Pounce or wait for Nehalem?

There are much more important things to talk about like for instance this whole Russian-Georgian invasion.

What does the Russian-Georgian invasion have to do with the Nehalem? Of course there are more important things going on in the world, but does that mean we should discuss them in a hardware forum? Why do people always say things like this? I see this on digg.com a lot; "oh no, I dont like someone's point of view! We're in a WAR... um.. in IRAQ.... over.... OIL! Stop keeping the conversation on topic, GAS PRICES!"

ugh
 
What does the Russian-Georgian invasion have to do with the Nehalem? Of course there are more important things going on in the world, but does that mean we should discuss them in a hardware forum? Why do people always say things like this? I see this on digg.com a lot; "oh no, I dont like someone's point of view! We're in a WAR... um.. in IRAQ.... over.... OIL! Stop keeping the conversation on topic, GAS PRICES!"

ugh
Ok fine. Lets continue to argue about the fact that Nehalem is coming soon and some of us will have bigger e-penis's than others. Awesome!
 
because talking about politics was the ONLY other option?
At least it is something important that really does have some sort of effect on your life. I'm just saying. Besides the OP has heard lots of opinions and should now be able to make their mind up at this point.
 
Wow just stop. You are just being contrary just for the sake of being contrary now. You are speculating WAY too much. I am going by the information given. The LGA 775 stuff will drop extremely low once the Nehalem hits. The best information we have if that the Nehalem will be out at the end of September possibly early October and there will be a $300-350 2.66Ghz part, $500-550 2.93Ghz part, and an extreme $999 3.2Ghz part. There is also evidence that the CPU's will perform anywhere from 33-80% faster at the same clocks as current quads depending on the app. That is about all we know about these CPU's. Intel is obviously close to release or else they wouldn't be talking about the new i7 moniker.

Yes a P4 to a Q6600 is a great upgrade. I never said it wasn't. Noone did. I simply said that if they were able to make it this far without upgrading they can stand to wait a month or so longer. The Nehalem will probably release on time. There is no reason I have seen to think otherwise unless you can provide me one. Like the guy said earlier. An educated decision would be to wait a little bit and size up the Nehalem. If you don't need it then buy something cheaper.

Buyers remorse will likely be an issue for some buying any current gen system. I don't see how you can claim that someone that bought an older gen system so close to a next gen release could still be happy about their purchase when they could have waited just a month or so more and have a system that is up to 80% faster for a couple hundred dollars more (perhaps). Hell if you spent money on a high end system now you could have easily bought a mid-range Nehalem machine in a month.

To take some kind of magical poll on some forum (even if it is this one) seriously is just damn stupid. Do you honestly think you know more than the news sites who hear this from REAL insiders? No don't answer that one cause you probably think you are an insider.

exactly. if you only upgrade your pc every 3 tears thenw ait for nehalem. if you upgrade yearly then fair enough go for a core 2. some of us dont have the privilege to upgrade every year. there is a reason why me and the OP STILL have a P4 lads.

might be easy for some of u to upgrade yearly but not me and most probably the OP. i can wait for nehalem as its just months away from being released
 
Look man we have different opinions about this. I honestly see no reason to argue about this anymore. There are much more important things to talk about like for instance this whole Russian-Georgian invasion.

LOL now i have seen it ALL. talk about going WAY OFF TOPIC.
 
exactly. if you only upgrade your pc every 3 tears thenw ait for nehalem. if you upgrade yearly then fair enough go for a core 2. some of us dont have the privilege to upgrade every year. there is a reason why me and the OP STILL have a P4 lads.

might be easy for some of u to upgrade yearly but not me and most probably the OP. i can wait for nehalem as its just months away from being released
I'm in a similar financial position, and my viewpoint is that if you can't afford to upgrade more often than once every 3 years, you probably don't have the cash to plop down for "bleeding edge" Nehalem until at least several months after its release. Therefore, the effective waiting time for the new platform (assuming you aren't willing to jump on it until prices come down to a more reasonable level) is more than just a month or two.

Being on a budget, I feel like I did the best thing I could by putting together an E8400-based system this week. Prices for C2D/C2Q (excepting the highest models) are probably near the lowest they're going to get before they're phased out, and DDR2 is unbelievably cheap. Buying all new from NewEgg, I spent less than $350 for my CPU+mobo+RAM - and this is still top-of-the-line technology right now! A great time for budget buyers indeed. Sure, Bloomfield 2.66GHz may be out in a few months, but the chip alone will cost almost that much, to say nothing of DDR3 prices.

