Micro Devices than Intel
 05/04/2010 03:08 AM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

Originally posted by: slaveondope

ok so like when you have those big explosion seens with alot of particles/physics it holds out alot better then


Correct

Better way to look at it is x6 = more even constant FPS game play in BFBC2.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)

Edited: 05/04/2010 at 03:19 AM by PC-GURU -
 05/04/2010 08:58 AM
User is offline View Users Profile Print this message

Author Icon
QB the Slayer
Case Modder

Posts: 1304
Joined: 01/23/2010

Originally posted by: PC-GURU -

I am going to say that if a core has an average of 15% or higher then it is used to some degree in game...


Now we are talking. I am very interested at seeing the final results.

So far we have BFBC2 as doing the best job using all 6 cores. And this is also evidenced by the average CPU clock.

Dirt 2 & Crysis Warhead seem to be trying to use all 6, but not doing a great job at it (maybe the other cores are not as needed as they are in BFBC2) and this is illustrated in Crysis with core 6 usage far higher than the rest. And also a slightly higher CPU clock.

QB

-------------------------

The MONSTER HTPC

CPU: AMD Phenom II X4 965 C2 (140W).||.Cooler: Corsair H80i
MB: Gigabyte 990FXA-UD7.||.RAM: 8 GB Mushkin Blackline DDR3 2000MHz (7-10-8-27-1T)
Case: CoolerMaster HAF 932.||.PSU: Corsair HX750
GPU:HIS IceQ 5 Radeon HD 5770 Turbo 1GB.||.Audio: Creative X-Fi Titanium Fatal1ty Pro w/ Logitech Z-5300e (5.1, 280W-RMS)
Drive: 2xKingston SSD 40GB RAID0.||.Storage: 6TB (4x500GB Caviar Black RAID0, 2TB Hitachi & 2TB Caviar Green)
 05/04/2010 09:10 AM
User is offline View Users Profile Print this message

Author Icon
pcnazz
Sub Zero

Posts: 2246
Joined: 05/17/2008

AMD's new Thuban 6 core is a nice future proof cpu for the ones who do not have to buy every new AMD cpu that comes out .

-------------------------
    Room Temp H20 | 2 Stage Cascade | PH II X6 1090T BE | PH II X2 555 / Asus CH4 v905 | i7 970 / Asus R2E v1914 | Corsair HX1000 1210W | Asus 5970 QF | Corsair Dominator GT 2000 | WD VR 3+0 Raid ( 450gb )
 05/04/2010 12:18 PM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

Metro 2033 posted with rather interesting results.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)
 05/04/2010 02:21 PM
User is offline View Users Profile Print this message

Author Icon
slaveondope
Case Modder

Posts: 1592
Joined: 01/25/2010

Originally posted by: PC-GURU -

Metro 2033 posted with rather interesting results.


Why? What? That its really not the hard on the cpu?
Looks to use 2 cores for the physics and what not then the other core feed the gpu's which is the hardest part of that game.

Next up Spider Solitare!!!

-------------------------
More Advanced Micro Devices than Intel
 05/04/2010 02:22 PM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

lol I was talking about boost usage in game buddy.

Clearly used all 6 cores .

as for your request, your funny lol.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)
 05/04/2010 10:01 PM
User is offline View Users Profile Print this message

Author Icon
QB the Slayer
Case Modder

Posts: 1304
Joined: 01/23/2010

Actually I took Metro to be poor at multi threading... Mainly used just 2 cores and because of this it used boost a whole lot more than any game so far. There is a inverse correlation to # of cores used to amount of boost used. More even core use=less boost used... That is how the boost works. So if you see really high boost values (avg) then there is at least 3 cores not in use very much. That is why BFBC2 had such a low boost avg, it spread the work out very evenly

QB

-------------------------

The MONSTER HTPC

CPU: AMD Phenom II X4 965 C2 (140W).||.Cooler: Corsair H80i
MB: Gigabyte 990FXA-UD7.||.RAM: 8 GB Mushkin Blackline DDR3 2000MHz (7-10-8-27-1T)
Case: CoolerMaster HAF 932.||.PSU: Corsair HX750
GPU:HIS IceQ 5 Radeon HD 5770 Turbo 1GB.||.Audio: Creative X-Fi Titanium Fatal1ty Pro w/ Logitech Z-5300e (5.1, 280W-RMS)
Drive: 2xKingston SSD 40GB RAID0.||.Storage: 6TB (4x500GB Caviar Black RAID0, 2TB Hitachi & 2TB Caviar Green)
 05/04/2010 10:45 PM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

True, but still used every last core just the same.

I think there is one thing being missed here. I seen a very real bottleneck with my 5870,s and my 965 BE.

This bottelneck was eliminated with the 1090T.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)

Edited: 05/04/2010 at 11:10 PM by PC-GURU -
 05/04/2010 11:11 PM
User is offline View Users Profile Print this message

Author Icon
slaveondope
Case Modder

Posts: 1592
Joined: 01/25/2010

I dont understand how or why the OS reports stuff like superpi a single threaded app. loads across all cores.

