If anybodies interested I can post some benchmark results with the P5B Deluxe vs the P5NSLI at the same clocks next week with my lowly 6800GS's
If anybodies interested I can post some benchmark results with the P5B Deluxe vs the P5NSLI at the same clocks next week with my lowly 6800GS's
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
why not...
Anybody else come to the conclusion that the low benchmarks in SLI on this board is because of very poor performance in the CPU tests? Particularly '03 and '05
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
No it's because the two pci-e slots aren't 16x (the second is 4x) and most importantly, the latter is controlled by the sb.
Project ZEUS II
Asus Rampage II Extreme
Intel I7 920 D0 3930A @ 4.50GHz (21 X 214mhz)
3 x 2GB G.Skill Trident 1600 @ 1716MHz (6-8-6-20-1N)
2 x Asus HD 6870 CrossFire @ 1000/1100MHz
OCZ Vertex 2 60GB | Intel X25-M 120GB | WD Velociraptor 150GB | Seagate FreeAgent XTreme 1.5TB esata
Asus Xonar DX | Logitech Z-5500 | LG W2600HP 26" S-IPS LCD
Watercooling setup:
1st loop -> Radiator: 2 x ThermoChill PA120.3 | Pump: Laing DDC-3.25 with Alphacool HF 38 top | CPU: Swiftech Apogee XT | Chipset: Swiftech MCW-NBMAX | Tubing: Masterkleer 1/2" UV
2nd loop -> Radiator: ThermoChill PA120.3 | Pump: Laing DDC-3.2 with Alphacool HF 38 top | GPU: 2 x EK FC-6870 | Tubing: Masterkleer 1/2" UV
Assembled in Mountain Mods Ascension Trinity
Powered by Corsair Professional Series Gold AX1200
Well this is what lead me to my conclusion
Top score is my P5NSLI, bottom is the P5B deluxe, exact same clocks. You will note the biggest differences are in the CPU scores. I've noticed stuttering during the CPU tests in both 05 and 03.
The CPU scores return to normal when not running SLI on the P5B
Both P5B, one SLI, one not. Note the CPU scores
Last edited by Dr_BenD_over; 09-15-2009 at 03:33 AM.
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
can we get some SLI bridge/ no bridge scores on the P5B Deluxe??? ?
I can run some, but I don't think there will be a huge difference with only 6800GS's.Originally Posted by TheGoat Eater
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
Ben Dover?...for shame!
Gigabyte EP45-DQ6 - rev 1.0, F13a bios | Intel Q9450 Yorkfield 413x8=3.3GHz | OCZ ProXStream 1000W PSU | Azuen X-Fi Prelude 64MB X-RAM| WD VelociRaptor 74HLFS-01G6U0 16MB cache 74GB - 2 drive RAID 0 64k stripe | ASUS 9800GT Ultimate 512MB RAM (128 SP!!) | G.SKILL PC2-8800 4GB kit @ 1100MHz | OCZ ATV Turbo 4GB USB flash | Scythe Ninja Copper + Scythe 120mm fan | BenQ M2400HD 24" 16:9 LCD | Plextor 716SA 0308; firmware 1.11 | Microsoft Wireless Entertainment Desktop 8000 | Netgear RangeMax DG834PN 108mbps; firmware 1.03.39 + HAWKING HWUG1 108mbps USB dongle | Digital Doc 5+ | 7 CoolerMaster 80mm blue LED fans | Aopen H700A tower case | Vista Home Premium - 32bit, SP1
The SLi Vs non-SLI scores are pretty convincing in terms of final value. What drivers were you using? You seem to be getting a much bigger benefit from SLI than Coolaler didOriginally Posted by Dr_BenD_over
Thanks for the tests
What FSB were you running? after 400FSB, the chipset strap changes, that might be part of it
84.56_Mod2, they seemed to be a bit more stable than the 85.96's
Those are totally stock. I can't say I've been impressed by my runs at 2804 (401x7). Almost thought about putting the P5NSLI back in...
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
Thanks 401 will kill memory bandwidth.. if you feel like it, please could you try both boards below 400FSB? 380 or something
As for the tests on the P5B SLI Vs non-SLI, I can only guess at whats happening. SB load is being passed to CPU? CPU needs extra cycles when SLI is enabled, at least it did in the good ol days
-> just tried to use my PCI-E 4x for some bench runs... damn SATA ports get in the way, so nowt happening here tonight except for thinking
Haven't had a chance to do any unbridged tests yet. Whatever the glich is it doesn't appear to affect '06 very much with 6800GS's. My 5717 run last night pretty much shows that.
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
I had 2x7600GT and got around ~5100 '06 marks on P5B Deluxe. then I've added the SLI bridge, and got ~6500. might be a reason to see slow SLI results because no bridge used ?
i7 920@3.8, 6GB @ 1540, 2 * 295@630, P6T, 1Kwt, 24" LCD.
6800GS SLI rips 7900GTX SLI on this board (comparing Dr Bens results to Coolalers)
possibilities (the way I see it)
*driver differences
*the lower spec of the 6800 is not eating up so much bandwidth, especially on the 4X slot so there is no bottlenecking, although there is a CPU hit
Any input? I`d love to see Coolalers opinion on this
@Dr Ben...sorry to ask more of you, but if you get a chance please can you try the other 3DMarks?
That was my hope when I got the board. I'd kick myself for doing so, if only the 680i boards weren't so expensiveOriginally Posted by K404
Anyway, I'll try some other benches this weekend (I hope).
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
4858 = bridged
4626 = unbridged
3.2Ghz, cards at 425/1000 (stock)
Last edited by Dr_BenD_over; 09-27-2010 at 10:31 AM.
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
How did you get the 9114 score in `06? (Post 105) Thanks for trying the extra benches Numbers look about right for full SLI results.
That was an '05 runOriginally Posted by K404
|Q6600 G0 @ 333x9|P5Q Deluxe 0803BIOS|HD5850|2x2048MB|XFI XM|
Bookmarks