Ok so I would really like a Mac Pro right now so I can use one HD to run OSX and another HD to Run Vista for some gaming ect. I will be buying a MBA alongside the Mac Pro for obvious reasons, but I have an eee pc 1000H and Dell Insp 1420 both fully loaded so the MBA is not a big issue here. I can certainly hold out to see if there is an update in sept/ oct to the pro, but waiting until Jan '09 may be a little too long. My main reason for not buying now is that the video card does not have an HDMI output and there is no blu ray writer. Another big thing holding me back is the technology coming out, mainly nehalem and DDR3 memory and Snow Leopard. While a MP right now would really solve my issues of just having one main computer for ALL of my music an DVD's (I have almost all of my DVD's down to iPod format) in one place that my laptops can feed off of. The power of the MP to do these conversions is another reason I really want one (I only need a quad core though- 8 Cores is overkill for what I need). Basically I feel like I should wait because it is not really a necessity, its just a pain to have a laptop hooked up to an external monitor when I could have a true desktop. I like how you can essentially future proof yourself with the Pro, but I feel like these next 8-10 months are the time that the casing of the current one and even the tech is going to get outdated fast.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Mac Pro- Buy Now or in 2009?
- Thread starter jfriedman8
- Start date
- Sort by reaction score
The way you phrase outdated sounds like this Mac Pro will be unusable in eight to ten months. 
There are 3rd party Blu Ray options you can find to put in the Mac Pro.
There are 3rd party Blu Ray options you can find to put in the Mac Pro.Wait.
Particularly due to your statement about converting video and gaming. The Nehalem processors at 2.6Ghz blow even the 3.0Ghz QX9650's out of the water. There's no comparison.
This isn't a minor spec revision or speed bump, it's the end of an architecture cycle for numerous components.
You're going to Nehalem, with integrated memory controllers. That's a big deal. You also have DDR3 being rolled out. PCIe 2.0, the 4000 series ATI cards, Blu-Ray perhaps becoming standard, etc. etc.
Even if you listed all the options, Nehalem is a good enough reason to wait. And if you plan on gaming, the leap between the 3000 series and the 4000 series ATI cards is night and day.
The resale value on a Mac Pro bought now will drop significantly more this update cycle due to the major revisions likely being pushed out.
Particularly due to your statement about converting video and gaming. The Nehalem processors at 2.6Ghz blow even the 3.0Ghz QX9650's out of the water. There's no comparison.
This isn't a minor spec revision or speed bump, it's the end of an architecture cycle for numerous components.
You're going to Nehalem, with integrated memory controllers. That's a big deal. You also have DDR3 being rolled out. PCIe 2.0, the 4000 series ATI cards, Blu-Ray perhaps becoming standard, etc. etc.
Even if you listed all the options, Nehalem is a good enough reason to wait. And if you plan on gaming, the leap between the 3000 series and the 4000 series ATI cards is night and day.
The resale value on a Mac Pro bought now will drop significantly more this update cycle due to the major revisions likely being pushed out.
Well the Mac Pro's use the X5400's, so why your quoting the iMac chip is a mystery.
The memory? It's only major deal if you believe the intel hype. And DDR3 is extremely expensive and will continue to be so until it is adopted across the industry, which given the price will not happen for a considerable time.
Blu-Ray wont become standard next year. It goes against Apple's recent policy of live streaming and would kill HD downloads from the iTunes Store.
As for your idea of the current Mac Pro's resale value falling significantly? I doubt it, most people don't know which chip can do what, let alone the new architecture.
As the original poster states he only needs 4 cores. so buy one know.
Because the only widespread demo has been of the desktop Nehalem, not the server variant.
And secondly, what iMac has the QX9650. No really, show me an iMac with a QX9650.
And in that Nehalem demo, the gains were in the 30% range. One would likely assume the same gains would be seen between the current Xeon and the upcoming server Nehalem chip.
