FSB/HTT on a8n32-sli

ninjacore · Nov 25, 2006

What msx FSB should I expect out of the A8n32? I just sold my a8n-sli premium because I could only get 286 out of it max. I'm hoping I can get over 300. Is this going to be possible?
thanks

Zx2Slow · Nov 25, 2006

I had 330fsb with mine, Perhaps your CPU is holding you back?

ninjacore · Nov 25, 2006

you had 330 on your a8n32 or did you get that on an a8n regular? My question is what max FSB everyone has been able to get on their a8n32's?

billb · Nov 25, 2006

ninjacore said:
you had 330 on your a8n32 or did you get that on an a8n regular? My question is what max FSB everyone has been able to get on their a8n32's?

I've been to 345 on the A8N32-SLI-D, though I think 330-335 max is typical.

Nebulous · Nov 25, 2006

billb said:
I've been to 345 on the A8N32-SLI-D, though I think 330-335 max is typical.

5

Is this with a divider or 1:1? Max i can get without it barfing is 285-ish, yet my ram did 310 on several DFI boards.

ninjacore · Nov 25, 2006

Nebulous said:
5

Is this with a divider or 1:1? Max i can get without it barfing is 285-ish, yet my ram did 310 on several DFI boards.

285!?!
that sucks. Was that just a fluke on yours or is that common? I need at least 300

Nebulous · Nov 26, 2006

ninjacore said:
285!?!
that sucks. Was that just a fluke on yours or is that common? I need at least 300

Dunno. I've tried several chips and different ram. Also tried several bioses to no avail. I can prolly get 290, but then @ that speed it'll fail to boot/post at times

It's gotton that now i just run stock speed and to hell with ocing. Kinda sucks too. :-/

QuietIce · Nov 26, 2006

I've tested my FSB to 310 and beyond that it wasn't worth the trouble for me since I wasn't going to use it. No that was not at 1:1, who has the money for 2Gb of RAM that can do that???

I know RAM timings are different than on some boards. I've read reviews on the Corsair 3500LL and a couple of them couldn't get it above 250 on other boards. I've had mine to 265 and didn't try to figure out the timings for the next step - didn't need it ...

ninjacore · Nov 27, 2006

so, you don't think at 300 will be a problem if I go with this board?

QuietIce · Nov 27, 2006

If you're talking about running a 165 @ 300x9 or a 170 @ 300x10 I don't think that'll be a problem. Also, vCore goes to at least 1.625v (been there, for a one-shot test) and vDIMM to at least 2.9v (and maybe higher, I can't remember).

Edit: Since I've never run my system as a whole on an FSB that fast (only did it testing the FSB speed) I decided to check it out. I really didn't want to experiment much and I'm very familiar with 246x11 (2706) @ 1:1 since it's my day-to-day setting. I set the chip @ 300x9 and the RAM at 166 (~245) with all other settings normal except the HTBus multi, which I turned down to 3x. The system ran SETI just fine for a quick, 15 min. test - no problems or unusual behavior at all ...

billb · Nov 27, 2006

Nebulous said:
5

Is this with a divider or 1:1? Max i can get without it barfing is 285-ish, yet my ram did 310 on several DFI boards.

Uh? What does a divider have to do with Max FSB?

I'm Just saying that the MOBO (A8N32-SLI-D) will do FSB 345 for me. Obviously, with a low multi and low divider. Though I do have some TCCD that will go 332.5mhz...286 x 10.5 with a 233 (7:6) divider, 3-4-4-8, 1T @ 2.75v,

Nebulous · Nov 27, 2006

billb said:
Uh? What does a divider have to do with Max FSB?

