Mac Pro Suitable Xeon (Nehalem) Processors Due March 29th

[whatever]

mini Tower....not again!

Apple will not do a mini-tower! They may shrink the size of the Mac Pro, but don't expect a mini tower.

People here really don't get it. The reason Apple was able to to capture market share was not because they lowered their prices and catered to the masses, but because they remain true to their ideals. As companies like Dell cut prices they end up losing more and more profit. This does not help them. Apple is better off selling one high price machine and make a profit than they are selling three cheap machines.

Another thing that needs to be pointed out, Apple is selling computers. On any given Saturday go into an Apple Store and hang out for two hours and count the number of computers leaving the store.

Yes, the economy sucks, but people are still buying....

 

[antominus]

how fast?

With the return of hyper threading these things should scream. It didn't work the first time around because the implementation was weak and parallel computing was a novelty in the high end segment, where people were running custom proprietary software. Now everything is expected to be run in multiple streams, and this chip can work this to the hilt.

By the look of things though, these xeons are going to be EXPENSIVE, and Intel's platforms have increased tremendously as well. The price point for the 2.8 harpertown is the 2.53 gainestown (which will, of course, be faster and have 8 threads per chip, 16 total ). I doubt they will offer the base model with slower ram (2.53 is the top clock on ddr3 1066), so look for a $400 price bump on the base model... OR a single quad core chip that cranks 8 threads sold as "see, still handles 8 ops like your old one", and it probably will be as fast. But... back to a $2499 base price. No deletion options on this one, folks, like the old 2299 quad. Two of em will be a $1000 BTO... $3499 for 8 cores of processing doom. $3999 for 2.8GHz

Citibank or Wachovia for two 3.2's, or maybe a small island. Like Oahu.

That's how I see it.

 

[Umbongo]

With the return of hyper threading these things should scream. It didn't work the first time around because the implementation was weak and parallel computing was a novelty in the high end segment, where people were running custom proprietary software. Now everything is expected to be run in multiple streams, and this chip can work this to the hilt.

By the look of things though, these xeons are going to be EXPENSIVE, and Intel's platforms have increased tremendously as well. The price point for the 2.8 harpertown is the 2.53 gainestown (which will, of course, be faster and have 8 threads per chip, 16 total ). I doubt they will offer the base model with slower ram (2.53 is the top clock on ddr3 1066), so look for a $400 price bump on the base model... OR a single quad core chip that cranks 8 threads sold as "see, still handles 8 ops like your old one", and it probably will be as fast. But... back to a $2499 base price. No deletion options on this one, folks, like the old 2299 quad. Two of em will be a $1000 BTO... $3499 for 8 cores of processing doom. $3999 for 2.8GHz

Citibank or Wachovia for two 3.2's, or maybe a small island. Like Oahu.

That's how I see it.

The only real price differences on components should be the processors, so with two 2.66GHz X5550s costing $300 more than the current ones we will probably see a price of $2,999. It's a good price point for such a system.

 

[RebootD]

Sheesh! Now I know why Tallest Skil requested on the Mac Pro thread that discussions for xMac be held elsewhere. This "mid level desktop" conversation totally hijacked this thread!!!!!!!!! Someone who gives a darn, please create a dedicated thread for it and stop muddying up the Mac Pro discussions.

That's funny because I've been in an iMac news discussion that was hijacked by a few Mac Pro Gainestown fanatics lately so if you stay out of ours we will stay our of yours.

 

[ktlx]

Sheesh! Now I know why Tallest Skil requested on the Mac Pro thread that discussions for xMac be held elsewhere. This "mid level desktop" conversation totally hijacked this thread!!!!!!!!! Someone who gives a darn, please create a dedicated thread for it and stop muddying up the Mac Pro discussions.

Unfortunately that's what MacRumors has degenerated to. iPhone rumors and Mac Pro rumors that completely get swamped by the xMac crowd. It's impossible to discuss any Mac Pro rumors without completely being overwhelmed by people who think Apple are complete idiots for not following business plans posted by people with too much time on their hands.

 

[Tallest Skil]

Unfortunately that's what MacRumors has degenerated to. iPhone rumors and Mac Pro rumors that completely get swamped by the xMac crowd. It's impossible to discuss any Mac Pro rumors without completely being overwhelmed by people who think Apple are complete idiots for not following business plans posted by people with too much time on their hands.

While this doesn't happen in every thread, it does happen too often.

