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.
New Mac Pros are here
- Thread starter goMac
- Start date
- Sort by reaction score
I think you're pretty much on the money. We're going to see a very different all-in-one coming in the next couple of months or in 2013. This story explains it better than I could.
http://www.cultofmac.com/169875/what-apples-vibrating-pen-tells-us-about-the-future-of-everything/
Suddenly the iMac folding down onto the desk and vibrating pen patents make sense. Craig Federighi himself made it very clear at last year's WWDC what Apple thinks about touch screens and hands on a computer. But a vibrating pen that works like described in the article above... well, that seems like something Apple will try to do, just to differentiate itself from the upcoming Windows 8 touch Metro interface crapiola, yet still be similar.
----------
I prefer the term "Final clearance".
The articles posted in other threads indicating a new desktop in 2013 does not explicitly state desktop = MacPro.
So such a system is a possibility, and the technology to connect it to high speed peripherals is already out (TB in particular). Granted, it's not equal to x16 or x8 slots in terms of throughput, but Apple's logic could see this as a viable alternative to a traditional box with slots in it.
Better yet, they could make the new system with at least one PCIe slot for a GPU card (gets around the integrated GPU issue better than an embedded mobile chip or attempting to use TB to run a GPU card). That would leave the TB port primarily for fast storage products.
This probably makes sense to Apple. We've already seen how an all-in-one workstation can be created by HP. It wouldn't surprise me to see Apple creating something similar, but seeing the direction they are taking with their mobile computers, I think the days of upgradeable Mac computers are at an end.
This latest Mac Pro "upgrade" is extremely puzzling. It seems like nothing more than a cynical move to clear out old stocks. I can't imagine many people buying this, which will give Apple the reason to phase it out since "there isn't interest any more".
It isn't a deal breaker in the future either. Those PCI-e lanes for for "lower and moderate speed" I/O implemented with discrete chips. Here is a block layout for one of the new Thunderbolt desktop boards that Intel introduced that Anandtech looked at.
[Full article here covering a ASUS and Intel boards. http://www.anandtech.com/show/5884/...part-2-intels-dz77rek75-asus-p8z77v-premium/1 ]
Here the dual 1GbE , thunderbolt , and extra SATA controllers are hooked to the support chipset. A very similar approach could be used on the workstation Mac Pro models where USB 3.0 would replace that additional SATA controller since didn't have USB 3.0 in the core chipset. Also Firewire 800 wouldn't be hiding behind some x1 PCI-e switch at the bottom either.
Since the C600's the E5 models use also have Storage control unit (SCU) it prbably isn't a good idea to hook Thunderbolt to it to saturate the DMI link with traffic. But something like USB 3.0 would work just fine.
If the Tick/Tock cycle for Xeon class is moving to 4 years then the chipset is probably going to be put on same cycle as the CPUs. I don't think it is going to be quite 4.... but likely more than 2.5 .
But that is even all the more reason they can't skip adding something like USB 3.0 . If the motherboard is static for 2 years then can't really pass up something that is about to explode in growth now.
For the E5 2600 this is a complete non issue. There are 80 PCI-e v3.0 lanes. That is a plenty to hook motherboard embedded controllers to and still have enough left over to fully support 4 PCI-e slots without resorting to PCI-e lane oversubscription with PCI-e switches.
Even for the E5 1600 with a moderate amount of switching still can have two "clean" PCI-e non-switched sockets. ( e.g., in a Mac Pro like set up pair one x4 socket with embedded TB controller and another x16 socket with embedded GPU). If the GPU is integrated into the CPU again this is a trivial issue. On the current Mac Pro's the two x4 PCI-e slots are switched now and no one is really howling about it.
they didn't change the board this time either... so yeah no USB 3.0 right now.
That's flawed, IHMO. There is a place for TB on any system with a integrated GPU. It is immaterial if that system is a "box with slots" or a classic laptop case. A very large fraction of new desktops shipped at this point (and going forward) have integrated GPUs in them. The only question is how much does that growing segment get expanded by also adding designs that embedded a mobile GPU configuration onto the motherboard also (e.g., iMac like designs ). It doesn't really matter how the GPU gets attached to the motherboard just as long as it is effectively permanently attached (minus some corner case repair modes).
Desktops don't 'rule' the PC market anymore. Laptops and all-in-ones do. There is going to be additional functionality desktops can leverage (thereby providing value) if they can tap into the same products mix that laptops do without adding much to the complexity of the desktop. Thuderbolt isn't desperately needed at the top end desktop market but at the entry-mid points it is a value-add.
The mindset of GPUs are on PCI-e cards skews the vision of where the overall market is going. PCI-e card centric GPUs .... that's a business with strategic misalignment problems with the overall industry.
This isn't a niche long term. Your missing where the laptop/tablet market is going if you think this is going to be a minor niche.