The X5400's are simply higher binned QX9650's. The only major difference is the locked multiplier. They perform relatively identical at stock settings, albeit with slightly higher stability. Just because one is socket 771 and the other is 775, doesn't mean anything.
The only difference is that the Xeon chipsets use FB-DIMMS, which grants them the advantage in certain multithreaded tasks.
It's a major deal because FB-DIMM production is all but halted. ECC registered RDDR3 has taken over.
You claim my comparison of a QX9650 with Nehalem is invalid and then use the desktop DDR3 argument? You do realize the Mac Pro's are essentially server parts, right? They don't use DDR2, they use FB-DIMMS. And they won't use standard DDR3. They will use R-DIMMS.
But keep going on like you know what you're talking about.
Which is why I included the word "perhaps". And you act as if you can't go out and buy a Blu-Ray drive and stick it in a Mac Pro. If people want to purchase hard copies of HD video, they will. They'll even go out and get an external Blu-Ray drive for their Mac Pro (which gives no money to apple on the hardware purchase) or buy an internal and have it installed (which gives only the installation costs to Apple, if they even choose not to do it themselves)
The only thing keeping Blu-Ray out of Macs is low write speeds (The fastest available commercially I believe is 4x, and the highest demoed is 6x) and the high cost of the drive/discs.
As soon as that starts coming down, you'll see more widespread adoption. Not everyone has Verizon FiOS, not everyone has the bandwidth needed to download HD video quickly from the iTunes store. The same adoption problems arose with DVD when it debuted. The players were expensive, the write speeds were slow, the discs cost too much.
And guess what happened? Prices fell, adoption became widespread.
The whole discussion on the Blu-Ray topic is moot unless the prices drop significantly by January. Apple does have a habit of adopting hardware earlier even if it does cost more, but the Blu-Ray is just slightly out of their range.
Until 20 Mbit is widely available to Americans, this notion of "hard media is dead" is a pipe dream. Most people don't poney up for more than 3-6 Mbit let alone 8/10 sometimes offered by the cable companies. And even then you rarely get the speed you pay for. Personally, I don't think we'll see widespread dropping of hard media distribution until we have fiber to the home in at least 40% of America.
No, most people don't care when looking to buy used Macs when the difference is just clock speed bumps, die shrinks and bigger hard drives. Because they know the gain is minimal. But those older hardware models still command slightly less.
But with major revisions like this, it's almost as big as a G5 iMac and the first intel iMac. Not in terms of the older hardware not supporting the new OS/software eventually, but in radical hardware departure. While they are relatively the same age, the hardware difference is so significant that people will assign drastically different resale values to them.
It recalculates the entire used market price.
Rubbish.
And DDR3 isn't faster than DDR2 for the same MHz speed due to higher latencies.
Agreed. I don't think Apple's going the Blu-ray route - ever.
4 cores or 4 faster cores?
Didn't Apple adopt DVDs early? And Apple offers high-priced BTO RAM options too...
I'm looking at my internal BD-RE/HD DVD-ROM drive sitting six feet away from me. It burns at 6x max. I am not a pretester. I bought it for $327 in March.
Apple has a habit of adopting late and costing more. Even Steve admitted that they were late in adopting CD-R/W drives.
Then you're a fool for thinking Mac Pros use standard DDR2/DDR3. And the latencies for 1066 DDR3 have already hit 6-6-6-15 and 5-5-5-15.
Even if they went with 1333 or 1600 or 2000 with the slower latencies, you're reading/writing at the same time, and the difference between FB-DIMM DDR2 800 and R-DIMM DDR3 1333 is leaps and bounds.
Not to mention the TDP gains. The FB-DIMMs had HORRIBLE TDP.
That's hardly the main contention. Apple goes where the mainstream market goes. If Bluray drops significantly in price and it achieves widespread adoption, they'll have to offer it as an option.
Or figure out a compelling argument why people should pay slightly less for a DRM'd copy that they can only play in iTunes and that can't be burned to a disc. Oh and that if you delete, you can't re-download despite the fact you already purchased it, unless you plead and whine to the sales department and they grant you a one time redownload.
