Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Originally posted by saint.duo
Looks like a $999 eMac to me. Unless you're in to games, the GeForce4 64MB DDR won't mean anything to you, and if you are that big into games, you can build that system for less.
Originally posted by saint.duo
Looks like a $999 eMac to me.

Yes, if it weren't for the fact that the eMac's processor is half as fast as the Dell system that was posted, the two systems would be very comparable (besides the fact that the OS and software on the Mac is infinitely better, of course!). Actually, the eMac is better in many ways from a hardware perspective because for the price of the Dell plus a 17" CRT (i.e. $1250+), it includes things like the Superdrive, a larger HD, Firewire, and built in antennas for Airport. However, there is, once again, the problem that the processor is much slower (in real terms, not just in terms of clock speed).

If you put a 1.6 Ghz G4 with DDR333 (and a processor FSB that can actually run double pumped), then I would say that the eMac would be very competitive with the Dell from a hardware perspective (and blow it away from a software perspective). But of course the problem is that Motorola isn't producing a G4 with a double pumped FSB, and their highest clocked G4s are sensibly reserved for the Pro towers. Hopefully the arrival of the PPC 970 will change this. In the best case scenario, we would actually see a new "G4" type chip (possibly by IBM) that would support the higher frequencies and a double pumped FSB, and this would be used in the eMac and/or iMac lineup. But even in the suboptimal scenario (i.e. that we must continue to use Motorola G4s because IBM has no replacement chip to bridge the gap between the G3 and the 970), by going with the 970 in PowerMacs, Apple will hopefully be able to put higher clocked G4s with faster bus speeds (but unfortunately still not DDR) into the consumer Macs. That will at least make them remotely competitive (in terms of processor speed...particularly when the MPC 7457 is released later this year) with the $1,000 PCs, even if it not entirely closing the performance gap.
 
Like the auto industry?

The comments about market share vs. margin, and Apple positioning itself as a higher value (luxury?) item makes me think of the auto industry as an interesting analgy. According to the recent Harvard Business Review, IT infrastructure has reached its peak and is no longer an issue of strategic advantage for corporations. It is now moving toward a cost-driven necessity -- similar to the rapid adoption of electricity, telephones, trains, or automobiles.

Looking at automobiles, one may equate Dell with Ford (low cost utility), or Microsoft with GM (multi-class products). I would think of Apple as a Mercedies or BMW -- inovation/quality/style driven products that maintain significant premiums and maintain value. These cars, for example do not generally attempt to increase market share by lowering price at the expense of inovation/quality/style. Instead they focus on a premium consumer experience and an association with higher-quality and status.

Personally, I wonder if Apple is indeed focused on maintaining the education markets (even though it is a historical market) and instead is focused on premium product differentiation in an increasingly competitive market place.

If this is the case, then its "consumer" products would be geared toward inovation/style/quality and ease of use at a premium.

Since IT infrastructure is now moving to a cost-driven model, it would also make sense for Apple to focus its business markets in areas were there is clear competative opportunity (i.e. multi-media). Thus, the drive toward performance seemes to be focused on satisfying the media professional needs -- where it can possibly be a market leader.

Price points for the "consumer" markets would remain as high as possible to maintain margin. If this were the case, the idea of low-cost machines would be counter-intuitive. Inovation would be geard toward premium consumer experience (e.g., ipod, powerbook, imac) and the needs of the multi-media industry (e.g., 64 bit processing, HD screens, xserve, xraid), with some cross-over (e.g., 803.11g for business and home device networking, professional multimedia apps and iapps).

I wouldn't expect apple to develop competative "productivity" apps, since they fit into the low-cost/broad-market utility model - unless there was a niche opportunity that took advantage of unique apple technology (e.g., mulit-media apps such as safari, keynote, quicktime, itunes). My point is, I don't expect Apple to try to capture large shares of business or consumer markets. They will focus on higher-value premium markets.

My thoughts, anyway.
 
