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.
$2999 for entry model: what do you think?
- Thread starter badlydrawnboy
- Start date
- Sort by reaction score
But it seems to focus the Mac Pro to one single market - people who need excellent video performance. The thing that was great about the old one, was that it was versatile... single CPU, dual CPU? No problem. Lots of internal storage? No problem. Add in some PCIe Fiber Channel cards, or a more meaty GPU? No problem. It was easily expandable and adaptable to whatever role you needed from your high performance workstation. The nMP seems to be focussing itself on a specific market and is making it an extremely unattractive option for other markets - especially given that any serious external storage (without getting into stuff like iSCSI) will need to be through a Thunderbolt 2 disk array - ouch.
I'm kinda sad because I secretly hoped they'd be affordable. Even with my work's computer refresh coming up (I get a fairly generous budget) and the substantial education discount applied to nMP (20% off!) I still can't afford one.. and simply can't justify one for home at those prices. Damnit.
According to this video [ http://www.complex.com/tech/2013/10/expensive-mac-pro-fully-loaded ]
your education discount is worth $2,760 on a fully loaded nMP w/o the monitor(s) [20% of $13,800 = $2,760], leaving your final cost, w/o monitor(s), at only $11,040.
The post may have been in reference to the Mini, but realistically one will want to compare against an equivalant desktop processor when going against a Xeon. Hence the links I provided. Anything will really beat a mini. It's whether or not the TrashCan Pro's base configuration can meet or beat the equivalant i7. In this case, it does not. Close, but still beaten. As I said before, where an LGA2011 system is going to shine is in the greater number of PCIe lanes. That alone will give distinct advantages in SLI/CF setups, especially given that AMD is now moving away from CFBI to XDMA CF in their Hawaii lineup.
No arguments here. Depending on the type of workload, those "old" E5 Xeons can smoke any LGA1150 part. But one pays a hefty premium for those CPUs. A real hefty premium.
What I wish Intel had done was at least given the i7 K series the Iris Pro 5200 IGP. While not as good as a discrete card by any means, it's leaps and bounds above the HD 4600 the Haswell chips are saddled with, as I found out when trying Diablo 3 at native resolution of 1900x1200 even with everything on low - 18-23 FPS tops. Yech.
I've got my Hackwell Pro built and am just awaiting my EVGA 03G-P4-3784-KR GeForce GTX 780 to really rip things apart. Combined with my overkill fan setup, that card should yield some really good results. I'd have gone 290X or my original choice, the Toxic 280X from Sapphire, but a) Sapphire took too long and missed a golden opportunity to get those Toxics out on time and b) the nVidia price drop yesterday made a louder, thermally constrained 290X reference board not much of a deal, especially given nVidia cards work "out of the box" in a Hackie.
So now I will have something that utterly rips the iTube to shreds for non-compute tasks (i.e. gaming) while offering full expandability and upgradeability. It should get interesting when Broadwell comes out.
Am I blind or did Dell discontinue everything over 6 core? Their premium model has 6 cores running at 2.6 Ghz, has 32 GB memory, Quadro 4000 and a 7200rpm HDD, for 3400$. Apple's 4 core offering probably is faster than this and has a SSD instead, pimp that to 32 GB and you'll get a similar price to 3400$ and for that you get a much better computer than what Dell is selling.
You just can't customise the workstation choices that will use the new CPUs online at the moment. They aren't shipping until the end of November. They are in limited supply and Dell's server market is more important which is where you will find these CPUs at the moment.
If you search for the spec sheets you can see they are just very minor refreshes of their current systems.
yea judging by what some people are saying some of the cpus and gpus cost .. it seems like the 8 core will be an extra 1200 .. the 12 core an extra 2500 and the d700s possible another 3k
Maybe more. If they're similar to the W9000 (according to AMD/Apple), that's $3500 each--$7000 total. My guess is 4 grand for D700 over D300.
Of course, if Apple just has them identify as 7970 in Windows, they could charge very little
Candidly, I think Apple knows exactly what it is doing. It is creating a new product for a niche group that can afford its offering. My gripe is the RAM configuration as one needs to pay for RAM that they may not want to use if they go third party.
I suppose some of us will want the present and past Mac Pros and of course the Mac Mini instead of these new machines. Candidly, I think the bottom of the line Mac (mini) Pro is a bit over priced for what one gets. It should have gone for around 2400 dollars not 3000 dollars. - It is not a very impressive item as a stock item. - Going hex at least gets one in the park.
Last - nothing like having to go through another wait for Thunderbolt 2 items. Look how long it took for TB1 to catch on and how expensive the devices and cable were. This is another rather pathetic exercise in emptying our wallets.
I suppose some of us will want the present and past Mac Pros and of course the Mac Mini instead of these new machines. Candidly, I think the bottom of the line Mac (mini) Pro is a bit over priced for what one gets. It should have gone for around 2400 dollars not 3000 dollars. - It is not a very impressive item as a stock item. - Going hex at least gets one in the park.
