Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

wally21

macrumors member
Original poster
Mar 7, 2009
53
0
Hello all,

I have been following intel's release of the nehalem processors (gainestown for xeon) with a lot of anticipation. I know that the new MPs have received some fairly negative reviews so far, but still remain optimistic that over the next few months with Snow Leopard and other new software coming out, these processors will start to show their true merit.

One thing that keeps bugging me about the current MP analysis is that the current tests and programs seem to rely more on processor speed than multi-threading capability. Given this, I was wondering if anyone knows if/when apple will include the 3.2GHz (w5580) as an option. I really think that comparing an 2008 3.2GHz octo to a 2009 2.93GHz octo isn't quite fair (price concerns aside).
 
This is what I've been wondering as well. I think that the 3.2GHz Nehalem will be silently upgraded and bumping each model down while keeping the prices the same, thus making the 2.66GHz the new base model when the 3.2GHz finally hits.

Also dont forget that Apple received their Nehalem Xeons (gainstown) earlier than any other vendors. Once the full line is out, I'm sure Apple will adjust it accordingly.
 
Never going to happen. The extra cost and power consumption of the 3.2GHz chip made it impossible for Apple to consider given the silly price of the chip and their new green credentials.
 
Never going to happen. The extra cost and power consumption of the 3.2GHz chip made it impossible for Apple to consider given the silly price of the chip and their new green credentials.

Oh, yes, I am going to pull a one-ton load with my Hybrid....

Not.

The MacPro is a machine that gets you get work done with, not worry about its power consumption.
 
Never going to happen. The extra cost and power consumption of the 3.2GHz chip made it impossible for Apple to consider given the silly price of the chip and their new green credentials.


The prices that I've seen have only been about $300 or so higher for the 3.2GHz vs the 2.93GHz. But I suppose if you double that, and add apple's premium prices it might add up to an extra $1000 or so
 
This is what I've been wondering as well. I think that the 3.2GHz Nehalem will be silently upgraded and bumping each model down while keeping the prices the same, thus making the 2.66GHz the new base model when the 3.2GHz finally hits.

Also dont forget that Apple received their Nehalem Xeons (gainstown) earlier than any other vendors. Once the full line is out, I'm sure Apple will adjust it accordingly.

your posts are erroneous and misleading.
 
your posts are erroneous and misleading.

How is it misleading? Its just my opinions. I never said its a fact, I said, "I think."

I wouldnt put that away though because macbook pros have been getting silent bump upgrades all the time.

Its true that the 3.2GHz machine will cost more than what the 2.93GHz does (intel's prices) BUT if you think about it, Apple is already OVERCHARGING all the 8 cores by ALOT more than it should be. If they do a silent upgrade I think it will match the prices accordingly with the speed bumps.

Maybe Apple wanted to keep the prices high because they know of the inevitable silent speed bumps are coming. And especially since they got their Nehalem chips first, they wanted to put it up for sale and make some extra $$$ in the next few months.
 
How is it misleading? Its just my opinions. I never said its a fact, I said, "I think."

I wouldnt put that away though because macbook pros have been getting silent bump upgrades all the time.

Its true that the 3.2GHz machine will cost more than what the 2.93GHz does (intel's prices) BUT if you think about it, Apple is already OVERCHARGING all the 8 cores by ALOT more than it should be. If they do a silent upgrade I think it will match the prices accordingly with the speed bumps.

Maybe Apple wanted to keep the prices high because they know of the inevitable silent speed bumps are coming. And especially since they got their Nehalem chips first, they wanted to put it up for sale and make some extra $$$ in the next few months.

I hope you're right, but as has been noted above, the trouble is you have:


CHIP Power

3.2 130 W
------------------------------
2.93 95 W
2.8
2.66
------------------------------
2.53 80 W
2.4
2.26

So, the logical offers clock-speed wise are 2.4, 2.8, and 3.2, but b/c of the huge power requirement of the 3.2, it is likely to not be featured, hence the only logical offers become 2.26, 2.66, and 2.93. I guess conceivably, Apple could upgrade the lowest one to the 2.4 in a quiet upgrade, but I don't see the 3.2 coming... hope I'm wrong!!!
 
I hope you're right, but as has been noted above, the trouble is you have:


CHIP Power

3.2 130 W
------------------------------
2.93 95 W
2.8
2.66
------------------------------
2.53 80 W
2.4
2.26

So, the logical offers clock-speed wise are 2.4, 2.8, and 3.2, but b/c of the huge power requirement of the 3.2, it is likely to not be featured, hence the only logical offers become 2.26, 2.66, and 2.93. I guess conceivably, Apple could upgrade the lowest one to the 2.4 in a quiet upgrade, but I don't see the 3.2 coming... hope I'm wrong!!!

I dont see why Apple wouldnt give a 3.2GHz a go even at the rated 130W. I mean its not like its 200W or anything insane, if its that fast I think they'll still offer it.
 
So is the world going to end for all you naysayers?