This time IBM will also be using this 970 thingermebob, I bet it will only be a matter of time before IBM brings out its own OS X machine (probably part of the deal, why else would they have shoehorned in the altivec unit? IBM doesn't need it in it's blade servers).

This would have the automatic effect of immediately increasing Apple's share of the market, people would then be using OS X at work and then be more inclined to buy it for home as well. The knock on effect.

If I'm not mistaken Apple has bought a ton of software firms recently, Shake and Logic for example, I bet these bits of software suddenly appear in 64 bit versions very soon, if not on the day of the launch of the 970. Its certainly an original way of garanteeing that the apps get written on time.

Another thought, maybe the iTunes song shop is just a marketing gimmick and really there just to raise the profile of Apple, after all when it reaches the other 97% it will be an "in your face" Apple app. How many TV ads and banners is that worth, particularly if its making you a modest amount of money at the same time!

The Ars Technica article was a good read and did point out that Apple has obviously been working flat out on something big for the past 2 years, I wouldn't be suprised if everything is related in some big way!
 
Originally posted by benoda
If the 970 does come to Apple, how is Apple going to deal with creating an optimized version of OS X along with pushing software creators to optimize for 64-bits etc. while still having G4's and G3's in the lower end? Seems like a bad idea to say "program XYZ is made for a powermac 970, but you other mac users are SOL."

Quite simple. You (the developer) compile twice: once with G3/G4 optimizations and once with 970+ optimizations. Or, you compile just once with 970+ optimizations and tell your customers that for better performance they can buy newer hardware ... not great CR, but better than being slower than the competition on high-end hardware!

The installer will install the correct version of shared libraries and your app, or your app will link to the correct version of shared library after an initial gestalt.

Note that these are optimizations only, meaning that the code itself will still run on "legacy" chips, albeit slower than code optimized for those chips. You still have to appropriately handle unsupported instructions (like Altivec on a G3 or 64-bit operations on a G3/G4) in conditional code. But that's nothing new for most developers (and, truthfully, most of it is handled via standard system libraries anyways ...)
 
Re: Apple employee discount

Originally posted by nicmac
Wasn't there information recently that apple was offering Power Macs to their employees for a large discount.
Or do they always offer large discounts?
Or was that just a rumor?

Apple always offers large discounts to employees, and the specific percentages in the rumour (30% off) are incorrect according to employees I know.

Rumour completely ignorable.
 
I can't quite figure out what the problem is that some of you have with apps running faster on the 970 if they are optimized for it vs. a G4 or G3.

Apps work better on my G4 than my G3 now just as one would expect.

Apps will still run on the "older" machines when the 970 comes out. Plus it sounds like the 970 will be faster than the G4 and certainly the G3 so apps will run faster on the 970 even if the apps are not optimized for it.

What's the problem?
 
Originally posted by anjaki
This time IBM will also be using this 970 thingermebob, I bet it will only be a matter of time before IBM brings out its own OS X machine (probably part of the deal, why else would they have shoehorned in the altivec unit? IBM doesn't need it in it's blade servers).

IBM added the AltiVec unit so that the PowerPC 970 would have one more market where it is attractive. Although IBM has said they are going to use it themselves in blade servers (and probably workstations), as others have said, IBM has not penetrated that market well yet.

There is a lot of risk in building the PowerPC 970 just for the blade server market and possibly the embedded market in the future. By adding the AltiVec unit, they almost certainly guaranteed that Apple would be a customer which would mitigate the risk.

The embedded market still likes the 750FXes a lot and although the 970 looks really cool, it is hard to see them jumping at it until at least it goes to the 90 nm process. Even then, if the later versions of the 750 really can go to 2Ghz, they still might not find the 970 attractive.

The blade server market is attractive but how many blade server customers are willing to choose a PowerPC blade over an Intel blade? Theoretically, if you are running Linux, it should not really matter. But many companies that run UNIX/Linux also run Windows NT/2000/2003 Server and may decide to standardize on Intel blades to reduce total costs, spare parts and maintenance contracts.

