[tugboat_2]

I really appreciate the response.

I didn not know the amount of motherboard reserved memory and will keep that tidbit in mind since our boards are very near the same.

Machine Specs:
Case..................CM ATCS 840
CPU...................Intel i7 920 (DO) 2.66 GHz (@4.0GHz)
CPU HS.............Prolimatech Megahalems w/2 120X25MM PWM 96CFM fans
MoBo.................Asus P6T7WS SuperComputer
Mem..................OCZ 6gig Reaper 1800 (3x2) TriChan (@1804)
Thermal Paste....IC7 Diamond
VidCard.............XFire Sapphire HD4890 1Gig DDR5 (OC to 930/1025)
HDD..................Seagate Barracuda 7200.12 1 TB
Opti Drv............(x2) Samsung SH223 sata
PSU..................CORSAIR CMPSU-850TX 850W PFC
UPS..................CyberPower CP1500AVRLCD 1500watt
OS....................M$ win 7 RC 64 bit

By the way, I am far from expert, but isn't it best to run either memory in 3 gigabit increments with socket 1366 boards and CPU's? I noticed you are running 4 vice 3 or 6. Also I had no thought that you were knocking the i7 processors. I was only trying to understand how that CPU was a bottleneck at that speed. I truly don't know and think it pertainate to my on endeavors.

[daveyd]

My bad it wasn't the cpu as i suspected, I'll figure it out shortly. I enjoy posting analytical results to clarify some things that I discover while in the process of system configuration, hopefully it helps others too.


You have a nice setup. That motherboard is awesome! Put that beast to good use. Also, the CPU that you have will be capable of running an additional GPU. No worries there, especially on that motherboard.

In reference to your question about my ram being dual channel and not triple, I have 2 gig denomination memory, and i'm using windows 32 bit, so there's a memory limitation in 32 bit architecture. It's a personal choice that may be subject to debate. lol However, 32 bit Widows 7 does take advantage of all cores that the i7 cpu has. In a nutshell, Windows7 64bit takes advantage of all the cores plus all the memory you can sling at it.

Dual channel or triple channel memory, it's really up to you. More memory helps on certain things that you do on a regular basis. The best way to find out which is optimal, is to test them and see, as not all motherboards, memory, gpu, and cpu combos are created equal. Some can clock higher than others, one manufacturer can come out with a new revision, so there will always be a slight variance in tolerance coming from the factory, despite both items being the same make and model.

A good example of this can be seen by looking at the GPU exhaust outlets on back of my gaming rig below:



All three are the same make and model. Did you spot the difference? As subtle as these differences may be, it can affect the overall performance, or thermal characteristics of the third GPU. Which is why I put the least vented one at the bottom of the range, so that it can pull in a larger, cooler, volume of air.

With memory modules, and triple channel motherboards, the differences may or may not be so subtle, and sometimes cannot be spotted with the human eye until you use an oscilloscope and give it a thorough examination.

Happy benchmarking and overclocking!

[tugboat_2]

roger that.

As to the memory I was just under the impression that, in the case of a 32 bit Windows OS, you took a bigger performance hit by trying to run more than 3 gigs of memory than by putting a single stick of one gig mem in ea pair of slots. As in, the CPU memory controller is optimised for triple channel operation. I know that Windows 32 bit can actually use a little over 3 gigs.

And yeah the mother board is great. In strictly practical terms over kill in most ways for me. But I was doing a first build and did not have to deal "legacy gear" this time around. So I wanted max number of expansion slots, maximum number of sata, and usb slots and a board that I could both play the hell out of and tweak and experiment with for at least a couple of CPU generations. I've learned more in the last two or 3 months than I have in the last 6 years using cookie cutters.

[daveyd]

AMEN to that tugboat. Rock the house with ASUS for taking the initiative to move ahead with innovation.
I agree with your concept on buying the best motherboard money can buy, and having addressed it's compatibility towards future generations of cpu's whenever possible. I too did away without the old IDE, and floppy drive interface, and don't miss them one bit. Have you tried the onboard SAS yet, and do you take advantage of ASUS ExpressGate when it boots up?

As to the performance hit, I'll look into it and try testing to see the difference. I believe you're correct.

Ten-four ROGER, over and out.

