Yea the die simulator is a requirement. No question.

But I am concerned about ONLY testing on the die simulator as I have been able to pick out a lot of mounting issues and problems that end users would experience by actually installing and testing on a real system too.

Seems to me that this is an important consideration for reviews, and that I should really continue to do such testing. Since SocketA is EOL, I will sooner or later have to upgrade. Socket 754 w/ IHS makes the most sense to me.

for mounting issues, yes
a deficiency of my testing was that it was limited to the wbs performance and did not address ease of use, robustness, tubing issues, etc.

Well Socket A will live at least a few more months under the Sempron, and it doesnt seem that motherboard manufacturers have given up on it - they're still coming out with new boards. Some blocks aren't Socket A compatable, but most still are or have Socket A versions... when 462 is really EOLed then you can start making 754 noises - they should be even cheaper then :)

Does procooling have an official beggar position - someone(s) who constantly is prestering OEMs for gear for testing and/or test bed freebies? I'm sure a few of them would love to hear their gear on the test bed just so that people can see it exists. Yeah, I know we have talked about it before, but lets talk about it again.

that would be me. I am not too active on this because the more I beg the more I have to test :)

But quietly I bugged Koolance and PolarFLO over the weekend...

Nah, you need to decouple the marketing department and the testing department. Less bias that way =) And more of your time can be spent testing. Any one here can do 80-90% of the ground work to get the testing gear - maybe you can just close the deal. That might end up saving you a significant amount of time/effort.

Let us help.

N.B. some mfgrs are PR whores - its all good
others may be very picky, particularly where ranked lists are involved

over the longer term pH has a problem:
how to induce a mfgr to test when that mfgr's product may not compare favorably ?

I suspect the 'front runners' will participate, and no others (why should they, can't 'win' ?)
- be prepared to buy stuff

but you are assuming perfect knowledge here. Some mfgrs aren't totally aware of our reviews and will send a few blocks before they realize they've been torn up. Others will think their new stuff is significantly better than the last thing we rip apart. There are tons of premutions on that theme.

But, yes, to get a "full coverage" approach, we will either need to start buying things OR we need to get in touch with a reseller (newegg, directron, etc) who have deep enough pockets to pony up our missing blocks. If the article is good enough (eg. a deep and technical 15 item round up) we could probably get many of those from a reseller - as long as we mention their name a few times. That has the added benefit of not inducing bias on the result field.

That works for the big names Tempus but smaller mfgrs (as in this thread) don't deal with resellers.

Polarflo have quite a few resellers..
http://www.polarflo.com/index.asp?Pa...on=Custom&ID=2

As stated before, any manufacturer who has done testing will know how they stack up vs. the competition, and avoid sending their blocks here unless they expect good results.

Maybe once you've got the die sim set up and can take all the numbers so that retesting wouldnt be required, you could auction off the blocks after use to make the money back?

Ah so they do. SVC looks like a likely sponsor; Joe's been working on them for MWL sponsorship I think.

might even be able to get the koolance block from them

Would buying a Polar Flo TT from Frozen CPU be bad?

not at all, Mark is a nice guy - the other fellows too
but I'm not sure Mark will give pH one

Bill, is that a typo in your location (switching the , for a .) ?

on the other hand, if you are putting effort into making a good testbed, you might as well future proof it and go 754.

I thought 754 was dying out and 939 was the way of the future

If I recall the last (hell, soo many changes) roadmap it was that 754 would stay the single channel (duron/sepron, mobiles, and possibly standard (re non-poweruser) A64s) and the 939 would be the dual channel workstation and desktop replacement mobiles while the 940 would stay on as the dual channel server (gotta love that registered crap.)

But I'm very possibly wrong and I'm more than likely out of date and just half remember a conversation.

If you have better info let me know.

From an enthusiast point of view, S939 is the way forwards.

S754 will effectively become the budget/low-end platform of the AMD line up to replace Socket A.

It would seem to me that it would make most sense to follow S939 from this point in time. S754 as a "flagship" enthusiast platform is effectively over, or at least will be within 2-3 months as availability and pricing gets better.

that's my (very limited) understanding as well

But both socket shares same mounting holes and very close packaging dimensions: total heights, length and width and thickness of IHS, as well as thermal output?

I need a DTR A64 CPU, anyone got the height of that bare CPU silicon?

Yeah, 939 will probably be the defacto standard for real power, unless the rumor about them changing sockets AGAIN when the release the DDR2 Cpus.

I think nikisub1 pops the lids of his CPUs, perhaps he would know that dimension

He's an Intel user though. Possible to just look at the tech documents for the AMD64-mobile maybe?

Right. Intel at this point.

yea, then make the assumption that . . . .
better a measurement I think