I don't believe you can make any assumptions about IBM bringing out a Mac OS X based machine themselves. In fact, that seems counter to their business strategy. I firmly believe the reason for AltiVec was to mitigate the business risk and improve the case for developing the PowerPC 970. Without the unit, the justification for even doing the PowerPC 970 at all is pretty weak.
 
I got this email fromIBM Microelectronics today (I added to the listserve) so enjoy if anything is new:

Standard products



64-Bit PowerPC 970 Targets Entry-Level Servers and Desktops

This Microprocessor Report excerpt, by the Reed Electronic Group, reviews the PowerPC 970: "IBM has created an impressive and affordable PowerPC chip for smaller servers, graphics workstations and desktop computers."

This report will be available for a limited time at:_http://www-3.ibm.com/chips/techlib/techlib.nsf/techdocs/A2CE393ABF2CE99787256D21006AE8A2
 
Apple = BMW?

You know, I have to admit I've thought of it this way myself.

However, a car is a self-contained object. You buy it, and all you have to do is maintain it with factory parts and put tires on it and gas and other fluids in it. Nobody has to design a road that works specifically for your make of vehicle so you have something to drive on.

This is the crucial difference. You see, it takes just as much manpower to develop a program for the PC market as it does for the Mac market. Except you have a ton more users on the PC side (NOTE: You have to eliminate a good percentage of this on big soulless corporate LANs). Although most of the design considerations for an application will be universally applicable, you still have to write good code and localize it to each OS. So having a small market share does hurt.

It also hurts to have to charge more for the hardware you sell to cover R&D costs because you have less of a user base to recoup those costs with. And this is two-edged too, because they have to pay more of a premium for specialized parts they get from 3rd parties the less of them they buy for inclusion in their computers.

I live in an area where there's a lot of Lowrider Customization, Eh, so I've seen a lot of customization parts around. You know what the easiest thing to find parts for is? Hondas. You know why? Because any given light structure can fit into like four years worth of Hondas, and there are a TON on them on the road, many of the drivers of which like to customize. But go look for a '92 Lincoln Mark VII sealed light enclosure. I bet you can't find one.

Anyway, I see your point, and I agree that Apple should not sacrifice quality for market share, but they do need to tighten their belts and sacrifice profit for market share (once they have a marketable product again), because this will only be a self-perpetuating process. The more market share you have the less margins you need, the less margins you need, the cheaper you can sell the same quality product, the cheaper it is the more people buy it = the more marketshare you have.

I'm done with my ranting Swede impression now ;) "That's ranty!"
 
Originally posted by ktlx
IBM added the AltiVec unit so that the PowerPC 970 would have one more market where it is attractive. Although IBM has said they are going to use it themselves in blade servers (and probably workstations), as others have said, IBM has not penetrated that market well yet.

There is a lot of risk in building the PowerPC 970 just for the blade server market and possibly the embedded market in the future. By adding the AltiVec unit, they almost certainly guaranteed that Apple would be a customer which would mitigate the risk.
I think that the PowerPC 970 could be part of IBMs strategy to kill the Itanium architecture. They try to bash it to death in the high-end server market with the POWER4+ and the upcoming POWER5.
The 970 could have been ironed to nip in the bud the Itanium workstation market with Apples help.
PowerPC64 proliferation is a strategic move, if Linux and Mac OS X could gain momentun as 64 bits platforms before IA-64 it would be a victory for IBM.
 
Theory, and practice

Originally posted by ktlx
The blade server market is attractive but how many blade server customers are willing to choose a PowerPC blade over an Intel blade? Theoretically, if you are running Linux, it should not really matter.

If you are only running the base Linux operating system components, you are right that it shouldn't matter. Get a PPC version of Linux, and go.

If you have additional programs that you've written (or open source programs that you've downloaded), and a team of programmers to port them to PPC, then test and debug them, it's a little more work and expense. Usually not much work, but sometimes significant work.