Not just data. http://www.aja.com/en/products/t-tap/#/overview
Or at least "data" in terms usually associated with files.
This doesn't really make sense. Apple is almost printing money right now. Buckets coming in so fast they really don't know what to do with it all. $100B cash pile and counting. And they are trying to squeeze a few extra drops out of the Mac Pro because they are hard up for cash.
I don't think Apple's depth of greed is that deep. If Apple thought the Mac Pro market was too small and want to just walk away they could. It would even show up as a blimp on their balance sheet. Even with the Mac Pro's large margins. Not even a blimp. There are bigger swings driven by seasonal iPhone buying trends that would be more noticeable.
I seen beancounters too. But Apple is pulling in so many beans with the iPad and iPhone that they are quite busy just finding places to stuff all the money. I could see at a company where margins were challenged. At Oracle, HP, IBM ... something struggling and somewhat unaligned like the Mac Pro might be dead. (e.g., IBM chucked their PC division long ago and HP almost did the same thing. )
This presumes that they were actually working on it. As opposed to some internal faction managed to mothball the product line while a "deathmatch" wrestling contest continued as to whether to EOL or not. I've seen that too. And externally, the 'death match" game results look alot like what this Mac Pro update looks like.
TB is part of what I'm thinking about, but in terms of the C600, my primary thinking is based on discrete RAID controllers (Marvell, LSI, PMC) where the designer isn't willing or able to utilize PCIe lanes from the CPU's PCIe controller, particularly a 6.0Gb/s spec'd component.
Real potential for bottlenecking in this instance (DMI throttles to ~1320MB/s, but 4x SATA III ports can push ~2200MB/s sustained).
Of course that would be the case (no way I see them doing a new chipset for each half of the Tick-Tock cycle). I'm not sure it will be that quick of a turn-around now though (2.5 year for the entire cycle), given they don't have any real competition in this particular segment as well as increasing complexity of both the design and process.
Hence the horrible conclusion as per the time frame. I truly hope I'm wrong on this, but I've that sinking feeling it's realistic.
I agree.
Unfortunately, I'm accustomed to MBA's running businesses based on accounting with ever decreasing regard to other factors (just focus on quarterly profits), such as long-term planning.
In this case, sure.
But I see the SP as leading the sales figures vs. the DP systems. Corporate purchasing budgets are stretched ever thinly, so they're looking very closely at cost/performance (ROI) on their tech purchases. And Intel has listened IMHO, by delivering some very powerful SP CPUID's to fulfill that need. They're certainly not the performance leader, but they do offer a lot of performance for the money, and are suitable to a great many tasks in the workstation segment (majority).
DP systems will still have their niche, but I see that sub-segment becoming ever smaller. Clusters will also chip away at this group, as it allows a reduction in labor.
This is certainly a possibility.
Some have noticed it affects them negatively, but it's not the majority.
But I can see it being more of an issue in companies with extensive IT infrastructures (dual GPU's + FC controller would consume all 40 lanes; if they need a RAID card as well, something will have to run on a reduced lane count, assuming there is another slot available).
I worded it incorrectly. I meant as a primary high-speed interconnect, particularly on a workstation.
There will be a few that need it for sharing peripherals, such as importing camera footage for example.
Now this could change if Intel gets the optical portion out in the wild, and in particular, adds networking capabilities (a cheap optical network would definitely attract enterprise users). But I've not seen a time-line on this, and given the limitation of the optical modulators (cost over-runs IIRC), will take time to get the component cost down to the level it would need (i.e. graphene based, which still needs R&D to perfect).
I'm not saying laptops, AIO's, and devices don't. But I've kept my remarks to the workstation market, which is where the MP is situated.
TB in the consumer space absolutely has a place, and I expect it to be an ever increasing one. But not so much in the enterprise market unless it can be formed into an inexpensive optical network.
Apple is in a unique position ATM, but I suspect they figure if they don't pay the same attention to it as other companies, they're more likely to make very poor decisions that will severely injure their bottom line.
As per the MP being barely a blip, there's not a lot of R&D in it compared to other products, such as what they're still designing in-house (iPad for example). I also expect they would like to keep their customer base in this segment if at all possible, so they released an interim solution until a replacement product can be brought to market .
Given the direction I think this is going, it will rely heavily on TB (whether it's an AIO or headless xMac <or whatever you want to call it without a screen>). But such a product would no longer be a traditional workstation system.
Foxconn would have done the hardware, leaving Apple with some firmware and software development.
The "deathmatch" concept is possible, but it seems like it's the result of changing directions to me rather than the indecisiveness that would result in the scenario you've mentioned. For me at least, the "deathmatch" scenarios usually resulted in two teams creating a prototype that was reviewed by management. Once they made their decision, it was usually pretty close to finished (i.e. replacement product v. upgrading on the existing Intel cycle, not just new CPUID's on the existing production models).