[daveyd]

[daveyd]

CRYSIS Benchmarks DX10 - Four sets of three runs at Ultra High Resolution - 2560 X 2048 with 8X Anti-Aliasing applied on all tests- GLOBAL QUALITY- Set to 'VERY HIGH' ON Test 1; Test 2 Global Quality- HIGH; Test 3- Global Quality-MEDIUM; Test 4 Global Quality- Set to LOW. No CPU over clock, everything set to 'factory default'. (Tested with Windows7 32 bit)

Results -
11/5/2009 5:25:59 AM - Vista
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX10 2560x2048, AA=8x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: VeryHigh
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 54.12s, Average FPS: 36.96
Min FPS: 27.77 at frame 138, Max FPS: 60.21 at frame 984
Average Tri/Sec: -13675064, Tri/Frame: -370029
Recorded/Played Tris ratio: -2.48
!TimeDemo Run 1 Finished.

Play Time: 42.95s, Average FPS: 46.57
Min FPS: 27.77 at frame 138, Max FPS: 60.21 at frame 984
Average Tri/Sec: -16709807, Tri/Frame: -358813
Recorded/Played Tris ratio: -2.55
!TimeDemo Run 2 Finished.

Play Time: 42.78s, Average FPS: 46.75
Min FPS: 27.77 at frame 138, Max FPS: 62.96 at frame 995
Average Tri/Sec: -16721587, Tri/Frame: -357696
Recorded/Played Tris ratio: -2.56
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Beginning Run #2 on Map-island, Demo-benchmark_gpu
DX10 2560x2048, AA=8x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 43.07s, Average FPS: 46.44
Min FPS: 36.90 at frame 1948, Max FPS: 77.80 at frame 1002
Average Tri/Sec: -31996060, Tri/Frame: -689009
Recorded/Played Tris ratio: -1.33
!TimeDemo Run 1 Finished.

Play Time: 33.97s, Average FPS: 58.87
Min FPS: 36.90 at frame 1948, Max FPS: 83.94 at frame 75
Average Tri/Sec: -40539648, Tri/Frame: -688605
Recorded/Played Tris ratio: -1.33
!TimeDemo Run 2 Finished.

Play Time: 34.09s, Average FPS: 58.66
Min FPS: 36.64 at frame 1959, Max FPS: 87.72 at frame 90
Average Tri/Sec: -40274812, Tri/Frame: -686538
Recorded/Played Tris ratio: -1.34
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Beginning Run #3 on Map-island, Demo-benchmark_gpu
DX10 2560x2048, AA=8x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: Medium
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 32.84s, Average FPS: 60.90
Min FPS: 49.90 at frame 1990, Max FPS: 96.26 at frame 988
Average Tri/Sec: 55267448, Tri/Frame: 907505
Recorded/Played Tris ratio: 1.01
!TimeDemo Run 1 Finished.

Play Time: 26.64s, Average FPS: 75.07
Min FPS: 47.67 at frame 1957, Max FPS: 96.35 at frame 88
Average Tri/Sec: 68665624, Tri/Frame: 914742
Recorded/Played Tris ratio: 1.00
!TimeDemo Run 2 Finished.

Play Time: 26.65s, Average FPS: 75.06
Min FPS: 47.67 at frame 1957, Max FPS: 100.18 at frame 127
Average Tri/Sec: 68731048, Tri/Frame: 915699
Recorded/Played Tris ratio: 1.00
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Beginning Run #4 on Map-island, Demo-benchmark_gpu
DX10 2560x2048, AA=8x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: Low
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 23.75s, Average FPS: 84.20
Min FPS: 71.97 at frame 74, Max FPS: 147.73 at frame 1025
Average Tri/Sec: 43268856, Tri/Frame: 513888
Recorded/Played Tris ratio: 1.78
!TimeDemo Run 1 Finished.

Play Time: 17.80s, Average FPS: 112.39
Min FPS: 71.97 at frame 74, Max FPS: 147.73 at frame 1025
Average Tri/Sec: 58291044, Tri/Frame: 518659
Recorded/Played Tris ratio: 1.77
!TimeDemo Run 2 Finished.

Play Time: 17.72s, Average FPS: 112.88
Min FPS: 71.97 at frame 74, Max FPS: 152.47 at frame 141
Average Tri/Sec: 58451432, Tri/Frame: 517799
Recorded/Played Tris ratio: 1.77
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

11/5/2009 5:25:59 AM - Vista

Run #1- DX10 2560x2048 AA=8x, 32 bit test, Quality: VeryHigh ~~ Overall Average FPS: 46.66
Run #2- DX10 2560x2048 AA=8x, 32 bit test, Quality: High ~~ Overall Average FPS: 58.765
Run #3- DX10 2560x2048 AA=8x, 32 bit test, Quality: Medium ~~ Overall Average FPS: 75.065
Run #4- DX10 2560x2048 AA=8x, 32 bit test, Quality: Low ~~ Overall Average FPS: 112.635

[tugboat_2]

Was that last set with or without dedicated Physix?

