1700+ Overclocking for DC effort. Dasm Stress testing :(

[Overkiller]

Hey all,

I'm testing my 1700+ overclocking abilities..with SEti of course

RIght now it is running 10x200 @ 1.5V (stock) fine...

Seti driver gives me an estimated WU time of 2:45...

That seems sort of high...wouldn't you say?

What are some WU times of 1700+ o/c systems...or I would like to compare them to Barton's around 2.0Ghz as well.

Thanks everyone!

 

[soni]

1700+@181*11.5 : 14 day avg. result from my setiqueue: 2h12m CPU time
This is with RAM @CAS 2 2 2 5 on a 8RDA+

 

[Wiz]

Don't believe the estimated time - it's never right

My two 1700+ systems (one at 2.1ghz & other @ 2.2ghz) both do 1:55 - 2:15 WU times depending on the Angle Range & how much the system is being used.

 

[Overkiller]

Setispy is saying 2.67hr/Wu now

I wont believe them. Thanks for the update guys..Going to check temps and then up to 2.1 ghz...hopefully i wont have to increase voltage!!

Will update accordingly

update: still 10x200 but now memory is 2-2-2-5..Voltage is stock for memory as well (only 2.6V i believe..)

Wow, this system has a LOT of headroom!

Going to find the limits of my ram with those 2-2-2-5 settings & then the cpu :Q

Corsair Pc-3200 is at 210 MZh (2-2-2-5) with a bump up of voltage to 2.7V...I do not want to increase voltage past this so the ram will stay @ 210.

The processor at 2100mhz after running DC for 30 min is only 38C...WOW!

more more more

 

[Coolkid]

See if you can keep raising that FSB
(Why u not online in MSN? )

 

[Overkiller]

Seem to be hitting a wall around 2.2ghz..

Upped the voltage to 1.525..no go...now up to 1.55V

Temps are ONLY 38C (and abits report higher temps anyways...the diode is right underneath the cpu) so i believe i just needed that extra voltage boost.

So right now:

Proc: 10.5x210 = 2.2Ghz @ 1.55 (up .55 V :/ from stock)
Ram: 210 @ 2.7V (up .1V) @ 2-2-2-5

Lets hope this is stable :Q
[edit] unstable and i dont really want to push voltage more than1.55..even though its not a big amount. Maybe i'll give it a go @ 1.575 or 1.6V because temps are still low

 

[ProviaFan]

My 1700+ Tbred-B (an older model that starts to get iffy at 2GHz) at 1.8GHz (11*166MHz) does a WU in about 2:40 or something (not exact because I'm visually "averaging" the times I see under the completed units list of SETI Watch). It might go a little higher, but this is my main system, and I don't like having to take it apart constantly to reset the CMOS due to a borked overclocking effort, or having it crash on me during a critical task.

 

[Assimilator1]

You got a good cpu there

1.6v is nothing to worry about for a Tbred B ,the higher end Tbred Bs run at 1.65v by default ,look here
And by AMD them selves here ,if you can handle the stupid side wards chart!

I've got mine running at 1.75v ,it does get rather warm in the summer though!
Oh yeah ,I've got an XP1700 @2.04GHz ,177FSB ,dual channel ,nfrc2 ,2.5-2-2-6 & I get 2hrs 28mins/ 0.4xx WU
SETIspy's prediction isn't accurate until the WU gets to nearly 50% complete

Your lucky you've got one of the 1.5v def core 1700s ,mines a 1.6v ,hence yours overclocks better.
If I were you I'd just crank up the MHz & vcore until you hit 1.7v

You've got a copper HS right?

[edit]hurry up & come back online I wanta know!

(I use too many smilies!

)

 

[Overkiller]

Lol Assim!!

Yeah, been running 10x210 @ 1.5V for many hrs now...its completing WU's in....2 hrs 4 minutes!!

Awesome! Basically that is equivalent to my drastically overheating (and thereby throttling) P4C rig (solution coming soon )


Wow, i am loving this AMD experience. Now i regret going intel

I should have never strayed so far from the roost

I am going to check temps now and give it 1.6V....be right back w/ updates.

 

[Assimilator1]

Those P4s with HT still rock at SETI ,with 2 processes they have great output ,though you don't get the low WU times & are much more expensive
Seems like your P4 & XP rigs are matched for output atm ,12/day,if you can overclock the P4 though it'll do better still
It'll be interesting to see which is ultimately faster.

Btw you didn't answer my question further up

 

[Overkiller]

Well the system crashed sometime last nite....