If you need to run third party software, you'll need a PPC version of the product. Not much is available - for example, try to find any Oracle products for PPC Linux.... Try to download the Sun Java for PPC Linux....
 
apple and bmw

"This is the crucial difference. You see, it takes just as much manpower to develop a program for the PC market as it does for the Mac market. "

I agree that is a crucial difference, but I wonder if Apple's move to adopt standardization and open source is an attempt to mitigate that cost (e.g., Unix base, rendezvous, safari).
 
That's true, but Unix portability doesn't help as much as Windows portability would. But it is cool (if you use Unix stuff) and the open source movement is one of my greatest sources of pride in Apple right now. Not to mention what a stark contrast it is with Microsoft.

Good point.
 
Re: Apple = BMW?

Originally posted by BaghdadBob
This is the crucial difference. You see, it takes just as much manpower to develop a program for the PC market as it does for the Mac market. Except you have a ton more users on the PC side (NOTE: You have to eliminate a good percentage of this on big soulless corporate LANs). Although most of the design considerations for an application will be universally applicable, you still have to write good code and localize it to each OS. So having a small market share does hurt.

Actually, other than thick-UI apps (which is generally a very poor way to program in any case, even without cross-platform issues!), it is actually fairly simple to use about 90% of your code base across platforms, and the remaining 10% (UI) should really be the easiest part of your app to write anyways.

OS X is largely POSIX-compliant. That means i can use things like BSD sockets and pthreads and stat and signal and mutex. Windows emulates most POSIX code. Know how difficult it is to take a Windows socket app run on OS X? Change the winsock header file to socket.h. Done. Know how difficult it is to change threading from Windows to OS X? Change the function call you use to start the new thread. Know how difficult it is to change file manipulation between Windows and OS X? Other than byte-order issues if you have them, there are really no changes necessary at all.

I work on cross-platform applications. Supporting OS X and Linux and Solaris in addition to Windows adds about 1% to my development schedule, and nearly all of that is solely in making sure a particular change didn't have an unintended effect in the other platforms (ie, regression testing). Yes, my apps are quite UI-light, but unless you're doing really complicated and unique things in your UI (meaning, you have to code and support your own widgets), the UI is the easiest bit to put together (certainly in OSX; in Windows the tools are more curmudgeonly, and I haven't even tried Linux or Solaris UI tools so I won't comment there).

Adding OS X support to any existing product should not add more than 5-10% cost on the first version, and 2-5% cost in additional releases. Of course, if you're not going to get 5-10% more sales the profitability doesn't add up.

But, still. It is far from twice as costly to support Win and Mac as it is to support just Win or Mac. Granted, there will always be those who can not afford the costs and for them Windows-only is really the most obvious target market. But for the rest ...
 
OK, but what scale software does this apply to? For apps that require a high amount of optimization, don't you have considerable porting costs? Look, at this point I will readily admit to being over my head -- I'm a Photoshopper at my best -- so I will pose this to you as a question: does it not take a considerable (more than 1-10%) extra time to port your product when it needs to be well optimized, such as graphics/rendering/video/games/etc.?

I can see for smaller apps or widgets how porting, especially in the new OS X environment, isn't as difficult as it used to be. But at that higher level, aren't the expenses considerably greater?

In any case, as was my original point, the smaller the market share the less motivation there is to do this -- unless that 3% is ALL professionals who each will need a license for that software. But I guess that will be relative to how exactly difficult that is to do, eh?
 
Look people, if all you care about is comparing hardware specs, you should get a PC. It's as simple as that. For my part, I choose Macs because of:

1. innovation
2. industrial design
3. the OS
4. basic apps like iSync, iTunes, iMovie, iDVD, etc.
5. integration of hardware, software, & services

For all this, the mark-up in price is worth it to me. You can't expect Macs to be price-competitive with Dell for the simple fact that Apple depends so much on R&D and invests heavily in it. It's one company doing all the R&D for the software, industrial design, motherboard, services, etc.