I suffered with my P4 system long enough, and while I haven't even fired up any games yet, just the difference in system responsiveness and application load times in XP is night and day. I fully expect I'll get by just fine on this build for at least 2 years, even if Bloomfield does post much better benchmarks. Then, sometime in 2010 when Nehalem is priced similarly to the current C2 lineup and DDR3 is completely mainstream, I'll probably swap my current setup for a mid-range Neha on the cheap.

I fully understand why extreme performance enthusiasts and those with money to spare are clammering to pick up Nehalem as soon as it drops, but for budget users still running a P4, I fail to see why it makes sense to stay on the waiting train. Seems to me that we're currently seeing about the best price/performance ratio in the low-to-mid-range market, relative to what's available at the time, that we will until at least 6-9 months after the new chips hit. (Because initially, all Nehalem chips will be out of reach for budget users, forcing them to either wait months longer or settle for the same C2D/C2Q chips they could have had months earlier, and likely for not much cheaper than they are now).
 
I am the type that tends to stick with tried and true tech.... jumping in both feet into a brand new socket and CPU and boards with new chipsets is a bit more thatn I like... I usually wait about 6 months for the bugs to get worked out.. and it has worked perfectly for me... I started with a E6600 on a P965 board then went to a P35 board and have never been happier with my decisions...
 
I currently have an E6400 (3.2ghz) + 4850 + GA-965P-S3.

I was very excited about a possible Nehalem build, until I took out my calculator and started counting the hard numbers.

In comparison to a similar Kentsfield or Yorkfield build (Quad core), it would cost me about $500 extra to go with a similar Nehalem build with the 2.66ghz quad because of the extra cost of ram and the cpu itself.

Now, is the extra $500 worth it for roughly the same performance granted that current quads will most likely overclock better? No.

Why spend more money for diminishing returns?

I strongly believe that the best choice for performance would be to go with a Penryn build now, then upgrade for a minimal cost to Nehalem/Westmere late next year. Plus, if you wait till the middle of next year, you will get the additional bonus of having USB 3.0.;)

So, upgrading to Nehalem within 3 months of its release is simply not worth it.
 
I currently have an E6400 (3.2ghz) + 4850 + GA-965P-S3.


Why spend more money for diminishing returns?

I strongly believe that the best choice for performance would be to go with a Penryn build now, then upgrade for a minimal cost to Nehalem/Westmere late next year. .

thats the catch. i cant build a new pc every year mate so the next one i build will last 3-4 years and a core 2 will feel REALY outdated within months whilst i feel my nehalem system will last 2 years+ before it REALY gets outdated.

if i wasnt a married man and wasnt saving up for a mortgage for next year, i would definetly build a new rig now. it realy depends on your situations but for the next few years unless i get a major payrise yearly from my job, i wont have a chance to build even a entry level nehalem next year.

the credit crunch is fucking us all up mate, fuel is going high plus bills and foods.
 
thats the catch. i cant build a new pc every year mate so the next one i build will last 3-4 years and a core 2 will feel REALY outdated within months whilst i feel my nehalem system will last 2 years+ before it REALY gets outdated.

if i wasnt a married man and wasnt saving up for a mortgage for next year, i would definetly build a new rig now. it realy depends on your situations but for the next few years unless i get a major payrise yearly from my job, i wont have a chance to build even a entry level nehalem next year.

the credit crunch is fucking us all up mate, fuel is going high plus bills and foods.

As a gamer, i dont really need Nehalem. The games coming out now and in the future should run fine when i get a Core 2 Quad. The only people who really need it are benchmarkers and what not, or people who need to have the best. But i may build a Nehalem build in two or so years when in college.
 
Yep, there is a catch with every decision.

If you do decide to go with Penryn for $500 cheaper and the same performance, then you will loose memory expansion (up to 24gb on Nehalem) and multi-threading technology in the future.
 
What about people with rather up to date systems right now? I run on a e8400, x38 mobo, and am wondering what i should do. I mostly game but I am also getting to the point in school where I can really benefit from a quad finally. I am in school for Digital Media Design in the game design major and while I dont mind using the macs at school I would like to do the work at home as well as I can make the learning process better, efficient, and get more out of it all.