I'm Just saying that the MOBO (A8N32-SLI-D) will do FSB 345 for me. Obviously, with a low multi and low divider. Though I do have some TCCD that will go 332.5mhz...286 x 10.5 with a 233 (7:6) divider, 3-4-4-8, 1T @ 2.75v,

You just answered your own question with this statement: " I'm Just saying that the MOBO (A8N32-SLI-D) will do FSB 345 for me. Obviously, with a low multi and low divider"

That was what I was asking, ty. Not all boards/ram will do 300+FSB @ 1:1. The A8N32-SLI-D (which I have in my current setup) refuses to do 300, doesn't matter which chip, ram or bios i use. Now on the DFI (Ultra-D & Expert) i was able to do 310+ and that was 1:1. I don't use dividers.

ninjacore · Nov 28, 2006

thanks guys. I didn't end up going with the a8n32 for two reasons: 1) When I upgrade next fall after basic, I'm going to want AM2 anyway, so I might as well go a bit cheaper than 120 now; and 2) When I upgrade at that point, I'm also going to want DX10, not SLI, so...
I went the DFI Ultra-D. It is supposed to go up to 335 stable 24/7, so that should be more than enough to max out my 4400.

thanks again for the advice

synthetic_fenix · Nov 28, 2006

I think the max I have gotten out of mine was 325 but I haven't tried since I went to watercooling and put the fan on the heatsink on the back of the board.

billb · Nov 30, 2006

Nebulous said:
You just answered your own question with this statement: " I'm Just saying that the MOBO (A8N32-SLI-D) will do FSB 345 for me. Obviously, with a low multi and low divider"

That was what I was asking, ty. Not all boards/ram will do 300+FSB @ 1:1. The A8N32-SLI-D (which I have in my current setup) refuses to do 300, doesn't matter which chip, ram or bios i use. Now on the DFI (Ultra-D & Expert) i was able to do 310+ and that was 1:1. I don't use dividers.

Perhaps I just see the same issue differently. If my MOBO will do 330 FSB, but with a !:! divider I can only get to a 285 FSB, I dont see that as a MOBO problem, I see it as a memory issue. Which leads to the question "Will changing latencies to get higher memory speeds (and hence a higher FSB) benchmark better or worse?"

As far as I know FSB has no effect what-so-ever on benchmarks.
In the attached, two benchmarks with equal CPU speeds and equal memory speeds, but different FSBs (200 vs 250) benchmark exactly the same.

QuietIce · Nov 30, 2006

Did you happened to measure temps under those conditions, too? My heat sensors are so bad I'd probably never see a difference.

It seems to me the higher "FSB" would use more power and generate more heat but maybe the reference clock is so small a part of the system it simply doesn't matter ...?

billb · Dec 1, 2006

QuietIce said:
Did you happened to measure temps under those conditions, too? My heat sensors are so bad I'd probably never see a difference.

It seems to me the higher "FSB" would use more power and generate more heat but maybe the reference clock is so small a part of the system it simply doesn't matter ...?

It was hard enough to find multis and dividers that would keep the CPU and memory speed the same at two different FSBs...and you want temps too. I don't think I can help as I have no way to measure NB temps. And at FSBs of 200 and 250, I'm so far underclocked that MOBO and/or CPU temp differences would be extreemely small.

Perhaps someone else????

QuietIce · Dec 1, 2006

I was thinking of taking the FSB/CPU to an extreme since you have that lovely FX-57. How about 300x10 @ 133 v 200x15 @ 200? That keeps the RAM the same even if it is under-clocked. It's just a thought. :shrug:

Not sure if it would prove much, though, without a decent temp sensor ...

jeep4x4055 · Dec 8, 2006

345?FSB

wait 345FSB from 200?!! how the hell do u do that the highest i could go with my 4400 x2 is 245 @ 2.7ghz is there something im doing wrong.?

ochungry · Dec 8, 2006

jeep4x4055 said:
wait 345FSB from 200?!! how the hell do u do that the highest i could go with my 4400 x2 is 245 @ 2.7ghz is there something im doing wrong.?

Its not a big deal really, the DFI expert can do 500mhz HT in lower multi.
I did 397mhz on the Asrock dual sataII (ULi1695 chip) and my best 24/7 was 345x8=2760mhz for an A64 3000 venice w/ that motherboard. I haven't tried the limit of this board yet but I am sure you can achieve 345mhz or 350mhz w/ a lower multi.

http://img216.imageshack.us/img216/1539/031720062200473kc.jpg

FSB/HTT on a8n32-sli

Member

Member

Member

Member

Dreadnought Class Senior

Member

Dreadnought Class Senior

Disabled

Member

Disabled

Member

Dreadnought Class Senior

Member

Risen From the Ashes

Member

Attachments

Disabled

Member

Disabled

Disabled

Member

Similar threads