In a year you'll be probably wanting something else...
Printable View
In a year you'll be probably wanting something else...
How come there are no board available in the US. I have one but I still don't see them available anywhere.
i have the v2000 also... awesome case but the upside-down thing does cause problems sometimes........ by not compatible I guess you mean the heatpipe would not flow properly?Quote:
Originally Posted by chunkylover77
I think this is the first time ever to compare DDR500 DDR400 timings on the same system, at the same CPU and Northbridge speed.
This was possible by implementing the hidden 5/4 divider on Athlon 64 Rev E CPUs.
Comparing DDR500 - DDR400
DDR500 1T, 3-4-4-10-14-18-3-4 Sandra=5775
DDR500 1T, 3-3-3-8-12-16-3-3 Sandra=5775
DDR400 1T, 3-3-3-8-12-16-3-3 Sandra=5514
DDR500 is faster by 260 points.
Comparing 1T - 2T DDR500
DDR500 1T, 3-4-4-10-14-18-3-4 Sandra=5775
DDR500 2T, 3-4-4-10-14-18-3-4 Sandra=5274
1T is faster by 300 points.
Comparing 1T - 2T DDR400
DDR400 1T, 3-3-3-8-12-16-3-3 Sandra=5514
DDR400 2T, 3-3-3-8-12-16-3-3 Sandra=4654
1T is faster by 860 points.
Comparing different timings with the same tCL at DDR400
DDR400 1T, 2-2-2-6-9-14-2-2 Sandra=5614
DDR400 1T, 2-3-3-8-12-16-3-3 Sandra=5574
Difference 40 points. (almost nothing)
Comparing different timings with the same tCL at DDR500
DDR500 1T, 3-3-3-8-12-16-3-3 Sandra=5775
DDR500 1T, 3-4-4-10-14-18-3-4 Sandra=5775
Difference 0 points.
Comparing tCL2 - tCL3
DDR400 1T, 2-3-3-8-12-16-3-3 Sandra=5574
DDR400 1T, 3-3-3-8-12-16-3-3 Sandra=5514
Difference 60 points. (small)
Comparing different timings
DDR400 1T, 2-2-2-6-9-14-2-2 Sandra=5614
DDR400 1T, 3-3-3-8-12-16-3-3 Sandra=5514
Difference 100 points.
Comparing DDR500 1Τ - DDR400 2Τ (BIOS Default)
DDR500 1T, 3-4-4-10-14-18-3-4 Sandra=5775, PCMark05=5759
DDR400 2T, 3-3-3-8-12-16-3-3 Sandra=4654, PCMark05=5621
DDR500 1T is faster by 1121 points.
In this case we have a 25% Sandra boost and a better PCMark05 score by 138 (2,45%).
Conclusions
The most important DDR speed factors in my case were 1T and memory frequency in this order.
Should I dissable 2T at DDR400 or leave it as is and raise the memory frequency at DDR500?
At DDR400 1T I got a 5514 score while by just raising the frequency to DDR500 2T I got 5214 points.
ASUS did a great job with 1009 BIOS.
Auto memory settings gave me the best stable speed as long as I enabled 1T manually.
My Ballistix are DDR500, 3-4-4-10-14-18-3-4.
With default BIOS settings on earlier than 1009 BIOS I would lose a quarter of my DIMM's speed and 2.5% of my PC speed.
This is how I checked.
Athlon 64 4400+ A8N32-SLI Deluxe BIOS 1009.
All tests run at the same BIOS default settings (CPU 201X11=2211MHz).
I used the latest Sandra 2005-10-10-69 version dlded from extremeoverclocking.com
My timings are: tCL-tRCD-tRP-tRAS-tRC-tRFC-tRRD-TWRT
DDR400 was 201MHz
DDR500 was 244.44MHz
(manually calculated, since no program understands a 5/4 divider)
Sandra's 2005 Memory bandwidth Benchmark scores were taken as an average of 5 checks.
I've decided to flash my BIOS from 803 to 1009 by CD as it said in the motherboard manual that you could do it by doing the ALT + F2 method. So I made an empty disc with A8N32SLI.ROM on it and press ALT + F2 during the bootup, flashed fine and after that it keeps saying bad BIOS sum error :mad:. It then proceeds to BIOS backup and it still says the same problem when it flashes even after redownloading the BIOS.
I also tried using the Asus support CD which has the original BIOS on and it still says the same thing :(
Hopefully a floppy drive (which I didn't have in the first place) will fix this problem as I'm getting on from a friend today :(
Decided not to be lazy right after posting that request and googled it. I used NVflash and the same Nvidia ref BIOS as you. BTW NVflash allowed me to flash both cards 'in situ' you dont have to remove them and only use the top slot. It detected one, flashed it, then detected then next and allowed me to flash that too, all went very smoothly.Quote:
Originally Posted by lowdog
I can also confirm it works, can see the whole 2048MB of ram now, it also seems to have fixed a problem I was having with COD2, it used to crash when I selected dx9 renderer, it runs just fine now. So far I have not noticed any detrimental effects, thanks again,
Paul.
This is not a problem.Quote:
Originally Posted by Cybertronic
The first time you may get a Bad checksum error but you should enter the BIOS and load setup defaults.
Someone on another forum said you always get the checksum error when you change the name of the BIOS file to A8N32SLI.ROM from whatever it was distributed as. It seems the file name is included in the checksum.Quote:
Originally Posted by Marios
I got it also due to the name change but all was well after loading defaults.
like M2 aka AM2? yes lolQuote:
Originally Posted by ReelMonza
just curious - what is the max people are able to set their K8>NB multi and also the NB>SB ?
i have k8>nb set to auto and it sets itself to 3x for my 275fsb but after i flashed to 1009 bios the nb>sb setting is 2x now??
just wondering if anybody can take this board over the 1000 freq for k8>nb and also can the sb>nb be = the k8>nb setting?
