Need Help OC'ing 3500+ Venice. Unstable at 210FSB

Berix

n00b
Joined
Jun 14, 2006
Messages
24
Hi guys. I'm trying to overclock my AMD Venice 3500+ E3 processor on a MSI K8N Neo4 Platinum motherboard. I got up to 210 FSB with an 11x multiplier (CPU at 2310 MHz), booted up fine, but when I tried to run the OCCT 30 minute Stability Test, my entire computer froze 12 minutes into the test. It was frozen for several minutes before I noticed that nothing was moving and progress had been stuck at 37% for a while; the mouse didn't move, Ctrl-Alt-Del didn't work, totally unresponsive so I reset the computer and loaded the default settings back into the BIOS. Here are the settings I used in BIOS for my OC try:

==DRAM Configuration==
Timing Mode = [Manual]
Memclock index value (MHz) = [133Mhz]
CAS# latency (Tcl) = [2]
Min RAS# active time (Tras) = [ 5T]
RAS# to CAS# delay (Trcd) = [ 2T]
Row precharge Time (Trp) = [ 2T]
Row cycle time (Trc) = [Auto]
Row refresh cyc time (Trfc) = [Auto]
Row to Row delay (Trrd) = [Auto]
Write recovery time (Twr) = [Auto]
Write to Read delay (Twtr) = [Auto]
Read to Write delay (Trwt) = [Auto]
Refresh period (Tref) = [Auto]
User Config mode = [Manual]
Bottom of 32-bit[31:24] IO = [D0]
1T/2T Memory Timing = [2T]
Read Preamble value = [6ns]
Async Latency value = [6ns]
S/W memory hole Remapping = [Disabled]
H/W memory hole Remapping = [Disabled]
MTRR Mapping Mode = [Continuous]

==Cell Menu==
**Current CPU Clock: 2310 MHz**
**Current DDR Clock: 136 MHz**
High Performance Mode = [Manual]
Aggressive timing = [Disabled]
Dynamic Overclocking = [Disabled]
Adjust CPU FSB Frequency = [210]
HT Frequency = [4x]
HT Width = [down_16 up_16]
CPU Spread Spectrum = [Disabled]
SATA Spread Spectrum = [Disabled]
PCIE Spread Spectrum = [Disabled]
PCIE Clock = [100Mhz]
SSE/SSE2 Intructions = [Enabled]
Cool'n'Quiet = [Disabled]
Adjust CPU Ratio = [StartUp]
Adjust CPU VID = [StartUp]
Extra CPU Voltage = [By CPU VID]
**CPU Voltage: 1.400V**
Memory Voltage = [2.75V]
NF4 Voltage = [1.50V]

I feel like I may have missed something or there may be something that's limiting my overclock that I didn't see. It's running at stock voltage, but I don't want to up the voltage, especially since load temps are 47'C and I'm not comfortable giving more volts. But it seems like that my OC should be able to go higher, other people I've seen running at stock volts can get like 2.5GHz so I'm thinking my chip may be a dud OC'er too.

Something important I should probably note is that SpeedFan says my +12V line is bouncing between 12.83v and 12.77v (which is out of the +/-5% spec), but Everest and the BIOS reports my +12V as switching between 11.92v and 11.86v. And for the +3.3V rail, SpeedFan reports it as 3.14v which is within spec but Everest reports it as switching between 3.36v and 3.38v (out of the +/-5% spec).

I've also tried using Clockgen to overclock but when I shutdown my PC after I've used ClockGen the system won't POST (it'll turn on though but nothing will happen, just a long beep and black on the screen) and I have to hit the power button to turn the system off and press it again to turn it on again before it'll POST. I don't know why Clockgen's doing that but I've avoided using that program to OC so far since I think it may be doing something to my computer it shouldn't be doing -- after all, it should POST if I turn off and turn on the PC immediatly after, shouldn't it? But I can give Clockgen a try if you guys think it'll be a good idea. I've looked on MSI's support forums and I've found someone else that's having this problem with Clockgen.

I'd like to say thanks for any help with my OC in advance. :) I'll gladly provide more information if I missed anything.
 
Although I do not have any decent advice on your overclock, I do have a little advice for this:


}Something important I should probably note is that SpeedFan says my +12V line is bouncing between 12.83v and 12.77v (which is out of the +/-5% spec) said:
I have never had a program read my volt levels correctly. If you really want to know them for sure, grab a multimeter and test them that way.




perhaps try locking your ram at 166 instead of 133? I've had ram be lame when I clocked it down too low.
 
I'll 2nd the ram speed. I've had 2 sets of ram that refused to be clocked too slow. Even oc'd the ram wanted it's factory speeds maintained ie: 2-3-2-5 1t, etc
 
I tried lowering the divider to 166MHz but this time OCCT freezes and the system becomes totally unresponsive in 6 minutes instead of twelve when I had the 133MHz divider. However, I did run Stress Prime 2004 for fifteen minutes on "small FFTs" for a little over 15 minutes and it ran fine. My load temp is 46'C.

CPU: 210 x 11 = 2310MHz
RAM: 165MHz (with 166 divider)
HTT: 840MHz (4x multi)
vCore: StartUp

I hope someone can help me get a higher OC. :)
 
i would double check that your ram is not intermittent..

http://www.memtest86.com/

also are you able to change the multiplier on a a Venice? sorry been out of o/cing for a while.. but i would leave the ram at stock timmings/fsb and work on seeing how high of a stable o/c you can get just tweaking the multiplier.. then you have that out of the way and you can work on attaining that o/c with your ram/fsb..
 
I had similar issues with my venice 3500+ on my dfi lanpartyUT SLI board. With phase cooling the max OC I could get out of it was ~300mhz over stock. It was very lame. Switched to an X2 and now I have a comfortable 800mhz OC with a 3800+. I think in my case I just had a bad chip, as the board will go to 305mhz and the mem to DDR500 @cas 3. If you haven't already, you should check out this thread to see if you may have missed anything.
 
hmm, that's doesn't seem right.

I consider my 3500+ venice not a great overclocker but I can get 2.5ghz at stock volts.

My computer seems to have a vaguely similar problem, in that it won't fail a benchmark, everything will just lock up completely instead.

What RAM have you got? Possibly even trying your RAM even closer to stock might help . . .
Mainly though, I would try and get a multimeter, so you can check the voltages of your rails. .
Dodgy rails can cause freezing under load.
 
Back
Top