So yea....good luck with that.
And they adopted LCD's early and a host of other hardware. And it was the correct decision. Because prices on DVD tech fell fast enough that they could recoup their early adoption costs. Blu-Ray simply hasn't been dropping fast enough, but it's getting there.
This time, their decision is wrong. They're basing their assumption on false averages for the connection speed of Americans. We rank near the bottom in terms of Mbit connections for consumers.
The average connection speed is likely around 3 Mbit, and downloading a Hi-Def movie at that speed (even when it's compressed) takes over an hour. And that's assuming you actually GET a solid 3 Mbit throughout the download.
We won't have total abandonment of a hard copy distribution method until we have 20Mbit to the homes or Fiber to the home.
I split my time between the US and Europe. We have 20 MBit standard there. Yet you don't see people chucking out DVDs left and right. People are not downloading Hi-Def. They are waiting for BluRay prices to drop down. It's the idea of having something tangible when you pay for it.
It's rebranded Crucial. So you're paying for Apple to rebrand Crucial and snap it into the socket. People who care about their budgets buy FB-DIMMs online and put it in themselves.
Ah well then the 6x burners are on the market. But I don't think you can purchase discs higher than 4x, rendering that extra burning speed useless at the moment.
They don't always adopt late. They were early with LCD.
Hey i can understand what kntgsp is saying about the whole major revision
about Nehalem, but is the same thing not going to happen from Nehalem to Sandy Bridge? I know people are saying there is always better technology on the horizon but this is not actually too long away, and i know people will then say that 2 years is lightyears in the tech world, but in the apple world it seems to be quite a bit slower then that...so its not that long of a wait for me since its such an expensive buy.
Sorry, im also looking to purchase a new mac pro and don't understand all of this stuff and it seems that you do. You say that many components will have to have new architecture types with nehalem...does that mean that when sandy bridge comes out (with the new cpu architecture) that people will soon stop developing (the cutting edge) hardware of the architecture required by nehalem? Since im planning on keeping this base configuration model that i purchase for about 4 years before doing any upgrades, id like to think i could bring it up to speed with some updates..but if the architecture is going to change all over again for sandy bridge which will be released in LESS then 4 years (if im right?) then is it not smarter for someone in my position to wait for sandy bridge at least?
Or will sandy bridge not require new component architecture too?
Sorry im confused!
(also, while i have you! i heard that nehalem's speed improvements dont scale very well past 4 cores? can you comment on purchasers wondering whether to invest in dual quad core, or sticking to single quad core? Thank you all and anyone who replies ^_^ ).
about Nehalem, but is the same thing not going to happen from Nehalem to Sandy Bridge? I know people are saying there is always better technology on the horizon but this is not actually too long away, and i know people will then say that 2 years is lightyears in the tech world, but in the apple world it seems to be quite a bit slower then that...so its not that long of a wait for me since its such an expensive buy.
Sorry, im also looking to purchase a new mac pro and don't understand all of this stuff and it seems that you do. You say that many components will have to have new architecture types with nehalem...does that mean that when sandy bridge comes out (with the new cpu architecture) that people will soon stop developing (the cutting edge) hardware of the architecture required by nehalem? Since im planning on keeping this base configuration model that i purchase for about 4 years before doing any upgrades, id like to think i could bring it up to speed with some updates..but if the architecture is going to change all over again for sandy bridge which will be released in LESS then 4 years (if im right?) then is it not smarter for someone in my position to wait for sandy bridge at least?
Or will sandy bridge not require new component architecture too?
Sorry im confused!
(also, while i have you! i heard that nehalem's speed improvements dont scale very well past 4 cores? can you comment on purchasers wondering whether to invest in dual quad core, or sticking to single quad core? Thank you all and anyone who replies ^_^ ).