I bumped down the ram to 200 just to rule that out of the equation..and bumped up the multiplier. Running 11x200 now = 2.2 GHZ @ 1.6 volts....Hopefully its stable. Going to leave this alone the whole day and see. If it crashes but temps are good should i just up the voltage some-more? As i said..temps = <40C full load for many hours..so i'm thinking that the 210mhz @ 2-2-2-5 was the upper limit of the ram last nite..oh-well we will see...

Oh and assim, the P4C i have..i just know it can hit at least 3.0..just gotta get that video card so i can set it up (it's in the mail..yay...) again with a proper cooling solution.
And yes, the HSF that i have on the Athlon does have a copper HS...its a:

Thingy

oh well..testing stability is always fun chore....:|

as usual, keep you guys posted :0

[edit] OK, i ran prime 95 under torture test and "blend" (tests ram a lot). failed within a few seconds @ 10x205 @ 1.55V...now running the Small FFT (ram not tested at all)...So i'm leaning towards Ram being the issue for the moment...updates coming later!

 

[Assimilator1]

Not a bad HSF ,though at full bore it must be noisey!:Q

Mine's the GW CAK4-88T ,the HS is 10mm taller than yours but the fan isn't as powerful (same size though).
Yeah mine's damn tall & damn heavy!,it exceeds the clip rating! ,mine is also an early version which only grabs 1 lug
(did you notice the fan specs of the CAK4 86!?:Q)

Btw only up the voltage more if its needed ,best program I've found to test CPU/RAM/mbrd stability so far is Prime95 ,use the tortue test to check stability.10mins for just a rough test & several hours for a proper test.Keep upping the CPU MHz until it crashes in a quick test of P95 & then up the voltage until it passes the 10min test ,then do a longer test.

[edit]I see you're already familiar with P95
Until you find your CPUs max speed & voltage that your comfortable with I suggest you keep your multiplier high & your FSB low(keeping RAM within spec) ,once you know your CPUs max MHz then you can drop the FSB & test the RAM seperatley

[edit]Ahh the joys of having free multipliers

 

[Overkiller]

At full bore the fan for the Cak-86 must be a demon!

I believe this fan runs @ 40-45db full bore

OH well the chip is kept cool...
~~~~~~~~~~~~~~~~

-I'm now using Prime95 to test stability. I brought the system all the way back to 10x200 (2-2-2-5) and am moving up slowly now.
-I had to up the ram voltage to 2.7V (2.713V actual) b/c it would not do the fast 2-2-2-5 settings properly (the ram torture test in prime would fail).
-I am now running the cpu @ 1.55V (1.503 Actual)
-Prime95 Torture test on Blend (tests fpu/l2 cache, max heat/powr consumption, and ram alternatively)...

If prime95 fails at this I'm going to test just the small fpu, then the large one..if those pass i know my ram is the factor.

Let's hope for the best!

 

[Assimilator1]

Best way to elimate cpu speed uncertainties & test your RAM is to underclock your cpu ,e.g 7x200+ ,then you know for a fact the only thing causing you instabilities is the RAM as you increase the FSB ,you can't relie on a program to tell you whther its RAM or CPU causing problems

 

[Overkiller]

10 x 200 @ 1.55V core(1.504V actual) 2.7V memory....

Stable 9 hrs prime.

 

[Assimilator1]

Nice ,now what about 11x200?

 

[screw3d]

I've got my 1700+ at 11.5x200 @ 1.70v, churning out an average of 10 WUs a day assuming that I don't do any CPU intensive stuff. RAM is sync-ed at 7-2-3-2.. a lil' conservative but it's fast enough for me.

 

[Assimilator1]

That's a sweet cpu you've got!,2.3GHz
I would think you'd get more than 10/day out of it though,what WU times do you get?

 

[Overkiller]

Ram means a lot also. I'm running 2-2-2-5 vs. 2-3-2-7..might be good for a few minutes here and there ( i think i noticed a 30 min improvement going to tight timings)...

finals weeek. little time for tweaking. bah...more updates to come

 

[screw3d]

Originally posted by: Assimilator1
That's a sweet cpu you've got!,2.3GHz
I would think you'd get more than 10/day out of it though,what WU times do you get?

I could make it run at 2.4GHz if I want to If I leave the computer completely (killing all unnecessary threads) it can come up with 11-12 WUs a day, but just 9-10 WUs with regular use.

I am not sure about the WU times though.. prolly about 2 hrs 10 mins? Donno..

 

[Assimilator1]

Ah, 11-12 is more like it ,& 2hrs 10 sounds about right.
Just wondered cos my XP1700 @2.04GHz does 'em in 2hrs 28mins or so.