I've fixed my BIOS issue by flashing to version 903 (probably was caused by 1009 flash) :stick:Quote:
Originally Posted by Cisco
I don't want to flash to 1009 again incase I get the same problem :stick:
I wouldn't mind trying to flash using the Asus update but I'm getting a problem:
Unable to create NT driver service for BS_DEF.DLL
and after that:
Setup program detect no ASUS motherboard.
Installation will be aborted now.
Does anyone know the solution to this please?:(
i dont have the a8n32 but I have the a8n-sli which is similar, and I have run 270HTTx4 without issues while testing ram (memtest). Which would be 1080, not much over spec but still, seems ok.
I'm having trouble OCing this board.. are you using the ram in your sig?Quote:
Originally Posted by LuckMan212
I flashed my bios from 903 1009 under dos with no problems
well my rigs are in serious flux right now........ no, when I tested the 270htt I was using 2 sticks of Patriot TCCD at 2.5-3-3-7 1T cpu at a 9x multi and used Memtest86+ 1.65 for 5 passes each test#5 and #8. How is your cooling? might be a good idea to lift your NF4 chipset fan and apply some good AS5 and re-seat...... 270htt is no big deal at all lots of DFI guys are hitting 350-400 HTT without much effort
Yup...I've gotten the same message the 2 times I tried installing it...I haven't bothered trying to troubleshoot it but knowing how poor a lot of ASUS instalers have been in the past...they'll probably get around to putting out another version soon.Quote:
Originally Posted by Cybertronic
In the meantime think of it as a "feature"...that windows updater is a dangerous piece of software. I wanted it for an easy way to change the boot logo but I would never use it for flashing the BIOS.
EDIT: Just realized in another thread someone already had the melon-head cat as their avatar...so I switched to one i know no one else has since i took the shot myself :)
I've gotten this message a number of times w/ my old Premium board. The only thing I found to fix the problem was a complete new Windows install. It's probably a registry issue or something.Quote:
Originally Posted by Cybertronic
Well Folks, I finally got my A8N32 this week. And only started overclocking today. I'm impressed. I may have been blessed w/ a good cpu but I reached 2.75ghz on my X2 4400+ w. hardly any tweaking. Most everything is still at default. Now, I set my vcore to 1.4v w/ cpu overvoltage enabled. Thru AiBooster, it shows 1.54v.
http://i20.photobucket.com/albums/b2.../A8N32SLI2.jpg
I flashed to the 1009 bios from the perfectly working 1008 except i wanted raid to work. Shouldn't have done that. Infinite BSOD just when the Windows loading bar is about to show up during boot. Hopefully it's only the OS install that got doomed. This is really a pita.
Just so long as you understand that it is probably 1.6v real. Add 0.06v to AiBooster every time, you'll be about right. DMM test points are either of the two capacitor holes bottom left (connector panel on your left) of the CPU.Quote:
Originally Posted by Capt_Caveman
Quote:
Originally Posted by mackanz
I'm using 1009 and raid is working fine for me. do you see your array before going into windows?
Yay!
Finally got the raid to work. New files needed to the floppy together with the 1009 bios.
Super Pi Mod 1.4 at 2Mb
CPU was at 201x11 in all cases. This was done with the 5/4 divider (250 limit) to get 250MHz of mem frequency.PHP Code:
DDR T CL RCD RP RAS RC RFC RRD RW Average
DDR500 1 3 3 3 07 10 14 3 3 01:25,862
DDR500 1 3 3 3 07 09 14 3 3 01:25,868
DDR500 1 3 3 3 07 09 14 3 2 01:25,897
DDR500 1 3 3 3 08 12 16 3 3 01:25,946
DDR400 1 2 2 2 06 09 14 3 1 01:26,338
DDR500 1 3 4 4 10 14 18 3 4 01:26,856
DDR400 1 2 3 3 08 12 16 3 1 01:27,328
DDR400 1 2 3 3 08 12 16 3 3 01:27,601
DDR400 1 3 3 3 08 12 16 3 3 01:27,896
DDR400 2 2 2 2 06 09 14 3 1 01:28,149
DDR400 2 3 3 3 08 12 16 3 3 01:30,324
All other BIOS settings were identical.
Average was extracted from 4 measures.
There are pictures for every measure available.
Here is an example.
http://users.forthnet.gr/ath/mloupos/125797.png
There was a of 0,150" deviation between the upper and lower score of the four measures.
So there is no actual difference between the average scores of the first four lines.
Notice that no program reports the correct mem frequency.
Hi everyone! I need some advice: I just received my new X2 4400+ CPU (CCBWE 0529) and in a couple of days I'll get my hands on a brand new A8N32-SLI Dlx (they sent me the P5N32-SLI board due to a shipment error, dammit!). I am shifting from a DFI SLI-DR with a venice 3500+@2,65ghz(265x10)/1,47v.
I am experienced in overclocking, but I am just a bit scared about this CPU voltage madness with this Asus board. I do not want to pass the 1,5v limit even if I have a Zalman 9500 cooler, so how should I set the bios (1009) to get the REAL value of 1,5v? I am a bit confused...:stick:
thanx in advance
Just don't enable the cpu overvolt option in bios and it should stick fairly closely to what you set the vcore to.Quote:
Originally Posted by arrarro
very nice comparison, thanks for sharing ;) :toast:Quote:
Originally Posted by Marios