Perhaps, but that's sooo far down the line. If you were still on a Penryn or Yorksfield by then it'd be crazy slow in comparison. And why are you worrying about that? I'd be like people who adopted Core2 more than 2 years ago worrying about the upcoming Nehalem release. It's pointless. Always better to be on the leading edge than the trailing. Nehalem is a leading, as is Sandy.
Don't concern yourself with the die-shrink to Westmere if you only upgrade every 3-4 years.
That's correct if those 2 years all occur within the same architecture cycle. There isn't earth shattering change. But this is a complete architecture change. I'm not sure you understand what that means, and I'm sorry if it's a bit technical, so let me try to explain a little better:
There is no more front side bus, the memory controller is on die. There is no more FB-DIMM, it's R-DIMM Etc, etc, etc. These are not tiny "bumps" or "shrinks" to the same device. They are complete redesigns.
It's akin to the motor in your car. Instead of just carburetor to fuel injection, this is like Gas to Electric. Completely different system. Much more efficient. Much faster. (cue Honda Insight jokes)
This isn't simply "oh we're bumping from 2.5 to 2.8 Ghz. This is the difference between Pentium 4 and Core 2 Duo. THAT'S why it's a big deal. And because the Apple crowd didn't see Pentium 4, they switched from G5 to Core 2 Duo, they don't understand how big of a deal this is.
Well look at it this way....
People don't make AGP video cards anymore do they? If a newer, better architecture comes along, the market tends to adopt it while offering limited legacy support. But what you're worrying about is a ways out since there's a die-shrink before Sandy Bridge.
But good question. Glad to see someone asking that.
You have nothing to worry about if you wait for Nehalem. There will first have to be the die shrink of Nehalem before the Sandy Bridge chips are released.
The same as Kentsfield to Yorkfield. Gains, but nothing earth shattering.
Likely No. You have to realize though that it's a while out. It won't even DEBUT until 2010. Let alone be released to consumers.
[/QUOTE]
Honestly it has more to do with the operating system and/or the program being used. The cores scale very well, but with any system you lose a certain amount of gain for each scale improvment.
For instance, DivX encoding doesn't scale well on the Nehalem past the first 4 cores, and any improvement is due to architecture and not core number. This is a program deficiency, not a hardware one. It simply needs to be rewritten to take advantage of more than 4 cores and new sets of instructions.
And for Apple programs, you'll likely see that more in the next year or so as 8 core variants become the norm. Now in terms of what you should buy in January? Depends on what you do.
You won't see an advantage unless the program is specifically written to take advantage of more than 4 cores. I don't use a lot of the Apple video/audio editing software so I couldn't tell you. Perhaps someone can weigh in.
But I guarantee that if it's an Apple application that they will add improvements to take advantage of those extra 4 cores, because they are selling them. People wouldn't buy them unless they received a performance improvement.
Now with non-Apple software, it'll be hit or miss in terms of whether or not it scales above 4 cores. There, you're at the mercy of whether they decide it's worth it to do so.
All that being said,
I still try to upgrade every die shrink though. So for instance, I upgraded once Core2Duo went from 65nm to 45nm. And likewise I'll do so when Nehalehm shrinks from 45nm to Westmere at 32nm
But for people who wait until architecture changes. It's better to be on the leading edge during those switches than the trailing edge. Because you don't get left in the dust as quickly.
I didn't want to quote all of that but thank you very much kntgsp, you've answered a great deal of my questions very thoroughly and it is much appreciated, now i think all my concerns about upgrading with nehalem are clear. Thank you again!
I read somewhere that a feature the nehalem processors will have is that when extra cores are not being used they are underclocked and the core that is in use gets overclocked balancing out heat dissipation etc. I know this is a silly thing but it gives me the impression of it being finicky like one of my cores (the one in use) is about to die...i understand no overall rapid change in heat but what about super small locally it will be right? I think its me being paranoid whenever i hear the word "overclock" after an experience with a friend and his graphics card!
