A64 OC Data

Status
Not open for further replies.
(cf)Eclipse said:
i've got that zalman.. it takes a while to heat up, but once it does.. it's not budgin any higher. the highest i've ever been able to get mine to go is 60ºc at 2.5ghz and 1.85v.. too bad i can't run high volts through my chip all the time.. *shrug*
what board do you have though? i don't really recall anything about any boards misreading low.. it's always high, so you should be safe for another 10º or so.
Click to expand...
Ha! Mine is better does 2.48 on 1.74... and I was almost worried my ClawHammer sucked...
 
im shocked your running 4x ht thats just sweet most people ive seen cant run that high.

the thing with HT is that lowering does not impair your performance. infact, its been shown that an HT of 2x is best for gaming. and the differences in benchmarks for 2x 3x and 4x is very minimal.
also lowering the HT one and 2 nothces absolutely allows for higher overclocks. that was one of the first things said about the HT technology. you really must lower it to gain top overclocks.
but you have proven that wrong.
man, that MSI board was my 2nd choice for mobo's and i wish i had gone with it, i have heard great things about it. this asus is good but overclocking is kind normal. nothing special.
my chip is the mobile and ive never, out of all theplaces ive asked had anyone be able to tell me what vcore voltages are my limits for low limit and high limit. anyone want to take a stab at that for me ? im already at 1.6 and have tried 1.625 but didnt notice anything good or bad about it.
So hats off to you man for crankin her so high! way to go.
 
The differences are so small that they fall well into the margin of error... I, too run 4x HT. nF3 250 are supposed to be able to handle 1GHz.. Your board isn't a very good one. K8N Neo Platinum is much better; but then again there's even a better one-- DFI Lanparty 250UT... that board is w00t. Up to 1.9vcc, and up to 2.11vcore in BIOS without any mods... not to mention up to 3.8vdimm. What's there not to love? ;)
 
Yep, I can concur on the OCability of the 3000+ CH's. After this past week burning in, A64 3000+ Clawhammer is up to 2.5 (10x250) with stock voltage, 1:1 mem (2x512 Kingmax Hardcore DDR500 TSOP's @ 3-3-10-3, stock vdimm) and a dab of AS5 under a Thermaltake Silent Boost HSF. All this on a VNF3-250 mobo. Temps are between 29°C idle and 41°C in games... It might help that I have 1x120mm intake and 6x80mm outake (including 2 in PSU).. That's when I have and ATA133/7200RPM HDD installed...

*HOWEVER*

When I use an SATA/7200RPM HDD, I can only do about 235 FSB stable with all of the same specs above...
 
what board do you have though? i don't really recall anything about any boards misreading low.. it's always high, so you should be safe for another 10º or so.
Click to expand...

I've got the DFI UT board. Everything works great on it, and I'm already happy with the overclocking results. It's funny, because I don't really have a NEED to push this any farther, just a WANT. :)

but I guess that desribes most people here...
 
I just bought it today. Haven't had time to test it yet; but I'm pretty sure it won't disappoint :D. Oh, what's this 4.0vdimm BIOS? excellent.
 
goss -

in order to keep using SATA with high overclocks you need to do 1 of 2 things
go get yourself a SATA pci controller card. i see them being sold/traded all the tiem, and infact here is where i got mine.
or
use the secondary SATA controller. my 2nd system is a VNF-250 with a 3200+ clawhammer. so you have to use the secondary sata to overclock cuz that particular controller is part of the pci bus and not part of the main controller. if you use the standard sata controller it does not fall into the pci lock. thus after you hit around 230mhz the sata controller frequency becomes too high and the drives become unusable.
switch it to the other controller and i bet it runs fine above 235.
the bad thing about my VNF3-250 is that i only have Corsair valueram on it,, and that shit is 150% fudging crap. im talking crap! even for not overclocking its crap but overclocking i can only get 215 out of it. and im not even using sata drives.
i need to find better ram ggrr.
 
