Survivability of LGA775 socket

[Nocturnal036]

Basically, does anyone have any idea on how long Intel plans on utilizing the LGA775 socket?
I'm thinking about upgrading my system. I originally went with a 939 socket mother board but not a month after I bought it, AMD moved to the AM2 socket. Just wanted to know if anyone has any insight or inside knowledge. Any help is greatly appreciated.

 

[Drsignguy]

With the chips that are out now, coming out and well slow to get out, using 45nm, I don't see them "going away" anytime soon. Nehalem is the next socket change. But with that info, I would love to hear the More Inclined than I do the speaking on this issue. As for 775 well, I am going to stick with it for quite awhile. IMO, I think you would be safe.

 

[GuitarDaddy]

Yeah the platform changes seem to come quicker these days. The new 45nm penryn parts will be the last s775 chips. Nehalem wich is supposed to debut in the fall or end of the year is not only a new socket but at least three new sockets. Intels new architecture roadmap seperates the classes of CPU's with different sockets. Budget desktops(current E2XXX) on one socket, midrange(current E6XXX,E8XXX) on another, and highend parts on yet another. Evidently putting to an end the ability to buy a budget level chip and match it with an overclocking board and overclocking it to the same levels as highend CPU's.

 

[superstition]

Evidently putting to an end the ability to buy a budget level chip and match it with an overclocking board and overclocking it to the same levels as highend CPU's.

Translation: Kill AMD and then give consumers their choice of any color they like as long as it's black.

 

[narreth]

omg intel is trying to make a monopoly

AMD really needs to survive to save the future of computers....sigh

 

[Drsignguy]

Originally posted by: superstition

Evidently putting to an end the ability to buy a budget level chip and match it with an overclocking board and overclocking it to the same levels as highend CPU's.

Translation: Kill AMD and then give consumers their choice of any color they like as long as it's black.

Someone will find a way around this. When it happens, the word will spread. Life finds a way. Again, the evolution of a new chip.

 

[Twsmit]

Socket 775 will be the mainstream socket until the start of 2009. Once 2009 rolls around there will be a new socket for Nehalem. For enthusiasts looking for the high end fastest CPU's available 775 is dead in less than a year, after that Intel's integrated memory controller will transition into the mainstream.

 

[Rubycon]

About five seconds if you're careless with a screwdriver. :Q

 

[Idontcare]

Originally posted by: Nocturnal036
Basically, does anyone have any idea on how long Intel plans on utilizing the LGA775 socket?
I'm thinking about upgrading my system. I originally went with a 939 socket mother board but not a month after I bought it, AMD moved to the AM2 socket. Just wanted to know if anyone has any insight or inside knowledge. Any help is greatly appreciated.

I think you have your answer based on your own post. When S939 was released (edit: meant to say when you bought it, not when it was released, doh!) there basically weren't any more chips made for it other than the ones that existed at time of the socket release.

Look at LGA775 and realize that pretty much all the chips that will ever be sold for the socket are on the market now.

Intel isn't going to put Nehalem on LGA775, and Intel isn't likely to release any wolfdale/yorkfield chips with higher GHz clocks than they have already announced.

So in the sense that you felt S939 was obsolete when released (edit: meant to say when you bought it, not when it was released, doh!), so to is LGA775 already obsolete in that it is EOL with currently available/announced CPU's. No future for LGA775.

 

[Drsignguy]

Quoted by "Idontcare" So in the sense that you felt S939 was obsolete when released (edit: meant to say when you bought it, not when it was released, doh!), so to is LGA775 already obsolete in that it is EOL with currently available/announced CPU's. No future for LGA775.[/quote]



So, do you believe that the market will still hold strong for the 775 for quite some time? Even though there is no future for the socket, the availability will be there for the consumer far more and longer than say like the 478's?

 

[Idontcare]

Originally posted by: Drsignguy

Originally posted by: Idontcare
So in the sense that you felt S939 was obsolete when released (edit: meant to say when you bought it, not when it was released, doh!), so to is LGA775 already obsolete in that it is EOL with currently available/announced CPU's. No future for LGA775.

So, do you believe that the market will still hold strong for the 775 for quite some time? Even though there is no future for the socket, the availability will be there for the consumer far more and longer than say like the 478's?

I do believe it will hold strong in the sense that 45nm Penryn will likely see a lot longer life than previously expected. (previosuly meaning before Phenom performance and clocks became a "known")

As 2009 wears on you'll see cheaper and cheaper Penryns, so folks who have LGA775 boards which are compatible with wolfdale/yorkfields will get cheaper and cheaper upgrade opportunities.

But if you are just now buying a LGA775 board and equipping it with a decent wolfdale/yorkfield now then you have likely set yourself up with an EOL'ed system that won't likely have many options for upgrading (ever). You might have a speedbin or two to upgrade to in a year or so for real cheap but you are not talking about going to the next gen stuff.

Now if you are getting into an LGA775 with a really cheap dual-core (think E2140) then you have at least one really good upgrade opportunity in a year or two in which you swap out your cheap E2140 for what will eventually become a cheap QX9770.

 

[The Boston Dangler]

the longevity of a platform (LGA775) means very little to me. i'll replace my comp's guts every few years. cpu, mobo and ram come together, and go together. upgrading the cpu only over time only makes sense if you buy the cheapest possible chip in that socket early on, then buy the last generation to replace it. even with that, do you really want to put a 45nm quad core on a 965-based (or older) motherboard? would it even work?

buy something not crap, and ride it out as long as possible.