I would also like to personally thank everyone who has read my thread for following my disclaimer.

When J the Ninja (post #11) started talking about the 35xx series and single socket computers, I got a little tingly feeling and added that disclaimer... I'm glad that I did.

 

[SACD02]

Xeons need to support 2x QuickPath Interconnects (one is needed per socket) and Gainstown support slightly faster memory, DDR3 1333 vs DDR3 1066 on i7.

So the ability to run on dual-socket isn't just simple gate being flipped on/off in the silicon... Does this justify the price? That is for consumers to decide

The extra Quickpath connector doesn't boost any performance. It is there so the processors can communicate to each other otherwise you cant configure them in the multi-processor setup. Xeon i7 processors are basically the regular i7 + another Quickpath connector for the DP setup

As for DDR3 1066 vs. 1333, you can go and view all the benchmarks out there. In the triple channel configuration, 1333 only shows its superiority in synthetic benchmarks. In real world situations, it barely improves the performance. Even the 1600 DDR3 memory doesn't add much to the performance. Thanks to Quickpath, the days of ram bottleneck is almost over. As benchmarks prove, even in dual channel configuration, the memory is pretty saturated and the triple channel doesn't add much. Ram speed becomes important when you talk about the ancient FSB system

PS: A lot of desktop motherboards support DDR3 speeds beyond 1333. My Asus P6T-Deluxe supprots 1333 and 1600 DDR3 just fine. Intel chose to use the 1066 as the official number and benchmarks prove that anything beyond triple channel 1066 add little to the performance

 

[diamond.g]

The only real price differences on components should be the processors, so with two 2.66GHz X5550s costing $300 more than the current ones we will probably see a price of $2,999. It's a good price point for such a system.

RAM should be cheaper (compared to FB-DIMM) as AMD would also be able to help with economy of scale. The mainboard may or may not be cheaper. The CPU costs go up though. In the end it is possible to have a Nehalem MP cost around the same as the current unit. It just isn't likely.

 

[iMacmatician]

By the look of things though, these xeons are going to be EXPENSIVE, and Intel's platforms have increased tremendously as well.

I can see Apple not upgrading other components much in order to even out the price increase from the CPUs.

The price point for the 2.8 harpertown is the 2.53 gainestown (which will, of course, be faster and have 8 threads per chip, 16 total ). I doubt they will offer the base model with slower ram (2.53 is the top clock on ddr3 1066), so look for a $400 price bump on the base model...

2.53 GHz is also the top clock for 80 W CPUs, which are the TDP used in the Xserve. If Apple wants to keep 80 W in them, then 2.53 GHz is the highest speed they can use in the Xserve, which is quite a numerical downgrade from the current 3.0 GHz.

And I also doubt they will use anything lower than the fastest RAM. That's what they did with the current Mac Pro (1600 MHz FSB vs. 1333 MHz FSB, even though the latter had more and cheaper CPUs).

 

[Umbongo]

PS: A lot of desktop motherboards support DDR3 speeds beyond 1333. My Asus P6T-Deluxe supprots 1333 and 1600 DDR3 just fine. Intel chose to use the 1066 as the official number and benchmarks prove that anything beyond triple channel 1066 add little to the performance

I wouldn't be suprised to see 1066 or 1333 being offered by Apple, but I suspect they will push for 1333 even if just for the marketing aspect.

 

[BenRoethig]

I've said this a thousand times. Blu-Ray is dead! There are so many cheaper, faster and better ways to store data.

The ease of digital delivery by NetFlix, Apple-TV, X-Box, is driving the nail in Blu-Rays coffin.

Heavy DRM and restrictive licensing is driving the nail into Blu-Ray's coffin. Unfortunately both also exist with digital movie downloads.

 

[BenRoethig]

Apple will not do a mini-tower! They may shrink the size of the Mac Pro, but don't expect a mini tower.

Who ever said anything about a Mini Tower? We just what what we had, a single CPU Mac Pro in the $1299-$1699

People here really don't get it. The reason Apple was able to to capture market share was not because they lowered their prices and catered to the masses, but because they remain true to their ideals.

They may have not lowered their prices, but they only cater to the masses these days. The ideals and the Mac user base got tossed out the windows when they figured out they could hit the motherload selling image products to the ordinary fad-hooked consumers.

As companies like Dell cut prices they end up losing more and more profit. This does not help them. Apple is better off selling one high price machine and make a profit than they are selling three cheap machines.