My main complaint with Apple is that Macs have stalled in performace, relative to the demands of software, over the last few years. I'm willing to take out loans or whatever (I'm a grad student now) to get that $3000 system from Apple. But when I spend that much, I want to get top-of-the-line performance. They're not delivering that now. With the 970, hopefully, Apple will get that performace edge back.

And no, Apple does not need to compete in the sub-$1000 tower market to be a viable platform. Not to be an elitist or anything, but Apple simply cannot survive as an innovator competing in that razor-thin-margin market. Apple is doing pretty well with it's 3% market share; I would be super happy with 6% market share and class-leading performance.

In 2000, Apple sold 1,436,000 Power Macs, according to their K-10 filing. That number went down to 766,000 in 2002. There is a lot of pent-up demand right now Power Macs. If they introduce the 970 at up to dual 1.8 ghz, that unit sales figure could easily double. That means another $1.5 bil in sales and a percentage point or two in market share. And the introduction of the 970 (and G4 7457) should have coat tail effects throughout Apple's lineup.
 
dongmin, while I agree with you that all computers shouldn't cost the same (do all hifis? TVs? ovens? cars?) Apple could really do with a sub $900 tower and drop the need for ADC and custom ROMs in gfx cards.

If people could buy a no frills PowerMac, I think more would. 1 AGP, 2 PCI, 1 FW, 2 USB, CDRW, 256MB RAM, 30GB HDD. No ADC, no audio in, no airport, no bluetooth. Give it CPU daughtercard and sell upgrades as well.

AIOs are great, but lots of poorer geeks want towers, not everybody needs a built-in monitor.

Maybe the 970 is a good chance to do this.
 
Originally posted by Blackcat
dongmin, while I agree with you that all computers shouldn't cost the same (do all hifis? TVs? ovens? cars?) Apple could really do with a sub $900 tower and drop the need for ADC and custom ROMs in gfx cards.

ADC probably adds $10 to the design but more importantly adds a lot of convienence. And Apple had a tower that could have gone in this direction: The PowerMac Cube. It tanked remember?

And Apple doesn't use custom ROMs in the gfx cards they sell - they use the Mac version of the ROM from the manufacturer. The reason why you can't get an nVidia card in the store and flash it is because to my knowledge the ROM code isn't out there. It has been in the past for ATi cards, but I don't know about recent models (yes, I would love to be able to order a Radeon 9800 in a new tower too).

If people could buy a no frills PowerMac, I think more would. 1 AGP, 2 PCI, 1 FW, 2 USB, CDRW, 256MB RAM, 30GB HDD. No ADC, no audio in, no airport, no bluetooth. Give it CPU daughtercard and sell upgrades as well.

If your gonna sells a stripped down model, then you don't need PCI slots or a CPU daughter card, both of which cost additional money to implement. And Apple doesn't sell CPU upgrades anymore - they learned their lesson with the old beigh 601 & 604 lines, if you sell CPU upgrades, people won't buy your newer systems. ADC & Audio In probably aren't worth squat on cost (after all, even low-end PCs often ship with audio hardware capable of sound input). AirPort & BlueTooth might add a cost worth removing for this theoretical model.

AIOs are great, but lots of poorer geeks want towers, not everybody needs a built-in monitor.

If your a poorer geek then you probably want the speed of a tower, but aren't going to need the expansion of it at all (if only becuase you are too poor to fill the case to the brim). But not all things are available to all people, so to the poorer geeks out there, you will just have to compromise. A cheap Tower wouldn't help you because it will be a cheap Tower - it will be yesterdays system, the old Low end of the PowerMac line with a single processor and subpar systems. Do you really want that [for the same price as an AIO]?
 
Originally posted by dongmin
You can't expect Macs to be price-competitive with Dell for the simple fact that Apple depends so much on R&D and invests heavily in it. It's one company doing all the R&D for the software, industrial design, motherboard, services, etc.

