I think the reality was those people were overly optimistic that SB-E would make it into Apple's hands before it was officially released. Its only thing to hope, its another thing to realistically count on something happening.
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.
"Intel demos Xeon E5 due for Mac Pro": does this not scream 'Mac Pro not dead'?
- Thread starter malch
- Start date
- Sort by reaction score
I think the reality was those people were overly optimistic that SB-E would make it into Apple's hands before it was officially released. Its only thing to hope, its another thing to realistically count on something happening.
Yeah, I just meant someone will always buy in to whatever gets posted in the news section on here and assume it is 100% fact as they read it and processed it and then argue it to death. Never changes, except now there are more nonsense rumours for more page hits and MR reposts them all.
I think the majority of us just wish that we had some better idea of what was coming and when. (obviously more important in the enterprise and pro segments than with gadgets) With most hardware companies other than Apple, you see what is in their pipeline for months ahead of time. I get Apple's style, but it is frustrating not really knowing what they will come up with, or not.
Nevertheless, we need to keep in mind that this is mac"rumors" site and not necessarily mac"facts".
Nevertheless, we need to keep in mind that this is mac"rumors" site and not necessarily mac"facts".
I think the majority of us just wish that we had some better idea of what was coming and when. (obviously more important in the enterprise and pro segments than with gadgets) With most hardware companies other than Apple, you see what is in their pipeline for months ahead of time. I get Apple's style, but it is frustrating not really knowing what they will come up with, or not.
Nevertheless, we need to keep in mind that this is mac"rumors" site and not necessarily mac"facts".
At this point, a rumor is better than nothing
OS X is morphing into iOS. SJ always said to focus on one product and not many different versions of same - his biggest critisism of Microsoft "spread too thin over too many products". (IIRC I read an article mentioning that actually the person in charge of iOS has a drive very similar to SJ and might be the best person to lead Apple except he's not interested.)
When I saw every granny and their cat and dog in Hong Kong dragging their last pennies from under their mattress to subscribe to a dotCom IPO in Hong Kong I knew the boom was over and started to sell short. (you should have seen the que's at the banks with crowd control, it was a circus) Similarly here - when every one is screaming "Apple, Apple" I know the turn around is getting pretty close. Am watching..... (Arrogance comes before the downfall.)
While this is true, I don't ever see them becoming the same OS. OS X will continue picking up iOS APIs, but we're never going to see OS X completely become iOS. It's just that a lot of technical work is being done on iOS and Apple doesn't want to do it twice.
A good example is AVFoundation, which started in iOS 4. Why deal with the painful mess that is QuickTime when they could just bring AVFoundation to iOS?
That basically happened once when they accepted higher tdp chips than the other oems if I'm not mistaken. Intel and Apple don't seem that close these days.
Well this rumor doesn't add any new information. We all see articles on Sandy Bridge - E.
Somehow I am guessing they will at least become very close to one another. OSX was supposed to last two decades initially right (something like that)? By that standard it's past its half life, and IOS seems to be their main point of concentration in terms of development.
I don't think that's what will happen. I think at the end of OS X's life span all Macs will be discontinued. There will never be a full merge.
But like you've pointed out, that's 10 years from now.
Perhaps because the Mac Pro is the only desktop computer Apple makes that has an easy to open case? Or the only one that has some ports on the front to make them easier to use? Or the only one that offers any expandability?
There are lots of reasons to own a Mac Pro even if your computing needs do not require the power of the Mac Pro.
from hardmac today:
"We missed something in the information package made available when the Sandy-Bridge E CPU were launched, CPU that would define forthcoming Xeon Sandy Bridge.
They have officially 6 cores and 15 MB of cache, but in fact they have 8 cores and 20 MB of cache on the die. Intel decided to deactivate 2 cores and 5 MB of cache that could be reactivated on next generation Xeon. Of course to maintain the same TDP, or similar figures, the clock frequency will have to be lowered, but we could then have Mac Pro powered by 2x 8 cores CPU, generating 32 threads.