Apple's accountants would disagree. Most of the profit doesn't come from the highest end models, in fact most Macs sold are 13" Macbooks.

Another thing that needs to be pointed out, Apple is selling computers. On any given Saturday go into an Apple Store and hang out for two hours and count the number of computers leaving the store.

Nobody is disputing that, just the way they're doing it. They could sell those computers without burning the bridges with the established user base. The switchers who are buying them are also traditionally much less likely to buy a second Mac than the veterans Apple seemingly wants to throw away. Return the Low end PowerMac/MacPro and the big screen iBook/Macbook to the fold and everyone is happy.

 

[twoodcc]

well i sure hope that apple updates the mac pro soon with these processors. i hope they don't stop making desktops though. i really don't think they will, but you never know with apple

 

[takao]

Unfortunately that's what MacRumors has degenerated to. iPhone rumors and Mac Pro rumors that completely get swamped by the xMac crowd. It's impossible to discuss any Mac Pro rumors without completely being overwhelmed by people who think Apple are complete idiots for not following business plans posted by people with too much time on their hands.

that's what you get for introducing lots of people to the platform with a more affordable desktop like the mac mini and then killing off/not offering anything in price/performance between it and a 2200 workstation

 

[Chicane-UK]

that's what you get for introducing lots of people to the platform with a more affordable desktop like the mac mini and then killing off/not offering anything in price/performance between it and a 2200 workstation

Precisely! Like plenty of folks here, I started on a Mini and was very happy with it and upgraded to an iMac. I felt given the features and the beautiful design it was worth it. But I feel I want more horsepower and I just don't think the Mac Pro is a sensible buy - it's completely excessive in both expansion capacity and price for me. I could AFFORD a Mac Pro but just think I'd be wiser spending my money elsewhere.

Why I don't understand about the so called 'Apple purists' is that they rail the comparative Apple newbies (like me) for being enthusiastic about a product and wanting something that fits my needs. I certainly don't agree that Apple should endlessly diversify into needless areas and with infinitely customisable systems... BUT.. the midrange of their product lineup flat out doesn't compete right now. I just feel let down by Apple when I'm basically left with no choice but to spend least £1,600 on a basic Mac Pro setup (including the screen I need to buy) when I could buy a Dell or something (or build my own) for considerably less with far more grunt. I appreciate that the Mac Pro has beautiful and durable design which no other system on the market can match but when you're talking that kind of money, and it's coming down to the bottom line, it's just not a good enough excuse.

 

[Macmel]

I'm not sure what you're referring to, but I'm referring to things like this:

http://news.worldofapple.com/archives/2008/04/28/apple-gets-first-dibs-on-intel-cpu/

http://www.tgdaily.com/content/view/37152/135/

Those are not regular processors. Those are custom made for the iMac, the same the made a special one for the MBA. Of course Apple gets those earlier than anybody else: they're the only ones using them. I don't know if any other manufacturer has taken advantage of these processors after they were made for Apple. Regular processors are used by other manufacturers (Dell, Sony, etc) before Apple does, probably because they're much more flexible regarding their configurations.

 

[-hh]

i simply don't get the compromise attitude ... and i'm simply not willing to pay premium for compromising left and right and only get old out-of-date hardware for it while lacking modern must have features

The problem is that we're all in a 'Catch-22': do you want your hardware to be out-of-date (Apple), or do you want your OS to be out-of-date (Windows)?


-hh

 

[-hh]

You seem to have missed a word in Aiden's post. A mini tower would attract additional sales...

Sorry, but I've heard that refrain a hundred times. The problem is that more sales doesn't automatically increase profits, and one can't simply try to ignore the underlying business fundamentals.

The reason why 'more sales' from adding the xMac won't necessarily increase profits is because (in simple terms), you can't absolutely eliminate cannibalization of sister product lines, which means that each one of these other products have fewer units of sales with which to amortize their fixed costs of manufacturing across ... which means that they each all became incrementally less profitable.

Thus, the key question is if the additional sales from a product line addition will be large enough to offset this profits decline.

Absolutely no one has made this case.

My own view of the problem is compelled to observe that since 70% of Apple sales are away from their desktop lines, the financials have a very very tough job of working out. You're free to disagree, but if you choose to do so, please put up some financial analysis to argue your case. I've said many times before that the whole debate with the "xMac Gap" isn't a technical issue, but is simply from the business side.