[daveyd]

It was with physX disabled. I can do a run and post it for you with it on. Jst let me know.

[tugboat_2]

That's ok. I was only interested to see how big a deal a dedicated Physix card is. Just simple curiosity since I don't have nVidia equip.

[daveyd]

Exact same test sequence and settings as the previous post- This time with dedicated PhysX enabled in the nVidia control panel, and a quick reboot. Since Crysis doesn't support PhysX, the difference between the two test results should be minuscule at best. (Windows7 32 bit)

11/5/2009 7:23:33 AM - Vista
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX10 2560x2048, AA=8x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: VeryHigh
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 54.60s, Average FPS: 36.63
Min FPS: 26.00 at frame 138, Max FPS: 59.96 at frame 985
Average Tri/Sec: -13578478, Tri/Frame: -370718
Recorded/Played Tris ratio: -2.47

!TimeDemo Run 1 Finished.
Play Time: 43.14s, Average FPS: 46.36
Min FPS: 26.00 at frame 138, Max FPS: 62.16 at frame 1006
Average Tri/Sec: -16609462, Tri/Frame: -358247
Recorded/Played Tris ratio: -2.56
!TimeDemo Run 2 Finished.

Play Time: 42.76s, Average FPS: 46.77
Min FPS: 26.00 at frame 138, Max FPS: 62.19 at frame 981
Average Tri/Sec: -16760350, Tri/Frame: -358352
Recorded/Played Tris ratio: -2.56
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Beginning Run #2 on Map-island, Demo-benchmark_gpu
DX10 2560x2048, AA=8x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 42.67s, Average FPS: 46.87
Min FPS: 37.22 at frame 1967, Max FPS: 76.87 at frame 1013
Average Tri/Sec: -32298600, Tri/Frame: -689152
Recorded/Played Tris ratio: -1.33
!TimeDemo Run 1 Finished.

Play Time: 33.97s, Average FPS: 58.87
Min FPS: 36.52 at frame 1973, Max FPS: 91.85 at frame 118
Average Tri/Sec: -40539608, Tri/Frame: -688584
Recorded/Played Tris ratio: -1.33
!TimeDemo Run 2 Finished.

Play Time: 33.67s, Average FPS: 59.39
Min FPS: 36.52 at frame 1973, Max FPS: 91.85 at frame 118
Average Tri/Sec: -40780892, Tri/Frame: -686629
Recorded/Played Tris ratio: -1.34
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Beginning Run #3 on Map-island, Demo-benchmark_gpu
DX10 2560x2048, AA=8x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: Medium
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 31.93s, Average FPS: 62.64
Min FPS: 50.93 at frame 1978, Max FPS: 99.64 at frame 921
Average Tri/Sec: 56851052, Tri/Frame: 907563
Recorded/Played Tris ratio: 1.01

!TimeDemo Run 1 Finished.
Play Time: 26.13s, Average FPS: 76.55
Min FPS: 48.07 at frame 1974, Max FPS: 99.64 at frame 921
Average Tri/Sec: 70047808, Tri/Frame: 915027
Recorded/Played Tris ratio: 1.00
!TimeDemo Run 2 Finished.

Play Time: 26.12s, Average FPS: 76.57
Min FPS: 48.07 at frame 1974, Max FPS: 99.64 at frame 921
Average Tri/Sec: 70119976, Tri/Frame: 915807
Recorded/Played Tris ratio: 1.00
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Beginning Run #4 on Map-island, Demo-benchmark_gpu
DX10 2560x2048, AA=8x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: Low
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 23.68s, Average FPS: 84.46
Min FPS: 72.90 at frame 75, Max FPS: 148.90 at frame 891
Average Tri/Sec: 43409244, Tri/Frame: 513948
Recorded/Played Tris ratio: 1.78
!TimeDemo Run 1 Finished.

