Depressed. Is it dirty silicon or noob syndrome?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

navecko

Junior Member
Dec 15, 2005
16
0
0
So a 0.05 fluctuation is bad? We have discovered something other than the silicon that could be hampering my overclock! Now to figure out why it is fluxing. The PC Power & Cooling Turbo 510 is reviewed to be the most stable psu. The a8n32 is claimed to be rock-solid with the 8 phase power. That leaves one thing, the fan controller. I will remove it and see how the voltages do. To be continued...
 

xtknight

Elite Member
Oct 15, 2004
12,974
0
71
Just think of me...I have to use an A8N-SLI Deluxe to overclock my 3500+ (which has been unsuccessful at best). And my memory is '100 MHz' at auto. :) Anyhow make sure HTTxLDT is <1000 and your PCI/SATA buses are locked?

0.05V fluctuation is very stable, for any of the rails. My Vcore fluxes about 0.10V with NeoPower 480 but the CPU is still rock solid (at 2420 MHz ClockGen and 1.425V).

Could very well be the mobo as well. My P4P800 D was great for OC, but I am not happy with this A8N-SLI D. I don't think I have gotten it to POST with any OC whatsoever and AI NOS causes stutter problems. However I can use ClockGen just fine. Have you tried ClockGen? Try raising HTT and test with Prime95 for stability each 5 MHz HTT or so. You may need to increase voltage off and on too. Good luck.
 

Diogenes2

Platinum Member
Jul 26, 2001
2,151
0
0
Something to think about..

If AMD was getting a high yield of 2.9 silicon, they would be selling FX85's or whatever ...

You got a good chip.. Too bad you paid FX60 price for it..

I'm happy to have an FX60 at an Opty170 price...:D
 

formulav8

Diamond Member
Sep 18, 2000
7,004
522
126
Whoever said voltage flucuation is bad IS wrong.

The Venice/Toledo based cpu's are Designed with core flucuation on Purpose. Normal voltages for stock are varying voltage from 1.35v-1.4v. That is normal.

So, your cpu is just fine with it varying the voltage.



Jason
 

markkleb

Banned
Feb 25, 2006
202
0
0
I just put together my system (again) but this time I used a Biostar mobo. In the overclocking section it said "DO NOT TRY TO OVERCLOCK THE FX PROCESSORS"

They must be close to their limit already.
 

3chordcharlie

Diamond Member
Mar 30, 2004
9,859
1
81
Originally posted by: markkleb
I just put together my system (again) but this time I used a Biostar mobo. In the overclocking section it said "DO NOT TRY TO OVERCLOCK THE FX PROCESSORS"

They must be close to their limit already.

Well being the fastest available chips (and at one time being the only high end chip on teh older production process), they will generally be closer to their limits than other processors.

But they are 'supposed' to be overclocked, and 'supposed' to let you reach the highest performance with the least hassle.

If ALL of them were capable of 3 ghz right now, there would be an FX62 (@2.8ghz) already, without a doubt.

Based on most people's OC results, 2.8 ghz is just out of the range that dual core chips can run with voltages that AMD is willing to release under warranty.
 

wwswimming

Banned
Jan 21, 2006
3,695
1
0

i wouldn't be too bummed. i mean, i know the feeling, i was bummed because budget limitations force me
to go for the Corsair value select instead of the Patriot 3-2-2-8 when i bumped my memory.

anyway, 3 GHz it is !

since you said that the system craps out when the temperature gets to a certain point ...

i suggest reducing the temperature of your input air. for example, buy a dehumidifier
and put the air intake for your computer next to the output of the dehumidifier. this
will lower your input air temperature.

otherwise, if you're keeping your input air fixed at "room temp.", whatever that is,
i'd say get a box & use the Zalman 9500, which has a thermal resistance of .12, line
up the fans so that you have 68 F air blasting through the Zalman.

if the CPU is putting out 200 watts, the base of the Zalma will be about 24 degrees
c above 68 f. 68 f is 20 c. the base of the zalman will be about 44 c.

given that that's right next to the top of the CPU, it shouldn't be too hard to
reduce the temp. differential from the bottom of the zalman to the top of the
CPU to less than 12 c.

jeez, nice computer man. OCZ low latency memory & the asus A8n32.
couldn't do better than that. i have an asus P5GD and a P4C800 that i use
the most.
 

robertk2012

Platinum Member
Dec 14, 2004
2,134
0
0
Originally posted by: Diogenes2
Something to think about..

If AMD was getting a high yield of 2.9 silicon, they would be selling FX85's or whatever ...

You got a good chip.. Too bad you paid FX60 price for it..

I'm happy to have an FX60 at an Opty170 price...:D

Lol then what do I have. An Fx-63 at the opty 165 price ($265) :)
 

robertk2012

Platinum Member
Dec 14, 2004
2,134
0
0
Originally posted by: formulav8
Whoever said voltage flucuation is bad IS wrong.

The Venice/Toledo based cpu's are Designed with core flucuation on Purpose. Normal voltages for stock are varying voltage from 1.35v-1.4v. That is normal.

So, your cpu is just fine with it varying the voltage.



Jason
If im not mistaken they were designed to fluctuate with Cool and Quiet. If it is of there should be very little fluctuation. My cpu voltages fluctuates .01 volts and I wouldnt want to see .05 or .1 volts fluctuation.
 

robertk2012

Platinum Member
Dec 14, 2004
2,134
0
0
Also the fluctuation is from 1.49 - 1.55 which is a fluctuation of .06. These leads to unstability as your voltage changes but clockspeed does not. What are the voltages under load? Is it the lower of the two numbers or does it fluctuate under load. How about idle?