So should I go for a york, OC it, and stick with my platform until 2010 and the 8-core nehalems come out or go with a nehalem at the end of 2008? I still want to game my butt off but I also want to do work that will benefit from the additional cores. However I usually upgrade every year and half to 2 years depending on what is happening.

I also dont want to get stuck in a chipset that is short lived and just not worth it. I am more into the efficiency of the upgrade/rebuild and not the shock and awe. So this is more from a "Will build once in the next two years for nehalem, should i wait or go for it now or in 2010?" kind of thing.
 
All I can say to that is that there are going to be quite alot of people with buyers remorse if they buy ANYTHING LGA 775 right now. Nehalem is gonna knock their socks off. Right now it would be a HUGE mistake to buy a LGA 775. I mean if you want to guarantee yourself no upgrade path whatsoever then go ahead and buy LGA 775. For me the choice is simple.

Don't fall in love with the numbers, man.

As it is, very little software I can think of realistically challenges 775 quads. That isn't going to change much between now and Christmas.

Meantime, you early-adopters are going to do what you always do and pay intel back some RnD costs. And that's fine. Some of us, on the other hand, are going to assess the needs of the moment vs the needs of the future, compare that to cost, and make a decision.

For my money, a q6600 now with a planned upgrade either next summer or next X-mas looks like a very smart play.

You should calm down a bit.

~S
 
As it is, very little software I can think of realistically challenges 775 quads. That isn't going to change much between now and Christmas.

Lots of applications can challenge 775 quads now, and will basically gobble up any processor power you give them. Video editing/encoding, 3D rendering, parallel code compilation, and scientific simulations are some applications, and that's just what I can think of in 30 seconds.

Now, if you meant to say that relatively few consumers actually use the applications that can use all that processing power, I would agree with you. For games specifically, a 3.0 GHz Core 2 Duo is all you really need at this point for 99% of them out there. There are some people for whom quad and Nehalem can be very useful for, though.

Meantime, you early-adopters are going to do what you always do and pay intel back some RnD costs. And that's fine. Some of us, on the other hand, are going to assess the needs of the moment vs the needs of the future, compare that to cost, and make a decision.

For my money, a q6600 now with a planned upgrade either next summer or next X-mas looks like a very smart play.

If it works for you, great! That's not necessarily the optimal path for others though.

I ran into that very issue with my first build a few years ago. I bought into socket 939 shortly before Core 2 was announced. And I was a bit frustrated that I didn't wait just a bit longer, though being my first real gaming computer in a while, perhaps it is just as well (otherwise I would have been stuck with a laptop with integrated graphics for another six to nine months or so).

If I had a half-ways decent but outdated computer today (i.e.: without integrated graphics, but perhaps a P4 or 939) and was overdue for an upgrade, I would without a doubt wait until February or so to build with Nehalem, after the initial early-adopter rush. Since I already have Conroe, I'm not in as much of a rush.

So what's the smart thing to do if you can wait long enough? IMHO, it's to buy mid-cycle. I was trying to hold out for Penryn, but got impatient and got Conroe right after the FSB raise to 1333 MHz. Turned out to be just as well, as by the time Yorkfield got easily available and affordable, Nehalem was just around the corner.

With the next generation, I plan to wait until around the time Lynnfield is out. If it performs well for a low-enough cost, I'll get it; if not, I'll wait for Westmere or just buy Bloomfield if it's cheap by then. If Nehalem is disappointing after all, I can get a Q9650 for cheap by that point (I might even end up getting a Q9450 anyway if there are any fire-sales of them). Lots more options to pick-and-choose from mid-cycle than the early adopters get, but nearly as cheap as waiting until the end of the cycle and you're not stuck with the old generation for as long.

That's my two cents on that matter, take them or leave them.
 
Upgrading my aging P4 system myself.

The performance gains being quoted for nehalem are unreliable. You can't use a single "per clock" speed when the difference is a much larger number of cores. Actual performance gain will depend on usage and the software taking advantage of cores which will take quite a while. How many games actually scale perfectly over 4 cores anyway?

In any case my logic is that if you buy a first generation nehalem chip it's still going to be obsoleted in a year or two by much cheaper and far superior iterations. So buying a cheap / mature C2D now and then buying a mature nehalem in 1-3 years is quite likely to be an economical solution if that's what you want.