The xMac supporters (not demanders) do NOT want Apple to sell machines at PC prices. We want a real desktop computer at a fair price. 50% more than a similar HP or Dell seems fair to me. That would put a modern quad core Mac desktop at the $1499 price point.

Yet even by saying "50% more", you're still baselining on the PC market.

Are you seriously calling the PC tower market niche?

Halfway, Yes.

70% of Mac sales are laptops, and over 50% of PC sales are too.

These numbers put "ALL desktops" in the minority, and this is a known marketplace trend.

Within the Mac product line, because of the non-tower form factors of the mini and iMac, the "tower" is already a niche. This hasn't happened yet on the PC side of the fence, mostly because the PC is still treated like a simple commodity, so PC manufacturers aren't generally able to demand any price premium for alterations to form factors (despite many tries).

Thus, PC consumers don't generally buy PC towers because they demand its feature of internal expansion utility - they buy PC towers because they're cheaper.

FWIW, I also recall hearing a statistic recently that claimed that something like 80% of all desktop PCs were being bought by businesses (I'll try to find it). Bottom line here is that businesses aren't as particularly inclined to pull DIY upgrades, because its cheaper to replace the hardware outright than to pay an IT Technician at fully burdened rates of $100/hour (or higher) to perform an incremental upgrade.

I believe an xMac would attract more than enough new sales to warrant having an additional product in the Apple lineup.

I'd love to believe you. Unfortunately, I've been involved in product development and manufacturing start-ups for 20+ years, so I disagree.

We need to go figure out how much all of Apple's NRE (Non-Recurring Engineering) and Fixed Costs would be: the new design, all of the OS & Software retooling & testing, and also all of the Capital investment costs for setting up the new production line for every new component and its assembly, right down to its packaging. These fixed costs need to be added to the variable costs to get the total, and the fewer units you're selling, the larger that its fraction becomes.

Thus, I know exactly what I need to be convinced otherwise: a reasonably detailed 'sharp pencil' financial analysis of the business case.



-hh

 

[Umbongo]

The reason why 'more sales' from adding the xMac won't necessarily increase profits is because (in simple terms), you can't absolutely eliminate cannibalization of sister product lines, which means that each one of these other products have fewer units of sales with which to amortize their fixed costs of manufacturing across ... which means that they each all became incrementally less profitable.

Thus, the key question is if the additional sales from a product line addition will be large enough to offset this profits decline.

I've said many times before that the whole debate with the "xMac Gap" isn't a technical issue, but is simply from the business side.

I agree with you. The argument that they must do an xMac because there is no draw back to having more has become tiresome on these forums. More products, more sales, more market share. Ignoring the fact Apple under Jobs are focused on doing things the way they believe is right even if that means sacrificing the almighty dollar. Companies don't need to "win" at everything to be the best they can be.

 

[takao]

The problem is that we're all in a 'Catch-22': do you want your hardware to be out-of-date (Apple), or do you want your OS to be out-of-date (Windows)?
-hh

software: can be upgraded/updated on both sides
hardware: not so much on the Apple side

Sorry, but I've heard that refrain a hundred times. The problem is that more sales doesn't automatically increase profits, and one can't simply try to ignore the underlying business fundamentals.

The reason why 'more sales' from adding the xMac won't necessarily increase profits is because (in simple terms), you can't absolutely eliminate cannibalization of sister product lines, which means that each one of these other products have fewer units of sales with which to amortize their fixed costs of manufacturing across ... which means that they each all became incrementally less profitable.

Thus, the key question is if the additional sales from a product line addition will be large enough to offset this profits decline.

Absolutely no one has made this case.

and us heard that about 1 gazillion times:
before the mac mini launch
before the ipod mini launch
etc.

then your niche argument: was an acceptable argument 5 years ago but just look at the apple line objective for once:

a super mini desktop using laptop parts: in non apple world that's a niche in sales
an All-In-One PC: in non apple world even more of an niche
an 2300+ dollar workstations: compared to normal desktops: ultra niche
Macbook Air: ultra thin frequent travelling executive laptops: niche
Apple TV: niche

on the non niche side there is
Macbook
Macbook Pro
iPod
iPhone

which ... color me "surprised" are apples top sellers by far
if apple would axed all niche products (being financial and all that) there sure wouldn't be much left of apple's line up
as for people not buying apple desktops: could it perhaps be that apples desktops haven't been updated like forever and are ridiculous bad value at the moment ?

and why is there still no HDMI from the company who called out "the year of HD" and called their computers "a personal media hub"

and why no netbook from apple ? been a raving non-niche success and ridiculously popular and apple does nothing

edit: what i forgot: isn't it better to have mac sales cannibalized by _other_ mac sales than having mac sales cannibalized by non-mac sales ? or do you need a business plan for that ?

edit2: forgot as well: 1 year warranty as standard for a premium brand: should i laugh or should i cry ?