In addition, if we trust the data published on a Chinese website, Intel might release next year CPU with even more cores:
Of course, only Pro users would be interested in getting a Mac Pro with 20 cores and 40 threads... In addition, cheery on the top of the cake, the socket compatibility will make it possible to upgrade Sandy-Bridge based hardware with Ivy-Bridge CPU, so it will be possible to significantly boost the lifetime of a Sandy-Bridge Mac Pro."
"We missed something in the information package made available when the Sandy-Bridge E CPU were launched, CPU that would define forthcoming Xeon Sandy Bridge.
They have officially 6 cores and 15 MB of cache, but in fact they have 8 cores and 20 MB of cache on the die. Intel decided to deactivate 2 cores and 5 MB of cache that could be reactivated on next generation Xeon. Of course to maintain the same TDP, or similar figures, the clock frequency will have to be lowered, but we could then have Mac Pro powered by 2x 8 cores CPU, generating 32 threads.
In addition, if we trust the data published on a Chinese website, Intel might release next year CPU with even more cores:
Of course, only Pro users would be interested in getting a Mac Pro with 20 cores and 40 threads... In addition, cheery on the top of the cake, the socket compatibility will make it possible to upgrade Sandy-Bridge based hardware with Ivy-Bridge CPU, so it will be possible to significantly boost the lifetime of a Sandy-Bridge Mac Pro."
No. Unless Apple was taking out-of-spec chips; which seems odd.
Apple merely launched 27 days before Intel (Mac Pro March 3, 2009. Intel Bloomfield March 30, 2009). That is less than a month and a window in which Apple and all the other large contract buyers were probably being shipped some chips. More likely, since Apple's volume is "just large enough" for a early contract and it had been over a year since last update (1/2008 -> 3/2009), they just shipped anyway with the early "low flow". Prior to this Apple had been a month (or so ) behind the launches. (more on the initial Intel launch but that is an outlier because "had to switch to Intel inside of a year" due to edict. )
This is similar to the current situation where some of the supercomputer vendors are getting them ( need a couple of thousand) before everyone else deploys them in volume. Plus the Top500 list only comes out every 6 months so they (the vendors and Intel) had a hard deadline of this week to get something going, least have to wait another 6 months.
With at least a 6 months gap between last Mac Pro launch, there is likely a decent backlog of Mac Pro orders out there. Folks sitting on the sidelines with money who will buy on product refresh. It would not be a wise move for Apple to bet that Mac Pro buyers were going to turn out in low numbers if try to launch on "pre launch... get production ramped up" supply. Even if they did it would dubious position: "Yeah we're late but hardly anyone is going to buy this anyway". That will certainly spawn a "Why are we selling this in the first place" kinds of questions.
Uhm, it's well-known that the new Xeons are 8-core.
It's the high-end users that only get the 6-core parts. Workstations get the "uncrippled" version...at a price
After all, the 6-core version is the salvage part.
Uhm, it's well-known that the new Xeons are 8-core.
It's the high-end users that only get the 6-core parts. Workstations get the "uncrippled" version...at a price
After all, the 6-core version is the salvage part.
Or you can think of it as having 2 cores turned off to achieve higher clock frequencies.
No. There was a small community that cared. The primary problems was that it was small and that computers get alot better every couple of product cycles.
The Mac pro only has 4 slots for PCI-e and RAM. It is the only "shove cards and disk" model in Apple's line up. There is no mid-tower. There is no mini-tower. No HTPC sized box where can put in your own card. ...... So as far as being "limited" it is in pretty much the very same boat as the XServe. So this really boils down to is it selling enough in number.
Utter nonsense. "OS X Server" has always largely been a set of applications on top of the mainstream OS. Beyond some minor tweaks to the kernel parameters that were already present there is not really special or highly optimized about the OS core for the server model. They (just like most other software and system vendors) just charged megabucks for the "Server" version because they could.
For many small businesses and departmental groups Mac OS X Server is just fine. Where the local support person is doing Mac OS X anyway and isn't predisposed to Windows and/or Linux, it is a pretty good fit.