Of course if you run perfectly scalable applications (like benchmarks) and can't stand not having the newest tech / highest numbers then you'll be waiting and paying a premium.
 
Upgrading my aging P4 system myself.

The performance gains being quoted for nehalem are unreliable. You can't use a single "per clock" speed when the difference is a much larger number of cores. Actual performance gain will depend on usage and the software taking advantage of cores which will take quite a while. How many games actually scale perfectly over 4 cores anyway?

In any case my logic is that if you buy a first generation nehalem chip it's still going to be obsoleted in a year or two by much cheaper and far superior iterations. So buying a cheap / mature C2D now and then buying a mature nehalem in 1-3 years is quite likely to be an economical solution if that's what you want.

Of course if you run perfectly scalable applications (like benchmarks) and can't stand not having the newest tech / highest numbers then you'll be waiting and paying a premium.

never buy stuff for todays software/games. buy stuff for tomorrows software/games ;) games in the future will make use of 4 cores and when that happens you would feel appreciated that you waited a while in getting a nehalem hehe
 
Lots of applications can challenge 775 quads now, and will basically gobble up any processor power you give them. Video editing/encoding, 3D rendering, parallel code compilation, and scientific simulations are some applications, and that's just what I can think of in 30 seconds.

Now, if you meant to say that relatively few consumers actually use the applications that can use all that processing power, I would agree with you. For games specifically, a 3.0 GHz Core 2 Duo is all you really need at this point for 99% of them out there. There are some people for whom quad and Nehalem can be very useful for, though.



If it works for you, great! That's not necessarily the optimal path for others though.

I ran into that very issue with my first build a few years ago. I bought into socket 939 shortly before Core 2 was announced. And I was a bit frustrated that I didn't wait just a bit longer, though being my first real gaming computer in a while, perhaps it is just as well (otherwise I would have been stuck with a laptop with integrated graphics for another six to nine months or so).

If I had a half-ways decent but outdated computer today (i.e.: without integrated graphics, but perhaps a P4 or 939) and was overdue for an upgrade, I would without a doubt wait until February or so to build with Nehalem, after the initial early-adopter rush. Since I already have Conroe, I'm not in as much of a rush.

So what's the smart thing to do if you can wait long enough? IMHO, it's to buy mid-cycle. I was trying to hold out for Penryn, but got impatient and got Conroe right after the FSB raise to 1333 MHz. Turned out to be just as well, as by the time Yorkfield got easily available and affordable, Nehalem was just around the corner.

With the next generation, I plan to wait until around the time Lynnfield is out. If it performs well for a low-enough cost, I'll get it; if not, I'll wait for Westmere or just buy Bloomfield if it's cheap by then. If Nehalem is disappointing after all, I can get a Q9650 for cheap by that point (I might even end up getting a Q9450 anyway if there are any fire-sales of them). Lots more options to pick-and-choose from mid-cycle than the early adopters get, but nearly as cheap as waiting until the end of the cycle and you're not stuck with the old generation for as long.

That's my two cents on that matter, take them or leave them.
I've been one of those early adopters for my two major builds so far, the first being Winchester/nForce4 and the other being Conroe/P965. I can say that I've learned my lesson and will always wait for mid-cycle before doing any major overhauls.
 
I've been one of those early adopters for my two major builds so far, the first being Winchester/nForce4 and the other being Conroe/P965. I can say that I've learned my lesson and will always wait for mid-cycle before doing any major overhauls.

why is that?
 
Why not wait, and buy a current-gen LGA775 processor + mobo for cheap when Nehalem comes out? There'll probably be lots of enthusiasts selling high-end parts (Core 2 Extremes, Quads, etc.) for cheap when they upgrade.

Just a though.
 
why is that?
If you wait and let the dust settle, you'll be able to get much more mature hardware. This is especially true for motherboards, where there are often different revisions with various fixes for problems in the early boards.
 
If you wait and let the dust settle, you'll be able to get much more mature hardware. This is especially true for motherboards, where there are often different revisions with various fixes for problems in the early boards.

They've made up their minds. We've covered all this with no effect.
 
This dude has waited for the latest technology.. and waited for prices... then waited for new technology long enough. His AMD 2800+ with 2GB RAM is coming to an end. I am ordering my new parts sometime within 2 months after Nehalem is released with a Q9550 and socket 775. I'll skip Nehelm and get the technology that improves on it within 3-4 years. ;)
 
Back
Top