Wouldnt that just mean that all this "it's only is coded for 2 threads stuff" goes out the window?

Or is it that a 2 thread app. will only still use half of a 4 core combined output across all cores?

Obviously superpi would be done quicker if it used all cores to the max but it doesnt because its single threaded.

I guess I understand the limitations of single threaded apps. by why on earth does windows report usage across all cores? Shared HT?

Mime please chime in

-----------------------------------------------------

It will be nice to see BullDozer cause to my understanding its suppose to take a single thread and actually load up all cores to the max some how and then spit out the results.

This is what goes on in my head no place for those seeking facts.

-------------------------
More Advanced Micro Devices than Intel
 05/04/2010 11:13 PM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

Originally posted by: slaveondope

I dont understand how or why the OS reports stuff like superpi a single threaded app. loads across all cores.



Wouldnt that just mean that all this "it's only is coded for 2 threads stuff" goes out the window?



Or is it that a 2 thread app. will only still use half of a 4 core combined output across all cores?



Obviously superpi would be done quicker if it used all cores to the max but it doesnt because its single threaded.



I guess I understand the limitations of single threaded apps. by why on earth does windows report usage across all cores? Shared HT?



Mime please chime in [IMG][/IMG]



-----------------------------------------------------



It will be nice to see BullDozer cause to my understanding its suppose to take a single thread and actually load up all cores to the max some how and then spit out the results.



This is what goes on in my head no place for those seeking facts.


super PI only used 1 to 2 cores bud, already tested that. Hyper PI is multi threaded.

I posted screen shots of super PI and the x6 in the 1090T thread only using 1 to 2 cores tops.

But your question has Merritt and I can answer this.

Windows 7 has threading optimization built in where it will force a single thread to be spread of multi cores.

2 cores can and will be spread over all available cores . It is called load balancing.

It is why 2 to 3 threaded games still show all 6 being used.

It is also why a 6 core allows more GPU,s to be more evenly powered .

windows vista SP2 supposedly added this as well.

Boost will sniff this out, If it see's a app trying to use 1 2 or 3 cores more then all of them then boost kicks in to allow a higher performance out of the used cores by that app.

Funny part is boost will use all 6 cores, just not at the same time.

Watch it in AMD OD and you can see it will shift cores with boost.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)

Edited: 05/04/2010 at 11:22 PM by PC-GURU -
 05/04/2010 11:19 PM
User is offline View Users Profile Print this message

Author Icon
slaveondope
Case Modder

Posts: 1592
Joined: 01/25/2010

Nope sorry Guru both mine show load across all cores on Pi

EDIT:
Ok I thought windows was doing something like that.

-------------------------
More Advanced Micro Devices than Intel
 05/04/2010 11:23 PM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

Originally posted by: slaveondope

Nope sorry Guru both mine show load across all cores on Pi



EDIT:

Ok I thought windows was doing something like that.


ya but if you watch 2 cores get used more then the others. Windows is using it,s load balancing to try and force it to spread out the load more to improve performance.

Just look at some of these games, Like our friend said. It is clear what games are programed to use 1 to 3 cores or threads.

But my point is 6 cores do get used, not fully or well optimized. but thanks to win 7 and 6 cores, powering multi GPU just became better then before.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)
 05/05/2010 01:46 AM
User is offline View Users Profile Print this message

Author Icon
Overmind
Assimilator

Posts: 8052
Joined: 01/22/2004

Please TEst STALKER CoP, FEAR, HL2 and Elite Force 2.

-------------------------

World's best Red Alert 2: Yuri's Revenge mod and Star Trek: Starfleet Command 3 mod: Overmind.ro
 05/05/2010 11:02 AM
User is offline View Users Profile Print this message

Author Icon
PC-GURU -
Lurker

Posts: 8018
Joined: 11/11/2008

Originally posted by: Overmind

Please TEst STALKER CoP, FEAR, HL2 and Elite Force 2.


ok will do bud.

Running stalker cop now.

seeing a 20 FPS boost in all out DX 11 in this game with the 1090T. places with high demanding spots are quite a bit smoother and higher FPS now lol.

Was a time all maxed out this game would dip to around 40 to 45 FPS in some spots. Now shes over 60 FPS in these spots.

Really opened up the GPU,s for sure.

-------------------------
AMD phenom II 1090T @ 4.23 Ghz custom water cooled
Asrock 890 FX deluxe 3 mother board
4 GB DDR 3 1600 @ 1688 Mhz 7 7 7 24 T1 2688 Mhz NB
2x 1 TB WD Sata III 64 mb cache HDD,s raid 0
2x 5870 HD @ 1015/1300 Asus reference
Dual OS win 7 x64 ultimate & win 7 x32 Ultimate (bench os)

Edited: 05/05/2010 at 11:24 AM by PC-GURU -
Statistics
112018 users are registered to the AMD Processors forum.
There are currently 0 users logged in.

FuseTalk Hosting Executive Plan v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.



Contact AMD Terms and Conditions ©2007 Advanced Micro Devices, Inc. Privacy Trademark information