Play Time: 17.74s, Average FPS: 112.75
Min FPS: 72.90 at frame 75, Max FPS: 153.78 at frame 148
Average Tri/Sec: 58475516, Tri/Frame: 518642
Recorded/Played Tris ratio: 1.77

!TimeDemo Run 2 Finished.
Play Time: 17.59s, Average FPS: 113.69
Min FPS: 72.90 at frame 75, Max FPS: 159.03 at frame 172
Average Tri/Sec: 58848304, Tri/Frame: 517642
Recorded/Played Tris ratio: 1.77
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

11/5/2009 7:23:33 AM - Vista

Run #1- DX10 2560x2048 AA=8x, 32 bit test, Quality: VeryHigh ~~ Overall Average FPS: 46.565
Run #2- DX10 2560x2048 AA=8x, 32 bit test, Quality: High ~~ Overall Average FPS: 59.13
Run #3- DX10 2560x2048 AA=8x, 32 bit test, Quality: Medium ~~ Overall Average FPS: 76.56
Run #4- DX10 2560x2048 AA=8x, 32 bit test, Quality: Low ~~ Overall Average FPS: 113.22

[tugboat_2]

I see what you mean. In fact since it's not coded for physiX it actually hurt you by dropping back to only 2 cards in sli. Which makes me think, that until "lots" of games are so coded, this proprietary ability is not worth what nVidia wants for it. Both in dollars and in performance in a 3 card set up.

[daveyd]

[daveyd]

BFG -GTX260 Core 216 896MB 3-Way Sli-


Crysis Warhead- Maximum Performance Benchmarks at 1680 X 1050 with Anti-Aliasing set to=16xQ Test 1: DX9; Test 2: DX10



Processor: Intel(R) Core(TM) i7 CPU 965 @ 3.20GHz @ 3207 Mhz
CPU ID: x86 Family 6 Model 26 Stepping 4
Operating System: Microsoft Windows 7 Home Premium
Physical memory: 2.49 GB
Display adapter: NVIDIA GeForce GTX 260 896 MB
Driver version: 8.16.11.9107 (20090927000000.000000-000)

==============================================================
TimeDemo Play Started , (Total Frames: 3502, Recorded Time: 92.69s)
!TimeDemo Run 0 Finished.
Play Time: 19.14s, Average FPS: 182.95
Min FPS: 81.65 at frame 92, Max FPS: 236.47 at frame 965
Average Tri/Sec: 60642912, Tri/Frame: 331469
Recorded/Played Tris ratio: 1.58
!TimeDemo Run 1 Finished.
Play Time: 18.61s, Average FPS: 188.21
Min FPS: 81.65 at frame 92, Max FPS: 236.47 at frame 965
Average Tri/Sec: 63527156, Tri/Frame: 337537
Recorded/Played Tris ratio: 1.55
!TimeDemo Run 2 Finished.
Play Time: 18.63s, Average FPS: 188.01
Min FPS: 81.65 at frame 92, Max FPS: 236.47 at frame 965
Average Tri/Sec: 63441380, Tri/Frame: 337440
Recorded/Played Tris ratio: 1.55
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

==============================================================
TimeDemo Play Started , (Total Frames: 3502, Recorded Time: 92.69s)
!TimeDemo Run 0 Finished.
Play Time: 22.03s, Average FPS: 158.97
Min FPS: 97.53 at frame 139, Max FPS: 217.57 at frame 1134
Average Tri/Sec: 52622724, Tri/Frame: 331022
Recorded/Played Tris ratio: 1.58
!TimeDemo Run 1 Finished.
Play Time: 21.92s, Average FPS: 159.77
Min FPS: 97.53 at frame 139, Max FPS: 217.57 at frame 1134
Average Tri/Sec: 53809840, Tri/Frame: 336790
Recorded/Played Tris ratio: 1.55
!TimeDemo Run 2 Finished.
Play Time: 21.85s, Average FPS: 160.30
Min FPS: 88.49 at frame 90, Max FPS: 217.57 at frame 1134
Average Tri/Sec: 54038396, Tri/Frame: 337104
Recorded/Played Tris ratio: 1.55
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

//////////// Summary \\\\\\\\\\\\\

11/06/2009 00:22:38 - Microsoft Windows 7 Home Premium

DirectX 9.0c PERFORMANCE 3X @ Map: ambush @ 0 1680 x 1050 AA 16xQx
==> Framerate [ Min: 81.65 Max: 236.47 Avg: 188.11 FPS]

DirectX 10 PERFORMANCE 3X @ Map: ambush @ 0 1680 x 1050 AA 16xQx
==> Framerate [ Min: 93.01 Max: 217.57 Avg: 160.04 FPS]

MAXIMUM GAME

[daveyd]

I'll have to redo these benchmarks to reflect triple channel memory, as my motherboard was acting strange for the past two weeks. 6GB of RAM on the motherboard had been operating in Dual-Channel mode during these tests. When in actuality, it's supposed to have been running in triple channel. So these benchmarks 'may' improve. We'll see!

In the meantime, spazzatack, share your benchmarks.

Happy overclocking!