Is my mobo the problem here?

[johnoh]

I've got a 6-8-1 1800+ with vcore set at 1.8v up from the default of 1.6v. Not sure of the other details because I can't read the cpu.

Its on an ecs k7vta3 v3 board which is not stable above 140 so I just keep it at 133. Generic pc2700 memory that runs fine at 180mhz on another board but is set at SPD on this board. Temp is 40C at load with an sk-7 and sunon 80mm fan.

The system is great at 15x133 but gives boot up errors at 16x133. I'm using socket pin connecting to set the multi and vcore. I see some people getting 2.2g-2.3g with their DUTC 1700/1800 at 1.8v and am wondering if I just got unlucky or if somehow the board is the problem. vcore fluctuates between 1.78v-1.81v. Any advice is appreciated.

 

[Gautam]

Sounds to me like you've got an early Thoroughbred B. Seems its doing too well to be a T-bred A. It's probably a lower batch, such as a JIUCB, instead of the JIUHB's that are all the rage these days. It should still prove to do well, but it will need more than 1.8v. Your cooling should prove to be good enough to take you to around 2.2, but with a vcore of 1.9v or so. Your motherboard is going to hold you back, but not in clock speed. That is to say that's its not going to restrict your overclock in any way, but is going to take a toll on your overall perfermance.

 

[johnoh]

Thanks Gautam that's what I was looking for. I didn't think the board would matter as long as the voltage was relatively stable and I'm using a 133 fsb but wasn't sure if somehow it mattered in another way. I read this morning where a guy said he got 200mhz more out of his cpu by switching boards at the same vcore but I don't know how that can happen.

I will probably stick with 15x133 since I prefer to run the fan at 5V for quietness but then it puts me at 60C at full load and another vcore bump would be too hot for my tastes. Not to mention 1.85v is the max when using pin connecting - the k7vta3 has no vcore adjustment built in.