Update- who the fuck was designing this board?! It violates just about EVERY specification. The 4-pin power plug is too far away; I have to stretch my cable so hard that the socket bends. The 12-pin power plug is on backwards. The floppy plug is on backwards. The chipset is WAY too close to the AGP slot; even closer than on the Neo Platinum! And the slots themselves are misplaced; I can't fit my video card without seriously bending it; and ripping off a capacitor, or two. GG...
 
imperator said:
I'm getting 260mhz x 10 = 2600 , and 4x HTT on my 3200+ 90nm winchester. I'm using air cooling with the Thermalright XP-120 on a MSI K8N-NEO2-Platinum/54 (yes it fits with a little bending). My memory is 2x 512mb Crucial ballistic PC4000.

Been running for 4 days now , seems entirely stable , played thru most of halflife 2 without any problems.

It boots into windows at 2700mhz but crashes on sisoft sandra, im only giving it 1.60vcore thou.
Click to expand...

I can't get much over 2.3GHz!!! I had to pull down the multiplier after 214MHz. :( I think its my ram holding me back, then again what more can I ask for out of generic?
 
Ok, giving what lestat said i've gone down to 2x HTT , and im now running 2680 mhz at 268x10 , and it seems stable, but i've only run a few sisoft runs and 3dmark05 will judge after a few more days use.
 
I have mine at 270 and 3xHTT and I use my Nforce sata with no problems. I run my main array off of it with a second array on the parallel ata raid - my board is a gigabyte K8ns Pro.
 
groebuck -

there are certain mobo makers who incorporate an SATA lock into the agp/pci lock. its sort of an unwritten lock. but it does exsist. but not many of them do it.

and i hate you guys so bad lol for getting such high overclocks. but damn is that sweet. and you 2 guys are only on air ? correct ? again way to go.!!

just out of curiosity i bumped my HTT to 4x and then started overclocking at 210, right now im at 220 and climbing.

there is one thing i have heard and infact can prove becuase of i own the K8N-E Deluxe.
even though we know our cpu/ram/mobo can handle above 235 fsb the asus board cant do a instnat jump to higher than 235. you have to go from 200 to 235 then reboot then go back to the bios and continue higher. for whatever reason it just wont allow you to just up so high without rebooting. now i dont know what that does or anything but its something some people with this mobo and others might want to think about.

ive always been an msi fan.. and ive always had nothing but troubles with asus, the only reason i got this mobo was everything i read, including the msi forums since i frequently was their with my 865PE NEO2 FIS2R was that A64 platinum board and infact all the msi A64 boards were very picky about ram. and i didnt want to take a chance that my ram wouldnt work,, so i got the asus as it wasnt as picky.
guess i should maybe start looking for a msi board. now thats no guarentee ill get better overclocks but it does put me back with a mobo maker whom i have always trusted to produce quality products.

i even have everything in my bios i dont use turned off. all the parallel ports and sata that im not using.
but one thing about this mobo, and also one i noticed with other boards is that you have to have the onboard SATA enabled or the pci lock dont work. you dont have to enble the raid, but just the SATA controller. why its like that i dont know... it just is.
 
Yes i'm on air Thermalright XP-120 with a medium speed Pappst does a really good job, infact my northbridge temperature is higher than my cpu temperature at full load. (NB is 47 celcius at load , while cpu maxes at 45 celcius.)
 
i wish someone could comment on my question i asked earlier, about the proper voltage limites of the 3200+ clawhammer mobile chip...
 
v_lestat said:
i wish someone could comment on my question i asked earlier, about the proper voltage limites of the 3200+ clawhammer mobile chip...
Click to expand...

What's their stock voltage? Most people don't go over 1.65 on regular A64s (ClawH and NC, both 1.5v stock) and 1.6-ish for 90nm A64s (1.4v stock), I gues mobiles would fall somewhere in between the two at best. That's all for aircooling.
 
well it's essentially the same thing as the regular desktop chips, so..
i'd say 1.65-1.7v with good air cooling or 1.75-1.8v with water cooling

edit: damn, i was beat. and the stock voltage on mobile chips is usually 1.4v unless you have a dtr, then it's still 1.5v
 
Dpends on your cooling in the end. Since I'm lazy and I don't feel like rumuging about for posts, I'll tell you what I know.

- On air cooling, Athlon64s shouldn't use voltages about 1.65v. That's hear-say. If you had some very powerful air cooling (XP-120 with a 100CFM fan) then I think you might be able to go higher.

