Hehe, the same thing had me pussy footing around for a while in fear of breaking it . The otherwise excellent Dell guide didn't quite cover that one.

Noody

I editted the original post to state that grounding either pin works. I didn't think about it causing confusion.

Oh and Michael, I think it should be stickied too

Since it works on a 9300 it seems my question is kinda answered indirectly already.

Now for the direct answer to my direct question... will this work on the XPS2?

yes

Here's a successful 1.7 --> 2.26 mod!!! So far at least, don't jinx me. I've only had SP2004 stress testing it for 10 minutes so far, but it's working currently. Many thanks to all that contributed to this idea.

I ran 3dmark on my 9300 with the stock 1.6 533 cpu (the cheapest one dell had). I have 2GB RAM installed (from newegg). I ran it 3 time and got scores of 2991, 2989, and 2983. I then did the pimod. I now got scores of 3023, 2992, and 2994. The performance increase measures to be between 0.03% and 1.3%. Did I miss something or is this all I can expect to get from this particular benchmark?

Yes, you missed something. Read the very first line of this thread that is bold and highlighted.

your pinmod did nothing
please read

damn you beat me

I missed nothing. By stock, I mean the CPU that came with it. That was test #1 which produced the first three scores.

AFTER, I did the pin mod with a 1.6/400 CPU I bought from ebay. The BIOS shows 2.13 for both current and max, so I assume that means I did it right.

I was more curious if anyone else did 3dmark measures and had better increases than I had.

I probably should have elaborate more on the sentence "I then did the pinmod" to "I then did the pinmod with my newly purchased ebay 1.6 400 cpu" ... my bad.

i apologize

np.

I guess what I really want is an explanation of why or why not 3dmark is an indicator of "bang for buck and effort". I don't plan to undo this updgrade, so I was curious if anyone found better performance improvments for "before and after" scenarios using different benchmarks.

I might even try a 1.8 if I get one for real cheap. Not sure if it worth the risk (by risk I mean accidentially breaking one of the little plastic parts on the casing or such and thereby causing an annoying imperfection ... call me anal). I've seen references to prime95, which I've yet to try. Not even sure if this is just a testing utility or offers some sort of benchmark. I also see many folks using FPS game benchmarks.

Hoof,

It sounded like you pinmodded the factory CPU. Sorry for the confusion. Anyway, my 3dMark scores only improved slightly. It is mostly video card bound. I was surprised to see that there was little change when I did it. You will see more of a difference in something like superpi that is CPU intensive. Personally, I wouldn't try upgrading to the 1.8. It may work fine, but I don't think the small change is worth the effort to go through it again.

Thanks. Probably wont do it, unless I go for the 6800 Ultra upgrade. Not sure if all the issues are worked out of that yet. Maybe combine it with a 7800 upgrade if the future holds that for me.

It was fun and quite easy though. I have to hand it to Dell for providing such a good service manual. I always thought them to be a company that would make stuff hard on purpose so folks would have to pay them to do any kind of service.

And thanks to WhyOWhy for providing such good and detailed pics for as5 and pinmod.

A good "real world" test of processor speed would be to zip or unzip a large (500 MB+) archive and time how long it takes at different processor speeds. RMClock could be used to control processor speed for anyone interested. I may give this a go once I get my i9300 and finish the pin-mod but if someone else wants to try it first let us know your results.

Martian