 

[Bubba Satori]

Apple is better off selling one high price machine and make a profit than they are selling three cheap machines.

And so they lose three customers who need reasonably priced, expandable Macs and not $2.3K Mac Pros or under powered, and non-updated Minis who would never consider an expensive MacPro. Brilliant.

 

[BenRoethig]

My own view of the problem is compelled to observe that since 70% of Apple sales are away from their desktop lines

Of course they are. There aren't any desktop lines to choose. That's like limiting your choices to only Fiats and concluding nobody wants a truck. We can't magically conjure up a desktop to buy and those polls aren't exactly asking about user satisfaction with the purchase. Unless you're willing to drop down $2300 on a workstation, you're getting a non-portable notebook that is actually less expandable than the portable version. How can you draw any kind of conclusion when Apple doesn't give you a choice? Its real easy to make the numbers what you want when you;re actively streering them in that direction and you're the only game in town.

 

[mcavjame]

Economy nothing. The people that these computers are designed for will be able to afford them.

That may be a little optimistic. The market has been flooded in my area with recent model MP because design firms are going out of business and/or cutting back. This downturn is a result of businesses trimming ad campaigns in the wake of layoffs and cutbacks. Apple will be cognizant of these realities and will certainly design and market their products carefully.

 

[BenRoethig]

That may be a little optimistic. The market has been flooded in my area with recent model MP because design firms are going out of business and/or cutting back. This downturn is a result of businesses trimming ad campaigns in the wake of layoffs and cutbacks. Apple will be cognizant of these realities and will certainly design and market their products carefully.

A lot of the smaller companies have also had their per computer price almost double from the G4/5 to the MacPro due to it move up market. The dual quad 2.8s and above are fantastic deals for the money, but the same can't be said of the single 2.8. I've heard of quite a few firms that are thinking about moving to PCs because they can no longer afford Mac Pros and the couple test iMac did not meet the requirements.

 

[AidenShaw]

Of course they are. There aren't any desktop lines to choose. That's like limiting your choices to only Fiats and concluding nobody wants a truck.

Thank you for saying this clearly.

All of the statistics about demand for Apples are warped by the limitations of the product line.

Mini-towers are the "4 door sedans" of the computer market. If you only sell Smart Cars and Hummers, though, you might conclude that there's no demand for 4 door sedans.


Fiat does make trucks, however.

http://www.fiatprofessional.it/LCV_...NERIC/1074015791/20080229/Kv_01_fiat_prof.jpg

The dual quad 2.8s and above are fantastic deals for the money...

Compare the $900 Dell Core i7 with the $2800 octo 2.8, and consider that:

• The Core i7 is much faster at anything using 4 cores or fewer
• The Core i7 has better multi-thread SPEC benchmark numbers than the octo-core (they are both 8 thread machines)
• The Dell has 50% more memory than the Apple, and 50% more memory channels
• Memory upgrade to 8 GiB is $250 on Dell, $1500 on Apple
• The Dell has a better graphics card
• The Dell has a larger hard drive
• The Dell has hardware RAID standard ($800 extra on Apple, although the Apple has more capable RAID with that card)
• The Dell has a 360 watt power supply, the Apple has over 1000 watt supply
• The Apple supports 32 GiB of ECC memory, vs 24 GiB non-ECC for the Dell
• The Apple supports 4 3.5" disks plus 2 optical, the Dell supports 3 disks plus 2 optical, or 4 disks plus 1 optical
• The Apple has 3 open PCIe slots - one x16, two x4. The Dell has 3 open PCIe x1 slots. The Apple has more power available for PCIe cards.

I find it hard to claim that the $2800 Apple is a fantastic deal.

It's only reasonable when compared to other dual socket workstations, or if you require more than 24 GiB RAM or ECC or more PCIe bandwidth/power.

In most other ways it's less capable than a single Core i7 that is less than a third the price. Of course, once Nehalem is in the Mac Pro the performance should be quite a bit better than the Dell on those "more than 4 thread" applications/workflows. But when that happens, it's possible that the Dell will be a quarter of the price of the Apple.