Can non K haswell be OC'd up to their max Turbo frequency?

[SOFTengCOMPelec]

You could also turn it around and say delidding is much easier without soldered TIM. I mean, any _true_ enthusiast would dellid it right? :awe:

True, but if it is soldered on, then there is much less need to de-lid it in the first place.

What I would love to know, is the real reason(s) behind the non-soldered TIM.

Is it because the new smaller chips 22nm have technical difficulties with solder, or something on these lines ?
In which case I have sympathies with Intel, although these technical difficulties may be solvable.

Or is it to just save $0.01 on the production costs, of a $349.99 (or whatever) end user priced part ?

Or is it to hinder overclockers ?
Perhaps because they don't want Xeon customers to save money by getting K parts and overclocking them.

 

[R0H1T]

True, but if it is soldered on, then there is much less need to de-lid it in the first place.

What I would love to know, is the real reason(s) behind the non-soldered TIM.

Is it because the new smaller chips 22nm have technical difficulties with solder, or something on these lines ?
In which case I have sympathies with Intel, although these technical difficulties may be solvable.

Nope, I believe there was some Chinese guy that confirmed presence of fluxless solder on IVB-E engg samples so if he's right in his discovery then the only logical reason for Intel to not solder their chips anymore is to save a few bucks as you rightly said !

Or is it to just save $0.01 on the production costs, of a $349.99 (or whatever) end user priced part ?

Or is it to hinder overclockers ?
Perhaps because they don't want Xeon customers to save money by getting K parts and overclocking them.

The Xeon target market will hardly get a consumer level K part because the former comes with lots of validation testing & certain other features, like ECC RAM support, so unless they're trying to save some major $$ an enterprise customer will not choose i7 over Xeons, but hey that doesn't stop Intel from being a cheapskate & skipping solder for even an i7 :hmm:

 

[SOFTengCOMPelec]

Nope, I believe there was some Chinese guy that confirmed presence of fluxless solder on IVB-E engg samples so if he's right in his discovery then the only logical reason for Intel to not solder their chips anymore is to save a few bucks as you rightly said !

The Xeon target market will hardly get a consumer level K part because the former comes with lots of validation testing & certain other features, like ECC RAM support, so unless they're trying to save some major $$ an enterprise customer will not choose i7 over Xeons, but hey that doesn't stop Intel from being a cheapskate & skipping solder for even an i7 :hmm:

If it is to save $01, it makes me very angry.
Because I am paying $349 (or whatever) to Intel.
They are saving $.01

Then I as an end customer (even without overclocking), have to put up with potentially increased fan noise and wear and tear (because of more fan RPM), due to the extra temperature.
In theory the chip life is considerably shortened, because of the potentially higher temperatures (depending on cpu fan type and control mechanism).
In theory the cpu is more likely to make a mistake (soft errors), due to a potentially higher temperature.

Also in theory, the increased cpu temperature (in general, depending on fan), will increase the cpus power consumption (as temp rises, the power consumption at the same clock freq, increases), which over the whole life of the computer, and in these days of ever increasingly more expensive electricity, will cost an extra $1..$50 depending on electricity prices, length of usage and ownership of computer, what the computer does, and other factors.