- With more ellaborate cooling like water, then I guess there shouldn't be any limit. Most Athlon64 boards don't go higher than 1.85v from what I can remember. And I think at 1.85v, you should have hit the wall...

All-in-all, it all depends on your temps. I'd just try to stay away from a load temp of 50Cs...
 
wow ive NEVER heard 1.8v shit.. im doing 250mhz on ddr with 2.7v and 1.55v on cpu @ 2.5ghz. i wonder,,, i guess i could go higher and see what happens. if it dont like the voltages the mobo wont lallow it to but so.. ill give it a try.

many thanks for the info gent's
 
You can do 2.5GHz on 1.55?!! I need 1.85 to get that stable... I'm moving to watercooling, so I won't fry my chip when I run it at 2.11vcore, though :)
 
well i usually run 1.60v and it dont make a difference, and i just tried 1.65 and it didnthelp either., looks like its one of 2 things either its the max speed of this chip or the fact that i am only at 2.7v vdimm which i think the vdimm voltage is the problem im so pissed at asus for only making 2.7 the max. im either gunna get the msi board like i wanted to in the beginning or ill find an ocz ddr booster and see if that helps. either way i knew asus was a mistake.

go ahead and go to 2.11vcore,, no water in the world is gunna save you, kiss that chip goodbye.. but chances are since the mobo;s are the way they are these days the mobo wont even let you boot cuz of the overvolt protection the chip and mobo have.

and i ran a full pcmark suite and got 4389 iwth HTT 4x and 4396 with HTT 2x so.. no difference at all in speeds. again 7 points is nothing more than a brain fart in the system causing it to have a higher score.
but that score was only at 230mhz or 2.3ghz so...
 
Of course it would let me go to 2.11vcore. And good watercooling should prevent it's untimely death.
 
should being the key word there. i would do it just to see how high i can get the chip, but i'll never run over 1.9v, even with water cooling
 
that might be true, but what would i do with the extra power? my gpu is most often the limiting factor cause i run games with the quality all the way up, and if not, i have vsync enabled cause i hate the flickering when it goes over the refresh rate of my monitor.
also note my sig. 1.45v right now ;)
 
I think, in your case, the CPU will become the bottleneck at some points. If it's not, then you obviously haven't OCed your video card enough :D

Then again, in my case, I have a desire for higher clock speed / lower memory timings, rather than a need.
 
heh, well this is as high as i can get my videocard without it overheating.. i'll do more once i get around to getting a water cooling system in my comp, but that's a ways off. and i remember that need for speed with my computer.. i kinda gave up after i had everything figured out though, and i knew what i could do at what voltages/timings.. though once i get back to college and pick up the ballistix that will be sitting in my mailbox, i'll have a field day. be sure to check for stuff around wednesday/thursday after thanksgiving...
 
Try the 1.4beta7 bios, worked much better for me than the official 1.4 bios.
 
i suppose you used to own an k8n neo not too long ago then? i've heard good things about 1.37 too. i'll try it out once i get back to my a64 system (i'm at my parents' for thanksgiving right now)
 
iddqd said:
I still have it. Going to sell it, though.
Click to expand...


what board you getting ready to sell man ?

and your 1.45v is man you really should go with 1.5 or 1.55 minimum and even 1.6v if your going to overclock at all. and running that much on air is no problem as long as you have a decent HSF.
and dont forget to yank your vdimm voltages up to atleast 2.8 or 2.9v and higher if you have good ram.
 
i noticed something though, with 1.45v i can do about 2225mhz stable, but i'll need 1.55 to get to 2300mhz. then 1.7 to get to 2400mhz. i like to just keep things cool for the moment cause i have nothing that'll use the extra 150mhz i get from raising the voltage by .25v
and about the ram, i always run them at 2.85v, the maximum my board allows without mods. how else do you think i got some old kingston hyperx that's probably based on ch-6 to 220mhz at 2-3-2-6
as sone as i get my new ballistix ram, and i can do 230mhz or more, i'll bring the cpu clock up to keep the ram near it's max.
 