As the rest of the industry has been taking design and architecture cuers from Apple for years, they are in away doing this for the whole industry.
 
Originally posted by Mr. MacPhisto
I've also had a thought about bringing back the cube on the low-end. The iCube maybe? Only question is how can it be made upgradeable. I have some ideas and drawings on how to do this with the eMac, iMac, etc cause I think this is extremely important.

Originally posted by Mudbug
Fortune 500 companies don't have thousands of P4 hotrods on desktops, they have netware boxes that are basically just end nodes of the network. What about just a simple network "appliance" (again, don't shoot me) for the enterprise that fits needs but doesn't overachieve?

Here's a cool idea.

Bring the cube back as an office network node. Small form factor... cheap cause it's not so big... easy to sell in large quantities with a great package price. Make them with 970's and you got a great machine.

Maybe even make a mini-(or non-mini)tablet for offices where an easy wireless network environment is key for productivity.

Here's another idea for enterprise: MAKE XSERVES WITH POWER4's OR EVEN POWER5's. If Apple ever got into the big enterprise market, a Power5 Xserve would be awesome. Ya think?
 
Originally posted by imaswitcheryeah
MAKE XSERVES WITH POWER4's OR EVEN POWER5's. If Apple ever got into the big enterprise market, a Power5 Xserve would be awesome. Ya think?

Do you mean "expand the server line to more capable systems" or "cram a POWER4 chip into a 1U Xserve box with limited expandability and availability"? I'd hope that it's the former.

A simpler, faster way would be to simply rebadge some of the IBM server line, and port OS X to the IBM POWER systems. Good way to test the waters for high-end servers without a huge upfront R&D cost.

For example, the pSeries 630 6C4 (http://www-132.ibm.com/content/home/store_IBMPublicUSA/en_US/eServer/pSeries/entry/6306C4.html) - a 4-way POWER4+ in a 4U chassis. Base configs from $14K (single 1.2GHz, 1GB) to $40K (quad 1.45GHz, 8GB). Memory expansion to 32GB, 6 PCI-X (64bit/133MHz) slots....

BTW, about the hot rod P4, try to find a business desktop from HP or Dell with less than a 2GHz P4 (or P4 Celeron). True that they're not "hot rods", but they're pretty fast machines nevertheless.
 
Originally posted by AidenShaw
Do you mean "expand the server line to more capable systems" or "cram a POWER4 chip into a 1U Xserve box with limited expandability and availability"? I'd hope that it's the former.

A simpler, faster way would be to simply rebadge some of the IBM server line, and port OS X to the IBM POWER systems. Good way to test the waters for high-end servers without a huge upfront R&D cost.

For example, the pSeries 630 6C4 (http://www-132.ibm.com/content/home/store_IBMPublicUSA/en_US/eServer/pSeries/entry/6306C4.html) - a 4-way POWER4+ in a 4U chassis. Base configs from $14K (single 1.2GHz, 1GB) to $40K (quad 1.45GHz, 8GB). Memory expansion to 32GB, 6 PCI-X (64bit/133MHz) slots....

BTW, about the hot rod P4, try to find a business desktop from HP or Dell with less than a 2GHz P4 (or P4 Celeron). True that they're not "hot rods", but they're pretty fast machines nevertheless.

Sure, whatever.:) I don't know much about enterprise servers and such... I just thought it would be pretty cool if Apple adopted the entire IBM processor line. I know the 970's were basically made specifically for Apple (so Apple can finally have a new generation chip). I was just painting a pretty picture of, let's say, 970's across the product line (yeah, even ibooks), with an xserve enterprise edition featuring the Power4-5 chips. Something along those lines.

Didn't I see somewhere around the rumor mill there might actually BE an xserve enterprise edition? I remember it being along side a rumor of xstations, a 2-4 way mac workstation. Word up, thanks for replying about my idea!!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.