Sure the c600 motherboard doesn't have built in TB or usb3 but it does include SATA III which at the very least would provide a connection that wouldn't bottleneck the currently fastest SSD drives.
Not to mention a faster memory bus as well as quad channel, right?
I can live without TB on a machine with four drive bays, but not with only SATA II. The only way to get full speed on SSD right now is to waste a PCI slot.
It's not that hard to tweak an existing intel board and shove it in an existing case, and even without TB or usb3 an update would sell more than this version and give pro users more faith that the product won't be killed. This sure looks like that one last update that could be done without spending a penny on R&D before killing the product.
Not to mention a faster memory bus as well as quad channel, right?
I can live without TB on a machine with four drive bays, but not with only SATA II. The only way to get full speed on SSD right now is to waste a PCI slot.
It's not that hard to tweak an existing intel board and shove it in an existing case, and even without TB or usb3 an update would sell more than this version and give pro users more faith that the product won't be killed. This sure looks like that one last update that could be done without spending a penny on R&D before killing the product.
After lots of thought
I'm in. But I'm not happy.
I'll tell you why I'm in.
I effectively have no MP under a desk at the moment (don't ask), there's 2 x 30" monitors on it, and it was specifically designed for them. There's no room for another, different, shiny iMac screen, and I need grunt.
Here's the cash bit; because this is a company purchase I can write down 40% of the net price back on tax and obviously the 20% of the gross through VAT. This is a hearty discount.
Effectively costing a very great deal less than a pro-sumer has to pay. This means the risk is reduced. Controversially I feel all this rubbish from iMac fans is driven by cash and I wish they would shut up, I'm sick of their subjective argument. A MP will effectively cost me less than they pay for an iMac, so **** 'em, if they were in my shoes power comes before money.
Oh, I feel better for that.
Why I'm not happy; I am a pro, power user. I have a infrastructure built around macs, so I'm stuck. I hate not being able to project and forecast my future. For me and my needs, Apple are not behaving in a professional manor. You MUST treat your professional market differently from your consumers. We're not fan boys or flippant consumers, we're running bloody companies here!
Anyway, that's why I'm pulling the trigger on a 12 core and I'm not happy about it.
That's all, carry on.
I'm in. But I'm not happy.
I'll tell you why I'm in.
I effectively have no MP under a desk at the moment (don't ask), there's 2 x 30" monitors on it, and it was specifically designed for them. There's no room for another, different, shiny iMac screen, and I need grunt.
Here's the cash bit; because this is a company purchase I can write down 40% of the net price back on tax and obviously the 20% of the gross through VAT. This is a hearty discount.
Effectively costing a very great deal less than a pro-sumer has to pay. This means the risk is reduced. Controversially I feel all this rubbish from iMac fans is driven by cash and I wish they would shut up, I'm sick of their subjective argument. A MP will effectively cost me less than they pay for an iMac, so **** 'em, if they were in my shoes power comes before money.
Oh, I feel better for that.
Why I'm not happy; I am a pro, power user. I have a infrastructure built around macs, so I'm stuck. I hate not being able to project and forecast my future. For me and my needs, Apple are not behaving in a professional manor. You MUST treat your professional market differently from your consumers. We're not fan boys or flippant consumers, we're running bloody companies here!
Anyway, that's why I'm pulling the trigger on a 12 core and I'm not happy about it.
That's all, carry on.
Same grabbing a 6 core and selling next year I guess....
Why a sad face? You'll have "the fastest Mac money can buy" for a full year. Chances are that this is the last of the true Apple workstations (I agree with Cindori's sentiment expressed in another thread), and there may be still be quite a few 1,1 through 3,1 holdouts next year who will be willing to shell out a pretty penny for your 2012 MP (esp. if you get Applecare) when that happens. If the XMac suits your needs - problem solved. If not - problem solved as well.
I like Akadmon's posts because while many of us are wondering what to do he actually did something.
It might not be what I'm going to do but it's a good point of view.
It might not be what I'm going to do but it's a good point of view.
I am as mystified as everyone else as to what happened yesterday. How could someone in Apple let this occur on their watch? it was a disaster, I can't help wonder why they didn't get this refresh out a few weeks ago so that yesterdays positive aspects didn't have a a bit of a whiff near it.
I pulled the trigger on the 3.33 GHz 6 core today, Im running CAD and Rendering on it, I couldn't bring myself to go for the 3 Ghz 12 core.