Given that the number of Mac OS X Server deployments shot up once the Mini Server bundle arrived, Apple is doing exactly the opposite of killing off X Server. OS X Server is growing. It does not have a "nobody is buying it problem".
Two fundamental flaws in this hand waving.
First, a recent rumors article came out with estimate that Apple is on track to sell 5 million Macs in a quarter. Being a bit conservative at 4.5M/quarter that's 18M/yr. 10K/18M ==> 0.0005 In other words, sell less than 0.1%. Fall out of the single digit percentage points and it is likely a dead product. An order of magnitude below that and it is not even a question. Even if Mac Pros are selling at 100K/yr they are on slippery slope. The minimal amount to avoid the death watch list is going up every year. The argument of selling into a stagnant market is flawed.
Second, you don't need a Mac Pro to compile a couple million line program. Using build farms is a simple way to get around the "compile everything from scratch takes a long time" problem. Indeed, there is already an Apple tool for that.
http://developer.apple.com/library/mac/#documentation/Darwin/Reference/ManPages/man1/distcc.1
Furthermore, almost no developers need to commonly do that. If tweaking a some functionality commonly requires the whole program be recompiled there is something majority flawed with the code. Developers spend a large amount of time staring and typing at the editing and debug screens. You don't need a supercomputer to accomplish those tasks. Likewise, recompiling a library module shouldn't require 16GB of memory to resolve the linkage. Sure, there should be some "build from scratch" integration builds to be done periodically, but there is no necessity it has to be in a single box.
Compiling software is one of those "embarrassingly parallel" class of problems. It is not hard to use multiple machines. If the Mac Pro didn't exist those who stand to make "Billions" in Mac Software sales will simply just buy multiple of the less costly machines and continue to make money.
It is a nice to have a cluster in a single box. It is not necessary to have a cluster in a single box.
$1099 Mini 2.7GHz Dual i7 with 8GB Apple RAM
x3 as compute nodes and a $999 min server as cluster file server: $4,296.
$3,924 Mac Pro Quad 2.8 with 24GB Apple RAM and two 1TB drives.
While higher priced, the mini cluster actually has 2 more "compute" cores (and two more file server cores). That's just the current mini. An affordable, faster cluster interconnect would make it even better. Likewise, each future node going to 4 cores even if the Mac Pro goes to a nominal 6. Even if the cluster is marginally slower and $300 more expensive... if that is gap between making $100-1,000K or not ... most businesses are going to pull the trigger.
The market for some apps that require non-mainstream video cards might dry up, (high end AutoCAD), but the vast majority of Mac Apps are generating money because they are being sold to that greater than 98% of folks who do not have Mac Pros. That's where the money is. Selling to those people who are buying significant numbers of Macs. Not in kowtowing to the groups that are not buying Macs. That is just backwards.
The Mac Pro doesn't need to outsell the other Mac categories, but it does need to keep up in terms of growth. If the growth dies for an extended period of time ( and falls into the very low single digits , or less, of the Mac product mix), then the product will probably be removed. Great products don't have negative growth. Apple is about making great products for people who find value in those products.
The Xeon E5's should add some value back into the Mac Pro feature set. Growth will probably pick up again once they are released. But it would not be surprising that it is on "death watch" by the folks who track numbers. If that growth doesn't show up there is no "get out of jail free" card here. Nor is there a "steal from Peter to pay Paul" card either ( eat into iMac sales to boost Mac Pro sales).
The problem with the XServe is that AppleCare didn't cut it, and the server market needs specialized 24 hour support with on site service. Apple had to either match that, or get out of the market. They chose the latter.
Mac OS X Server was also notoriously buggy in the IT community. Same problem as above.
I worked in a job where we were literally shipping entire XServe RAIDs back to Apple for debugging (note that Apple was not coming out to look at the problem, see what I mentioned above.)
Apple has become a mass-market company -- iTunes, iMac, i-everything. The profit margins for workstations are not worth the hassle.