 

[qliveur]

LGA775 has already had a very long lifetime in the industry, as I think it came out in mid-late 2004.

It should come as no surprise that it's finally nearing the end of it's life.

 

[Zap]

Originally posted by: The Boston Dangler
do you really want to put a 45nm quad core on a 965-based (or older) motherboard? would it even work?

If it did, why not? The P965 was a decent chipset. I don't see newer chipsets as being vastly superior. Now, if you were talking about 945/915/925 series... yeah...

 

[GuitarDaddy]

Originally posted by: Zap

Originally posted by: The Boston Dangler
do you really want to put a 45nm quad core on a 965-based (or older) motherboard? would it even work?

If it did, why not? The P965 was a decent chipset. I don't see newer chipsets as being vastly superior. Now, if you were talking about 945/915/925 series... yeah...

I just upgraded my P965 with a Q9450, and it seems to equal the P35/X38's in overclocking ability and on performance benches. Nice upgrade to an almost 2yr old board IMO, should last me at least another couple of years

 

[demiurge3141]

Originally posted by: Idontcare
I think you have your answer based on your own post. When S939 was released (edit: meant to say when you bought it, not when it was released, doh!) there basically weren't any more chips made for it other than the ones that existed at time of the socket release.

Look at LGA775 and realize that pretty much all the chips that will ever be sold for the socket are on the market now.

Intel isn't going to put Nehalem on LGA775, and Intel isn't likely to release any wolfdale/yorkfield chips with higher GHz clocks than they have already announced.

So in the sense that you felt S939 was obsolete when released (edit: meant to say when you bought it, not when it was released, doh!), so to is LGA775 already obsolete in that it is EOL with currently available/announced CPU's. No future for LGA775.

There is one more chip to look forward to, the almighty e3140! Basically if you view socket 775 as a super budget overclocker's paradise you will not be disappointed. But if you want to always stay at bleeding edge, then yeah there will always be new stuff coming out. Even if the socket stays the same you will be force to buy new motherboards for new chipsets anyway. P945 isn't exactly good for penryns.

 

[The-Noid]

Originally posted by: GuitarDaddy
Yeah the platform changes seem to come quicker these days. The new 45nm penryn parts will be the last s775 chips. Nehalem wich is supposed to debut in the fall or end of the year is not only a new socket but at least three new sockets. Intels new architecture roadmap seperates the classes of CPU's with different sockets. Budget desktops(current E2XXX) on one socket, midrange(current E6XXX,E8XXX) on another, and highend parts on yet another. Evidently putting to an end the ability to buy a budget level chip and match it with an overclocking board and overclocking it to the same levels as highend CPU's.

Socket775/T lasted for 5 years, how can you say that is quick?

 

[GuitarDaddy]

I wasn't referring to S775 specifically, it has had a good lifespan. And S775 may be the last platform we see last that long. AMD has gone through three platforms in that time frame, and I'm guessing nehalem won't be around for 5yrs either.

 

[v8envy]

Socket form factor for 775 may have lasted for 5 years, but just being able to plug a QX9770 into a socket originally meant for a 133 mhz FSB Netburst P4 without bending pins doesn't tell the whole story -- PWM and other circuitry changes over the years mean there have been at least 3 different socket 775s.

That said, yeah, I was orphaned by s939 before I had a chance to upgrade as well, and some tiny bit of bitterness remains. AMD has gone through *5* desktop sockets over the last 5 years -- Socket A, 754, 939, AM2 and AM2+. Many of these overlap, of course -- but few socket form factors are economically viable for more than 2 years. Doubly so if you migrate to the platform once it's mainstream.

 

[whosjohnny]

If everyone in the United States donate just $1 , AMD will survive. For $1, you get hundreds of dollars in savings for the lifetime of competitive future. Who's with me on this one?

 

[bryanW1995]

I used to buy budget boards and upgrade the cpu after a year or two, now I just buy good boards and oc the hell out of whatever chip I get. I'm pretty sure that my x3350 @ 3.6 will be better than just about any non-oc'd nehalem a year from now

 

[v8envy]

Yes! Let's have a charity drive to make Hector's parachute truly golden! Reward poor judgment and execution and we're sure to guarantee great results in the future.

That, and 250M wouldn't buy AMD more than an additional quarter or two. Giving them a fish is definitely the wrong answer.

 

[Jax Omen]

So how do we teach AMD to fish?

 

[v8envy]

Not our job. Hopefully the stockholders will wake up and go chumming using some 'key executives' as bait. That should do it.

 

[Winterpool]

We're getting a bit off topic now, but like everyone else on these forums, I don't think the customer (especially the enthusiast) will win in a monopoly market. I want AMD to recover and indeed to kick arse. Yes, we need AMD to learn how to fish (again). Which brings me to my question: why (how) did AMD do so well last time? I think especially of their arc in the first half of this decade, from Athlon XP to Athlon 64, when even if they didn't always beat Intel, they were almost always competitive and better value. I wonder if some geek site (perhaps Ars Technica) could do a good job examining AMD's recent history.

As much as I appreciate the gains of Moore's Law, I'm half-dreading the arrival of Nehalem for this and other reasons. It's the first new architecture that Intel will usher in from a position of relative complacency.