Sorry they are 2gb
Hey any clue why clue why cpuz reports mem running at 266 instead of 276
Printable View
Sorry they are 2gb
Hey any clue why clue why cpuz reports mem running at 266 instead of 276
Newegg
Brand Crucial Technology
Series Ballistix
Model BL12864Z503
Type 184-Pin DDR SDRAM
So is this part at Newegg the same :banana::banana::banana::banana::banana:en fast Z503?
I forget, are you running a divider? That would give your a different setting in CPU-Z and would account for the difference your seing.Quote:
Originally Posted by yasoumalaka
Yeah I'm using a divider I had it set at 366fsb using the 3/4 divider which I thought would run the mem at 274 but mem test and cpuz recognize it as 266. I eventually tested the bandwidth at 266 at 1/1 and got the same result as 366 at 3/4. So I think some how the mem is running at 266. I'd like to get it to 274 though I have no clue how to do it.
Now I don't know exaclty how to explain this but just a few weeks ago somebody either here or over at guru3d did explain to me why this is! What I mean is that you could be using a FSB setting of 300Mhz with a 5/6 divider but end up getting an actual RAM speed of 244Mhz instead of 250Mhz which is what the math would suggest. So there is an actual reason this happens yasoumalaka I just forget.
CorrectQuote:
Originally Posted by Roger_D25
The memory speed on A64 is not divided from the FSB(HTT) but the CPU speed. It has to be this way since the memory controller is integrated in the CPU.
If you run 300 HTT with 9x multi and "5/6" divider the RAM divider is actually CPU/11 which makes 2700/11 = 245Mhz
The divider is not a constant. It varies as you change multi.
Some examples with 166/333 "5/6" divider
300*9 - 2700/11 = 245
300*10 - 3000/12 = 250
300*11 - 3300/13 = 253
300*12 - 3600/14 = 257
All the dividers might not be correct, but it shows why the memory speed never seems to make sense compared to the "old" way of calculating.
That is exactly what I was told the first time but forgot, thank you very much to explaining that again, much apprecaited!
Thank you
I went throughall the possible combinations and I can't get my mem to run at 275-280 without the 1/1 divider while keeping my cpu clockin the 2.9-2.97 range. Any help you can offer?
Which BIOS offers the most options????
Update: I just flashed to the latest bios 04/06/06 So Ill be playing around with it for now. Anybody know about odd divisor selection?Quote:
Originally Posted by yasoumalaka
Try 328x9=2952MHz
and 5/6 mem/cpu divider
5/6x328=273MHz.
If your CPU can go a little higher (2.97 you said?) you can get ram to ~275 with 330MHz FSB.
Good luck. :)
just leave it disabled.Quote:
Originally Posted by yasoumalaka
Thanks for the suggestion, but from what I understand LlamaOfDoom that would make my mem run at 268-270.
2952/11 - 2970/11
My results so far with Samsung UCCC. Memtest stable up to 280mhz, but so far prime stable large ffts at 270. haven't tried more, but 280 immediately errored. My drive strength at default was normal in a64 tweaker, so i found 4 and 5 to be the best. For some reason I only need 2.5v to be memtest stable, but 2.65v to prime :(
http://img141.imageshack.us/img141/1...able5kn.th.jpg
Any ideas?
I was wondering if the Crucial Value with Micron -5B D has the same problems as the Ballisix?... The dieing problems..
290. Oh boy. :)
http://masterwoot.com/ocresults/9x290.jpg
Validated.
New batch of ballistix, got them straight from Crucial so i have great hopes!
http://img149.imageshack.us/img149/4...01276zr.th.jpg
My new rev ballistix clock about 5mhz worse than my old rev, but they can run tighter timings. Good luck with yours eBoy.
Considering they use the same chips, I suppose they do. I've got the Crucial value myself, and they've been working flawlessly for six months now. They OC to around 260 with 3.1 vDIMM, and 240 with 2.7 - hardly spectacular, but OK value for money.Quote:
Originally Posted by S!1v3rB@cK_Dk
I do not believe the memory modules were dying on the bad ballistix. I believe I read somewhere that it was infact the PCB's that were failing. If that is the case it could very well be that only Ballistix were effected. :(Quote:
Originally Posted by _damien_
and the fact that the .16TD1 revision of the ballistix don't seem to have any issues at all and still uses micron 5b D.... ;)Quote:
Originally Posted by LlamaOfDoom
Ok thx for the info. Oh btw, is 3.1v too much for the crucial valueram? And what times were u running at 240 & 260 Damien?
i wouldn't give 5b D more than 2.9v for 24/7 with good active cooling ;)
PM's sent to you, EnJoy, for the database.
Don't hold your breath, it has been almost four months since the last update....Quote:
Originally Posted by masterwoot