After reviewing the docs on i7 and looking at the tech information at FSINSIDER I have come to the conclusion that there MAY BE some value in running HT with i7 and FSX if the affinity mask edit is used


I personally do not have time to drop what I am doing and go into this and test it correctly .. therefore I am going to post this for people using i7 to try


NOTE: IF you are running 4GHz and enabling HT may push the temp of the i7 processor over the 80c limit you should reduce the clock on the processor to ensure safe operation. You may wish to stress test with HT enabled and establish your clock first before running this test.


Enter the BIOS and ENABLE this setting if it is not already enabled:

Intel HT Technology


In the FSX.cfg add the following if it is not already there

[JOBSCHEDULER]
AffinityMask=N


use these settings for N as you test and you must EXIT the sim and relaunch between changes

255 - all 8 in use

254 - 7 in use with the first remaining free

252 - 6 in use with the first entire CORE remaining free

127 - 7 in use with the last remaining free

63 - 6 in use with the last entire CORE remaining free


Verify the use and the perf then report back please.


There are some unanswered questions about the JOBSCHEDULER and what I posted in affinity mask settings above with i7 would answer a lot of them.

Test Results.

JOBSCHEDULER test in FSX

1. System:       i7 920 OC to 3.8 Gh
     3Gb Mushkin tri-channel DDR3 7-8-7-20@1603
               BCLCK 200
               Multi 19
     Vcore 1.34v
               QPI/DRAM 1.34v
     DRAM 1.64v
               TWO cores
               HT on
               AffinityMask=255      
               All else AUTO
               All other BIOS settings as suggested by Nick
     MB ASUS P6T Delux version 2
     BIOS: not sure but when the she boots I now  see 'NVRAM   Check'. Boot seems faster.
               EVGA 260 OC 15%
           

2. Add-ons: GTXn, UTX USA, REX, TrafficX (not used in test), FSGensis, various airports, aircraft

3. FSX settings:         1920x1200@32 on a 24" Gateway
                     Frames: locked@28
                              All other settings per Nick's advice except Autogen @ 75%. traffic: airline=30, GA=25 (I like this one and ran the test with my perferred settings)

4. Nhancer: as suggested BUT anti-alasing @ 8xSQ

5. Cfg file mods: TEXTURE_BANDWIDTH_MULT=80
               UPPER_FRAMERATE_LIMIT=28
               WideViewAspect=True
               RUNWAY_LIGHTS_SURFACE_SCALAR=.1
               RUNWAY_LIGHTS_VASI_SCALAR=0.2
               RUNWAY_LIGHTS_APPROACH_SCALAR=0.2
                RUNWAY_LIGHTS_STROBE_SCALAR=0.2
                [Main]
                DisablePreload=1
                [Bufferpool]
                PoolSize=70000000
                [JOBSCHEDULER]
                AffinityMask=255

6. Test aircraft:      Baron, F18,C172, RealAir SF260

7. Test conditions:      Clear, Building Storms, Fogged in, CBs
   Test area:      Around Richmond, VA and near Whitbey NAS (lots of trees)
                   New York@ night

8. Results: All flights smooth with no tree shimmer(due to the 8xSQ), smooth turns ( was not possible with 8xSQ before)      

9. Temps: Using Real Temp 2.90 max during flight: 63C
     OCCT ver 3.0.0.b16 passed with lower temps than when in FSX

10. Conclusion: On my test system, this was a surprise. I will continue to use TWO cores for a while.
FYI, I could not get 4.0Gh stable and am perfectly satisfied with 3.8 considering what I see on screen
               
If this fails, I will post immediately!

Jack S
               

Sorry for the delay, I was still testing.

Advanced
Intel(R) HT Technology
Active Processor Cores[All]  - default

I have the choice of 1,2,4,all.

Jack

OK, thanks. I was slightly mistaken; the P6T manual: "Allows you to choose the number of CPU cores to activate in each processor package.
Configuration options:[All],[1],[2]"

Doesn't explain exactly what this means so I elected to gice it a try.

Jack

All,

Based on what Nick pointed out, my test is flawed! (I cut out 1/2 of my processing power, DUH). I went back to where I was supposed to be (All) and reran the test.

My experience I believe is smoother with Affinity mask=255, any other setting on my system resulted in frames jumping all around when locked at 28. The 255 setting was definitely the best for what I have.

Live and learn I guess.   Thanks Nick for the correction.

Sooner or later, I will get this fun stuff.

My temps maxed at 72C while in FSX with HT on so I was getting close even at 3.8Gh.

Jack

Yes and No. Smoother flight with HT but I failed to try maxed with HT off but believe I remember that I had a micro-stutter flight with auto maxed and HT off.

Also found that the aircraft does, as you have said, make a difference.

When I changed from the 260 to the 172 over NYC at night, the ride was more fluid.

I'm hopeful someone else will come aboard with test results so you have vaild data points, especially with a 940 or 965.

Jack

PS why would ASUS offer a choice of cores?

When I get my mobo back I will be onboard this test and let you know. I'm running a 240 and will be running 3.5-4 ghz

Thanks Jack!

Absolutely it helps.. !

Nick, I tried the test and when I did I goot my OOM problem back. I even lowered the usera down as low as you mentioned and still no luck. Then I removed my OC just to see if it would make things better and still no luck. SO for us 32 bit suckers Hyperthreading does not seem to be a option if you have a GPU with a ton of ram on it. Put all my settings back to as they were and all was good again. Will try the Mask now with no HT.

Nick,

Update on AM=15, HT off.

I don't know how you do this  , but here are my results:

Autogen = MAX (finally!!)
Scenery = MAX
Locked frames@30
AntiAlaising = 8xSQ!

Beaver: t/o from New York Skyports, day, fair weather: SMOOTH, frames dropped 2 during t/o run, otherwise 30.

C172: building storms at RIC: SMOOTH

SF260: aerobatics over mucho trees N of Whitbey NAS: SMOOTH

Baron: building storms, landing at KFCI: SMOOTH

I am not changing anything and will keep all settings after reinstall due to my moving to XP64 as soon as the disk arrives.

I still can't do unlimited without spikes but the flight was much smoother.

THANK YOU!!!!

Jack

jwenham wrote on Jan 28^th, 2009 at 10:13pm:



It locks FSX to the 4 physical cores of the proc and my theory right now is that it also resolves thread collision problems the i7 presents since FSX is not a hyperthread application and the changes to i7 design may be making the app work a bit confused as to what is a real core and what is a virtual core.

i7 design change makes the virtual core appear like a real core. Just because HT is disabled does not mean the other part of the core is not seen by the app.. .and that may be causing FSX to spawn threads for cores that do not really exist. By the time the app figures out that the threads being spawned are not being addressed it spawns another thread to a real core and a collision occurs and a stutter may ensue.

By locking FSX to the 4 physical cores the collisions may be eliminated

ehhhh.. you cant really use that readout to make a solid estimate of what is happening.. but I get the idea

wsciutti wrote on Jan 29^th, 2009 at 2:16pm:




LOL!

gets'em every time

HT disabled for FSX

affinity mask is problematical and somthin you need to work out personally for the system


HT is only useful with applications like A/V editing and engineering programs designed to make use of the virtural threads and should be enabled if you use such software

macca22au wrote on Feb 15^th, 2009 at 1:17am:



All those tweaks are "Try and see" there is no such thing as "one size fits all" in the FSX.cfg, although TBM 70 I have always found to be a constant with Disable Preload,... but as I said, your mileage may vary