Once again I will join the fray and offer my newbie opinion to all you hardcore computer professionals, overclockers and the like. :>)
From what I have read on lately on Tom's also from alot of you guys discussing it here, it seemed to me the pin array *looks* more solid and sensible than I imagined. I can see how if the pins were flimsly or broke easily that would be a major deal for the motherboard makers. However, from looking at the new article:
<A HREF="http://www.tomshardware.com/motherboard/20040619/socket_775-04.html" target="_new">http://www.tomshardware.com/motherboard/20040619/socket_775-04.html</A>
<A HREF="http://www.tomshardware.com/motherboard/20040619/socket_775-12.html" target="_new">http://www.tomshardware.com/motherboard/20040619/socket_775-12.html</A>
as well as reading the comments the "land grid array" seems to make a bit more sense to me now:
"LGA (Land Grid Array) technology comes from the high-end server arena and allows for higher pin densities as well as very consistent electrical and mechanical characteristics. Less force needs to be applied in order to secure physical contact. According to an Intel PR manager, the main reason for introducing LGA 775 was the larger pin count, which makes it easier to supply power to future processors. Although Tejas has gone, we can be fairly sure that future CPUs will require a decent amount of power - particularly if you think dual core."
"While the pins have been part of the CPU before, they are now relocated to the socket. Most of Taiwan's motherboard makers are afraid that their motherboard return rates may skyrocket due to damaged LGA 775 sockets. Of course they would have to answer for that type of damage in some way. Considering the ongoing price war between Asus and most of the smaller manufacturers, costs will become a matter of life or death."
from the same article. Also:
"It is not difficult to handle an LGA 775 motherboard and processor with care, as long as you keep a few things in mind. Try to avoid touching the underside of the processor as well as the pins inside the socket. Make sure you insert and remove the processor without any insertion or removal angle. There are two notches in the upper and lower center of the frame (see third image on the next page) that allow your fingers to grab the processor without having to fumble around in any way."
"When you close the socket frame, make sure that the processor is inserted properly, and check that pressure is applied equally to both sides of the chip before locking down the lever. You're done!"
and furthermore:
" Everybody is complaining about the new socket being fragile, and MSI decided to do something about it. G52-M7001X3 is the code name for a CPU insertion tool that will be included with all MSI LGA 775 motherboards. It's an interesting add-on, but careful users will find it unnecessary."
From looking at the picture the pins actually seem thicker and sturdier than the ones on the chips currently. It still could be an annoying problem for some motherboard manufacturers if people try to put the cpus on with a hammer and duct tape, but it doesn't look like it will be nearly the issue I had invisioned when reading about it. Also, perhaps for some of you techs who make a living assembling computers the "hysteria" about it might mean less people want to do it yourself. Just a thought.
--------------
From reading that article it looks like the pin array is the least of the worries with the new platform. I would think before they sent samples to Tom's they would have worked some more of the kinks out. At the very least they should package the cpu with a fan that will cool the chip. That seems like a glaring newbie error in my book. I am sure it was a rush to get it out, but still that would have been a simple easy remedy to one of the concerns the Tom's review voiced. On the other hand I imagine, due to the prestige of Tom's testing reputation, that some of the feedback will be applied to production. Indeed it appears to me that the whole thing is still somewhat in development:
"This article was the biggest project THG has ever done, the cope of which is analogous to Intel's ambitions with its eye popping number of new features. Considering the amount of snafus that arose over the course of last week, I personally would say that Intel over extended itself by insisting on the platform launch at this very moment. Four more weeks could have ensured that all components were available. "
As far as overclocking goes I can see why they wouldn't want people to overclock if it runs that hot stock. Something I read in the article seems to indicate to me that perhaps the current design is approaching "maximum warp" anyway:
"The new top model will be the Pentium 4 Processor 560, which runs at 3.6 GHz. Four other versions, with model names 550 to 520, run at 3.4, 3.2, 3.0 and 2.8 GHz. While the TDP was 103 °C for the socket 478 Pentium 4 Prescott at 3.4 GHz, the maximum power output now will be 115 W. As a result, there's not much overclocking margin with the top models. More than ever, you should keep a close eye on the thermal parameters if you plan to go for a LGA 775 machine."
I certainly wouldn't buy any of this stuff, but then again I never would buy the brand spanking new technology. I think it is interesting how ddr 2 has been out for a bit with presumably nothing to run it on in anticipation of this release. It occurs to me that as amd has the jump on advancing 64 bit intel has made the jump in advancing several other new techs. It seems all pretty hypothetical to me at this point but I suppose they have a detailed roadmap of how they plan to develop the tech. On the one hand one company gets the jump and has their patents on that "path" on the other hand the other company can sit back and see what they did wrong. I think it is interesting to see how the two companies develop their strategies.
I imagine I will get a couple "thank you captain obvious comments." :>)