Anyway this is fantastic news and now i cannot wait to upgrade! *high fives everyone in the room*
I read somewhere that a feature the nehalem processors will have is that when extra cores are not being used they are underclocked and the core that is in use gets overclocked balancing out heat dissipation etc. I know this is a silly thing but it gives me the impression of it being finicky like one of my cores (the one in use) is about to die...i understand no overall rapid change in heat but what about super small locally it will be right? I think its me being paranoid whenever i hear the word "overclock" after an experience with a friend and his graphics card!
Anyway this is fantastic news and now i cannot wait to upgrade! *high fives everyone in the room*
Hahaha that's funny. Because I do quite a bit of overclocking in my spare time.
As long as the heat sink keeps the core temp below spec, it doesn't matter.
It's meant as a legacy support somewhat. Older applications are single threaded. And since the Nehalem won't be clocked very high (Sandy Bridge is supposed to hit 4Ghz standard), it helps with that. When you get a single threaded app, it slows down the other 3 cores and ramps up the one, allowing it to crank out that single threaded app more quickly.
Also you won't see a rapid change in temp if the heatsink is large enough.
For instance, clocking an E8400 from 3Ghz to 4Ghz, if you use a good enough cooling system, you'll only see maybe 5C difference in idle temps.
It's when it's under load that the temps go up, but that will happen regardless of what speed it's clocked at. And with that being a feature, you can bet the heatsinks will be properly designed to dissipate the heat during those "overclocking" moments.
Also, it's an Intel included feature, not some guy in his garage, haha.
I see, that's actually very reassuring. Look at me questioning intel developers ^_^.
I read :
I know that you probably don't have access to a magical crystal ball, but do you think that would be possible to do by just swapping out processors? Or will i have to get a whole new mac pro?
i love this, best tech support i've ever gotten! It's much appreciated
I read :
I know that you probably don't have access to a magical crystal ball, but do you think that would be possible to do by just swapping out processors? Or will i have to get a whole new mac pro?
i love this, best tech support i've ever gotten! It's much appreciated
You can still buy upgrades for computers from 2000, you have nothing to worry about. Upgrades will be availible.
Sandy Bridge is another "Tock" in Intel's "tick tock" strategy. Ticks are die shrinks, tocks are new architectures. Intel is going to continue this strategy, because it works really well. While the silicon will look different, much of the stuff will be the same. For example, PCI-e isn't going anywhere any time soon. There is no reason to skip the upcoming Nehalem unless you're happy with what you've got.
Probably not, they most likely won't be pin compatible or electrically compatible.
I see, thank you m1stake. I think it was mainly the updates i was worried about and you both cleared that up for me.
I don't earn a living from using my computer so I can't personally justify so many whole system upgrades. I know macs hold their value quite well, but I would prefer to avoid selling everything as i've been shafted through transactions a few times before. Anyway!
I think to be honest i am getting carried away and that nehalem will be more than enough for my needs!
Edit: By the time Sandy Bridge 4GHz processors are in mac pros (2011-12?) I will be nearing the end of my planned cycle for the mac pro anyway, right?
I don't earn a living from using my computer so I can't personally justify so many whole system upgrades. I know macs hold their value quite well, but I would prefer to avoid selling everything as i've been shafted through transactions a few times before. Anyway!
I think to be honest i am getting carried away and that nehalem will be more than enough for my needs!
Edit: By the time Sandy Bridge 4GHz processors are in mac pros (2011-12?) I will be nearing the end of my planned cycle for the mac pro anyway, right?
Well Kentsfield and Yorkfield were pin compatible, so it's tough to say. Die shrink doesn't necessarily mean that they use a different pin set.
But switching out a CPU on a Mac Pro would most certainly void any remnant of warranty, and I haven't heard of many people doing this, so I think it's a moot discussion point.
Sandy Bridge is NOT a shrink of Nehalem. That is Westmere. Sandy Bridge is a new microarchitecture, like Core and Nehalem.
Wrong decision or not, all signs point to no Blu-ray.
You missed my point. The high prices of BTO RAM mean that cost isn't a factor in why Apple does not use Blu-ray.