I think it is perfectly reasonable for Apple to upgrade the speed of the whole 8-core line in 6-8 months, with the bottom end being 2.66. Just because Apple doesn't have a history of upgrading often with Mac Pros, (often in Apple terms is 6-8 months, in the rest of the industry is 3-4 months.) In the PowerPC days, Apple would upgrade the PowerMacs as soon as they could get their hands on newer faster chips. I think that they realize that many (not all) people willing drop $2800 on a computer in November and December were also willing to wait for the next upgrade, upon hearing how great it would be.

Most reasonable people would think that it is ridiculous for a company to sell a product for the same price, 14 months after it is introduced. Just because most of us are fanboys here, doesn't make everything Apple does, the right thing to do.
 
Has nothing to do with it. The 3.2GHz Harpertown was the same wattage.
And the 3.0 GHz Clovertown was 150 W.

I think it is perfectly reasonable for Apple to upgrade the speed of the whole 8-core line in 6-8 months, with the bottom end being 2.66. Just because Apple doesn't have a history of upgrading often with Mac Pros, (often in Apple terms is 6-8 months, in the rest of the industry is 3-4 months.)
They skipped the initial set of Clovertowns in late 2006.
They added one Clovertown in 2007.
They skipped the 3.4 GHz Harpertown late last year.
I'm not betting on a silent speed bump, even though it would be good. Possible, yes. Likely, probably not…

Most reasonable people would think that it is ridiculous for a company to sell a product for the same price, 14 months after it is introduced. Just because most of us are fanboys here, doesn't make everything Apple does, the right thing to do.
When predicting release dates, I say what I think Apple might do, whether I agree with the prediction or not.
 
I think that Apple will do an upgrade in fall 2009.
That will give them some room to go to Xeon Westmere around June 2010 (the consummer/laptop Westmere will be late 2009/early 2010), then if all goes well we'll see SandyBridge in June/Fall 2011.
So If we look at the current Xeon Processor available now (march 29),
the next update can be : Quad : 2,93 / 3,20 Octo : 2,40 / 2,80 / 3,20 (for the current buyer those speed doesn't change a thing)
Btw in may 31, Intel will update the Core i7 to 3,06 & 3,33.
 
I too hope they upgrade to the 3.2 at some point.

I laugh every time I see someone say "Google the Megahertz Myth"

Megahertz still has so much to do with the speed of the computer. Why is it then that the 2.93 is so much faster than the 2.26? Why is the 2.66 faster than the 2.26? They all have the same amount of cores? If cores are more important than the speed of the cpu, then why even have different speed chips? So wouldn't the 3.2 be that much faster than the 2.93 and 2.66? Given the poor reviews of the 2.26 due to many things still relying on cpu speed, I agree with those that say they should realign the specs and do the 3.2 as the top and 2.66 as the bottom.

As far as pricing goes, that's another story... Since the Nehelam 2.93 system is roughly $1400 more than the 3.2 system was previously, I don't see how they could raise the price any more than they already have and that might be an issue. If they could adjust the specs and leave the pricing as is, they'd have a winner. If they would have to raise the price on the top end, and since it's already so much higher than the previous gen, I'd say it'd be a tough call...

At this point, so many things rely on cpu speed rather than just how many cores are doing the work, I'm holding off buying a MP untill they either prove something with snow leopard and put the cores to work, or I'll wait till they come out with a faster cpu.
 
I too hope they upgrade to the 3.2 at some point.

I laugh every time I see someone say "Google the Megahertz Myth"

Megahertz still has so much to do with the speed of the computer. Why is it then that the 2.93 is so much faster than the 2.26? Why is the 2.66 faster than the 2.26? They all have the same amount of cores? If cores are more important than the speed of the cpu, then why even have different speed chips? So wouldn't the 3.2 be that much faster than the 2.93 and 2.66? Given the poor reviews of the 2.26 due to many things still relying on cpu speed, I agree with those that say they should realign the specs and do the 3.2 as the top and 2.66 as the bottom.

As far as pricing goes, that's another story... Since the Nehelam 2.93 system is roughly $1400 more than the 3.2 system was previously, I don't see how they could raise the price any more than they already have and that might be an issue. If they could adjust the specs and leave the pricing as is, they'd have a winner. If they would have to raise the price on the top end, and since it's already so much higher than the previous gen, I'd say it'd be a tough call...

At this point, so many things rely on cpu speed rather than just how many cores are doing the work, I'm holding off buying a MP untill they either prove something with snow leopard and put the cores to work, or I'll wait till they come out with a faster cpu.

That's not what the megahertz myth is.
To summarize the argument, it's that you can't draw any conclusions on performance based on the frequency alone.

You have to consider several factors such as instruction set and extensions, pipeline length, etc.

Basically, "how many cycles does it take do do X on processor Y compared to processor Z".
 
Right, but I'm talking about the speed of cpu's within the same architecture, meaning the instruction sets are all the same.. Therefore the higher the clock the better....
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.