My current Mac Pro was a day one 3 Ghz Dual Core Intel 1.1, it never once missed a beat and has never let me down in 6 years, its getting a little tired and I had an opportunity to use a 2010 Mac Pro recently at it was a screamer compared to mine, and I really mean it, these things may read on paper with similar GHZ speeds and amount of cores, but I can tell you they are worlds apart, and so I felt comfortable ordering a hex core, the speed and difference for me will be worth it. These things as a tool for me pay for themselves in a few weeks, my last purchase in 2006 cost €5k and I run it 24/7, thats 2190 days ago, so it costs about €2.20 per day. Todays drop will pay for itself outright in a couple of weeks.
I also included 2 x 27" Cinema Displays, (which I note went EOL yesterday) an SSD, a 2TB HDD and Apple Care, because I was getting a great business discount pricing from Apple. I'll get 24GB of Crucial ram direct, So here's to the next 3 years. I'm really looking forward to setting it up.
I had a right laugh with the sales guy on the other end of the Apple Store Business Line about it. He was clearly embarrassed at the offering, I asked him could he give the box a wipe and he offered to blow any cobwebs before they deliver.
Some of us can get 6 years out of these things, and some of us get 2 years out of them, and hey look at Apple its getting 4 years out of them! its a fools gold. but it is what it is, and life goes on.
I can't wait to start using it, because the fact is, it will be new computer to me.
Having said all that, the big embarrassment happening under apples chin right now is this.
The "new" 3.3 GHZ 6 core with "6GB" ram is $2999
http://store.apple.com/us/browse/home/shop_mac/family/mac_pro
The "refurbished" 3.3 GHZ 6 core with "3GB" ram is $3149
http://store.apple.com/us/product/G0LF1LL/A
Its fair to assume here that the Mac Pro line has no one at the helm right now.
I pulled the trigger on the 3.33 GHz 6 core today, Im running CAD and Rendering on it, I couldn't bring myself to go for the 3 Ghz 12 core.
My current Mac Pro was a day one 3 Ghz Dual Core Intel 1.1, it never once missed a beat and has never let me down in 6 years, its getting a little tired and I had an opportunity to use a 2010 Mac Pro recently at it was a screamer compared to mine, and I really mean it, these things may read on paper with similar GHZ speeds and amount of cores, but I can tell you they are worlds apart, and so I felt comfortable ordering a hex core, the speed and difference for me will be worth it. These things as a tool for me pay for themselves in a few weeks, my last purchase in 2006 cost €5k and I run it 24/7, thats 2190 days ago, so it costs about €2.20 per day. Todays drop will pay for itself outright in a couple of weeks.
I also included 2 x 27" Cinema Displays, (which I note went EOL yesterday) an SSD, a 2TB HDD and Apple Care, because I was getting a great business discount pricing from Apple. I'll get 24GB of Crucial ram direct, So here's to the next 3 years. I'm really looking forward to setting it up.
I had a right laugh with the sales guy on the other end of the Apple Store Business Line about it. He was clearly embarrassed at the offering, I asked him could he give the box a wipe and he offered to blow any cobwebs before they deliver.
Some of us can get 6 years out of these things, and some of us get 2 years out of them, and hey look at Apple its getting 4 years out of them! its a fools gold. but it is what it is, and life goes on.
I can't wait to start using it, because the fact is, it will be new computer to me.
Having said all that, the big embarrassment happening under apples chin right now is this.
The "new" 3.3 GHZ 6 core with "6GB" ram is $2999
http://store.apple.com/us/browse/home/shop_mac/family/mac_pro
The "refurbished" 3.3 GHZ 6 core with "3GB" ram is $3149
http://store.apple.com/us/product/G0LF1LL/A
Its fair to assume here that the Mac Pro line has no one at the helm right now.
Last edited:
Speaking for myself, the reason I am not happy has been explained, was I not clear enough for you? Perhaps you should read my post. And all this blah about xMac, possible releases next year, etc is exactly why we're not bloody happy.
Sorry, I did not read your post. Was replying to someone else (in case you missed it). Some people -- always starved for attention
Sorry, I did not read your post. Was replying to someone else (in case you missed it). Some people -- always starved for attention
Explains a lot really, he was quoting me - which you didn't bother reading. If you are going to get into a pissing competition, remember to open your fly.
Last edited:
I have a quotation for this screen with the same part number and its price is discounted with the notation "EOL"
Last edited:
DMI has a limit on how much bandwidth is assigned to the SATA ports in order to keep the controller from utilizing the entire bandwidth.
So whether or not 6.0Gb/s SSD's would throttle depends on the configuration. For example, if you run 4x SSD's, each capable of say 500MB/s sustained, and configure them into a RAID0, then you'd be trying to push 2GB/s of data over a pipe that can handle ~1.32GB/s.
Memory bandwidth is quite good, so it's far less of a concern, particularly in workstation use. Think of it this way; any usage that requires disk I/O will be the bigger issue, as it's far slower than memory (SSD as well as mechanical).