somethying is wrong if you need to raise your voltages to 1.7 JUST to get to 2.3ghz. thats definately wrong.
you should run 2.3ghz at 1.55

the reason you cant get any higher isnt your cpu voltages its that ram.
drop it to 2.5-3-3-10 or 5 you WILL NOT i repeat NOT see any performance loss anyohne who tells you differently is full of crap. there is no visual performance loss by raising tha CAS that little bit. and you should run either 5 or 10 on the last cas setting simply cuz with the A54's those are the 2 best settings to use.

but your voltages are whack man there should be no reaosn in the world to run 1.7v just to get to to 230mhz. its either that or you got a bad overclocking chip.
 
did you even read my post. i need 1.55v for 2.3ghz... 1.7 is for 2.4

and i can lower the ram down, and give it a 5:6 divider after i go past 230htt (230mhz at 2.5-3-3-10 is the limit for my ram.. 231 fails prime after about 5-10 min), so the ram definitly isn't the limiting factor when i really go for an overclock.
 
ok i got ya on the ram

and as for reading your post it was a reading error i didnt get enough of reading rainbow fucker lol kidding.

did you even read my post. i need 1.55v for 2.3ghz... 1.7 is for 2.4

you have to raise you voltages .15 to go to 2.4 from 2.3 ? i still say thats wrong, meaning your voltages are too high for that little of an overclock.
now im not saying my chip is better or yours is bad but i can run 2.5ghz @ 1.55v becuase mine is a mobile it runs lower voltages, is yours the desktop chip clawhammer or mobile,, i thought it was mobile. also the mobo can be a factor so..
who knows. as long as you dont go much higher than 1.7-1.8 you'll be fine i guess.
can someone back me up on that,, 1.55v @ 2.3ghz i fine but 1.7 just to go to 2.4ghz seems like too large of a bump.
maybe its just me..
oh lastly without looking thru this huge thread what core is yours CG or C0 ? mines the CG
 
it's a cg.. though yeah, it's a really poorly oc'ing cg. kinda frustrates me, but whatever.. i can live with it. and i would raise the voltages if i really wanted to go faster, but at the moment i'm perfectly happy with my undervolted overclock, plus it runs cooler than stock, so it makes me happy (small dorm room + overclocking = too much heat for me.. even with the windows open)
 
Couple of things about your posts lestat.

and your 1.45v is man you really should go with 1.5 or 1.55 minimum and even 1.6v if your going to overclock at all. and running that much on air is no problem as long as you have a decent HSF.
Click to expand...

1) The best voltage is the LOWEST one that will stably run the given speed. No one should pump more voltage than necessary through the chip.

somethying is wrong if you need to raise your voltages to 1.7 JUST to get to 2.3ghz. thats definately wrong. you should run 2.3ghz at 1.55
Click to expand...

2) The reason the chip was sold as a 3200, not a 3800 is because it was only running at the 3200 speed within the margins they wanted (mostly, there can also be variability due to demand and suchlike, but . . . ) so the fact that someone else doesn't get the overclock you do doesn't mean anything is wrong. It can just as easily be that they didn't have the same overhead you did.
 
That's besides that point that that 3200+ is s754 a ClawHammer, while 3800+ are s939 NewCastle. :D
 
fine, then make it a 3700+, it's similar enough anyhow :D
etruscan said:
2) The reason the chip was sold as a 3200, not a 3800 is because it was only running at the 3200 speed within the margins they wanted (mostly, there can also be variability due to demand and suchlike, but . . . ) so the fact that someone else doesn't get the overclock you do doesn't mean anything is wrong. It can just as easily be that they didn't have the same overhead you did.
Click to expand...
thanks for pointing that out to him, i was getting a tad frustrated ;)
though more cause i can do 2.3ghz at 1.55v, but no, nobody listens.. hehe
 
i totally agree that the lowest voltages to higher overclocks is best.

i was just saying he was making too big of a jump in voltages from 2.3ghz to 2.4ghz he want from 1.55 to 1.7 volts just for 100mhz, thats not right by any overclocking thats too big of a jump for such a small overclock jump
i just meant exactly that,, not that he should CRANK it.
and besides for any chip the defualt volts is 1.5 or 1.55 and ill gladly take correction on that if im wrong but id swear without running mine at defualt, that the defualt voltage when you use the mobile chips in a desktop is 1.5v but ill probably be wrong there.
 
Status
Not open for further replies.
Back
Top