Meanwhile, pros and prosumers are becoming less and less loyal to Apple because Apple doesn't bow to them anymore. Plus the pricing went patently insane when the July 2010 model came out.
So, anyone who runs Final Cut, Adobe Suites, Pro Tools, Logic or any pro applications, will more and more begin to build their own Linux or Windows systems which have infinitely more options for expansion and "incremental" upgrading.
At the end of the day, few people are going to spend $6,199 for a "bottom-of-the-top-of-the-line" product when they can double their bang elsewhere.
That said, I'm not there yet... I still run Pro Tools on a Mac Pro. But I also have a PC and am not afraid of Windows.
Meanwhile, pros and prosumers are becoming less and less loyal to Apple because Apple doesn't bow to them anymore. Plus the pricing went patently insane when the July 2010 model came out.
So, anyone who runs Final Cut, Adobe Suites, Pro Tools, Logic or any pro applications, will more and more begin to build their own Linux or Windows systems which have infinitely more options for expansion and "incremental" upgrading.
At the end of the day, few people are going to spend $6,199 for a "bottom-of-the-top-of-the-line" product when they can double their bang elsewhere.
That said, I'm not there yet... I still run Pro Tools on a Mac Pro. But I also have a PC and am not afraid of Windows.
Which would be a great plan if Windows 8 didn't also look like it was dumping pros.
Uh? You kinda had me until that $6199 for "bottom-of-the-top-of-the-line" comment.
The basic "top of the line" Mac Pro has 2 processors that each would cost nearly $1500 if you bought them from newegg. By the time you get everything else, you're looking at around $4K for the total package.
While the roughly $2K difference sounds like a lot, it isn't that much when you think about the fact that these are computers people earn there livings using. The time alone it would take you to build this computer will likely cut substantially into that $2k. Ie., maybe it takes you a week of shopping for and buying all the required parts, a day or two for the actual build and installation of the OS and other apps. So, if you would earn maybe $20/hour using a store bought computer and you lost roughly 7 days of time creating your own, you just lost about $1000. Then, what happens when a part fails? You have to tare down your computer and send it back for a manufacture warranty? Which will probably take weeks, so instead of waiting you probably have to buy another part. Sure this can still happen with an Apple, but the chances are greatly reduced that you'll be out of commission for prolonged periods of time.
The entire foundation of OS X is open source.
The problem is re-writing Cocoa and Carbon and the GUI. That's a LOT of work.
compatible with Unix (POSIX) ? Type "linux workstation" into you favorite internet search engine. Slightly more risky but Solaris also ( will have to trail Oracle on source drops and tech releases but would have an POSIX stamp. )
Compatible with OS X (Cocoa APIs like AVFoundation ) ? Exactly clone Mac OS X APIs. You don't need Unix for that. You need a large law firm.
It is a dubious move. Exactly why Apple or any other serious commercial entity is trying to copy Windows. One, you are going to sink almost as much money into lawyers as developers. Two, it is such a large set APIs that by the time you clone it, it will have changed. Users who need the newest APIs (e.g., a new app leverages one) will take the risk adverse approach and just buy real Mac OS X.
I think you are confusing the "Finder" with Mac OS X. You can put another "Finder" on the current Mac OS X. I don't see how you make the OS services simpler and different , but maintain compatibility.
Self limited to a stagnant market. Good luck with that.
I never had a problem that needed onsite specialty. I was the onsite help. Their shipping was like overnight for any parts you needed and I had spare kits as well. Server blows up just grab a logic board, yada yada. Most large enterprise business' only need 24Hr phone or chat. Nothing was ever not resolved. They send you code, binary patches, whatever. I had only 10 Xserves going at any one time. But they anchored the phenomenal client growth over the last few years.
10.5.8 is pretty good still. Using it all over the place. Adobe had a hand in killing OS X as a major player what with ACL, SMB, AFP, POSIX changes every freakin release. Apple went one way and Adobe went the other. Respected permissions one day not the next. If a server OS can't get permissions right I don't see the point. But Adobe really needed to respect ACL's over AFP.