Ahh I completely mixed up Westmere and Sandy Bridge, thank you.
Wow, I need a cup of coffee if I made a slip up that bad. Too many threads at once.I've amended the post to reflect that. I don't want him getting that confused.
I assumed he was worried about the Nehalem die-shrink in relation to the Kentsfield/Yorksfield comment and how people worry themselves with minor revisions. I'm far more concerned with the Nehalem prices and specs than anything else at the moment. I can't imagine why anyone would worry about Sandy Bridge at the current date.
Worrying about anything that far out right now is pointless anyway though.
The signs point to no adoption of Blu-Ray right now. Not forever. Apple has to cede to market forces. As mighty as they are in certain areas (digital music distribution, music players, and now phones), they still have to offer what consumers want. If Blu-Ray adoption becomes widespread, they will have to offer it as an option to stay competitive.
If Blu-Ray drives become standard in every household computer, and Apple clunks along with DVD, it will hurt sales and they know that. I'm not arguing that Blu-Ray adoption will become widespread. I'm arguing that IF it does, Apple will have to follow. Look at firewire, they thought that was a great idea for consumer products until the rest of the market went USB 2.0 with EVERYTHING and people thought of firewire as a gimmick.
And cost IS a factor. It's just not the cost of the player alone. The one time price of the Blu-Ray drive might not be an issue and neither the Apple markup, but people have to buy the Blu-Ray discs. And those are still very expensive.
People don't have to buy additional objects to use the RAM they buy. It's initial cost vs. cost to use.
RAM has no cost to use. Apple consumers who suckle at Jobs' teat will gladly pay a $100 premium on RAM if Apple says it works better. But people don't buy their DVD-R's and CD-R's from Apple.
Not even the most blindly loyal customer will pay $50 for a 10 pack of DVD-R's.
Sorry, I'm a bit of an amateur, but I have a few questions...
1) What do you mean by "die shrink"?
2) What are you referring to when you say "nm". For example "from 45nm to Westmere at 32nm"???
3) When you mentioned intel's process or "tick-tock," what do you mean by this? What's the more important of the two, in terms of upgraded hardware?
4) If sandy bridge is a entirely different architecture just like nahalehm is compared to core2duo, what is the incentive to buy a new mac laptop when they come out in 2009 vs. just waiting another year for sandy bridge? (considering it's a fairly big jump from nahalehm, just like nalalehm is from core2duo)
nm is an abbreviation for nanometer. It's the fabrication process. A die shrink is taking the same architecture and simply shrinking the fabrication process.
i.e. 65nm Core2Duo (Conroe) to 45 nm Core2Duo (Penryn). They are the same architecture (Core2Duo) but smaller fabrication. This gives advantages of less power consumption, sometimes added instruction sets, higher clocks, etc. The amount of improvement this die shrink yields is relatively minor. Nothing earth-shattering.
I actually didn't mention tick-tock, but it's a great way to explain the situation.
Tick-Tock is bound to nm design spec. So the tick and tock portions both fall under 45nm or 65nm. That is to say that you can't have tick be 65nm and tock be 45. Tick is the die shrink and Tock is the architecture change.
Penryn is the Tick, and Nehalem is the Tock, both of which are 45 nm fabricated processors. It's always better to adopt on the tock cycle.
Whether or not you wait for Sandy Bridge over Nehalem depends on how long you want to keep using your Core2Duo. It's akin to individuals who have Pentium D's waiting until Nehalem. Do you know many people who still use Pentium D's?
The absolute earliest you'll see any Sandy Bridge chips at all is 2010. But Intel has stated repeatedly that it will update architecture every 2 years. Meaning Sandy Bridge will debut 2 years from the debut of Nehalem. That means late 2010, possibly even Jan 2011.
Just because Intel says 2010 for Sandy Bridge doesn't mean you'll see it in 2010. It's somewhat dependent on when the previous architecture debuts. If there are production hangups, problems with the design, etc. that pushes it back further.