Last - nothing like having to go through another wait for Thunderbolt 2 items. Look how long it took for TB1 to catch on and how expensive the devices and cable were. This is another rather pathetic exercise in emptying our wallets.
It looks like Thunderbolt 2 will most likely just keep prices where they are, and some companies are even offering free upgrades if you buy a Thunderbolt 1 device just now, but then at current prices you have to figure you're probably paying for both devices anyway.
So Thunderbolt 2 shouldn't make things more expensive, but they'll likely just replace Thunderbolt 1 offerings and prevent the price from going down for a while. It also means Thunderbolt 1 items may not be around at lower prices at the same time; there'll more likely just be a brief window where existing stock can be bought up for less and then everything will be Thunderbolt 2.
This is why I'm probably just going to make the best of USB3 that I can using a custom storage case, so that I can hopefully switch it to Thunderbolt later.
The benchmarks aren't out yet, but looking at results from similar processors that is incorrect. The i7 27" iMac will actually score about the same if not a bit better in most benchmark tests in pretty much everything other than those that take advance of the dual GPU's than the entry nMP.
entry model
It looks like Thunderbolt 2 will most likely just keep prices where they are, and some companies are even offering free upgrades if you buy a Thunderbolt 1 device just now, but then at current prices you have to figure you're probably paying for both devices anyway.
It looks like Thunderbolt 2 will most likely just keep prices where they are, and some companies are even offering free upgrades if you buy a Thunderbolt 1 device just now, but then at current prices you have to figure you're probably paying for both devices anyway.
It should take no where near as long. Two of the major hangups in TBv1 roll out were
A. Vendors had never done it before. Never did a TB certification process. For many never had to deal with DisplayPort. (TB's backward compatibilty mode requires this if have two ports, one of which can to put into "pass through" mode for backwards compatibility).
B. related to the first but on the different dimension: Software. Most drivers needs some small adjusting. Some folks were waiting on Apple to fix bugs/glitches and to provide basic foundation support for stuff not included with Macs.
This second pass the second should have little to no impact. TBv2 doesn't really require anything new. New components on devices may need new drivers but just moving current device forward isn't going to require any new software.
There are some new DisplayPort v1.2 twists but still the basic framework has been laid out.
The clone/chop shop vendors have had plenty of time to study the working systems other folks have done (copying other folks homework always goes faster) .... so if they can get past certification should also have more vendors.
For folks that have TBv1 devices for vast majority there is no need at all of rush out and buy replacements that for the most part do the exactly same function. There likely won't even be TBv2 versions of stuff like the TB->Ethernet and TB-> FW dongles.
A) vendors were afraid to get into the TB mix given that it was an unknown as far as demand and also a great lesson learned from Firewire where it was practically abandoned by Apple who could have provided the first devices and promotion.
B) Apple has a fine reputation of getting some things wonderfully right and some thing terribly wrong or rather "crippled."
C) New hardware and device drivers are needed for TB2. That is a non-issue as far as being a fact. Again, it will require a cost and a market and makers willing to invest in product creation. TB2 has the potential to be as one put it well "a slightly slow extension of the on board add ons."
I certainly hope that TB2 catches on and that prices are not as ridiculous as the early 2 years of TB devices.
As I just said in another thread, perhaps TB drive enclosures in particular, will be stuck in an expensive niche for the foreseeable future. What if everyone's waiting for cheap enclosures and cheap enclosures can only become a reality if people buy? Perhaps TB will be relegated to specialty peripherals for video, audio, SAN, 10GigE, etc.
Let's face it, USB 3 is good enough for single drives and even dual drive RAID0 arrays, so the only place TB really makes sense for storage is among home enthusiasts or SOHO users who want external SSDs or large RAID arrays. Most organizations with more than one person are going to have networked storage and won't have a need for TB enclosures as a general rule.
I have to agree with you on this one. USB3 (and even TB1) is plenty for the typical user leaving TB2 for niche set ups and certain professionals.
While Apple got the TB2 in time for the new Mac (mini) Pro, we'll soon see (maybe mid next year) the next round of USB3 which is substantially faster than the present version. Somehow, with Apple (grin) we seem to always get caught inbetween.
Firewire was not abandoned by Apple in the initial FW start up years. Apple flipped flopped on licensing fees. That spooked tons of folks and pissed off Intel most importantly among others.
The vast majority of the PC market vendors are afraid of anything new with a different price/performance ratio. That's why PC innovation is sliding back toward glacier pace. If it isn't cheaper then most sit and wait someone else to work out the kinks. When mature they might jump in.
TB going from v1 to v2 is actually a sign it is maturing. Apple is pouring 10s of millions of dollars per year into Thunderbolt. ( 10M Macs * $10 => $100M ) Money talks. Other system vendors holding their breath ( money) and stomping their feet over cross licensing is not going to have much traction when some one else is saying they will keep dumping of millions each for the next several years.
There are more than a few things that Apple has done in that first category that are initially panned by folks primarily looking the rear view mirror as being the second.
At system/device boot and firmware level yes, but that is coupled to the system/device. In terms of PCIe device drivers, for perhaps unimplemented optional PCIe standard's hotplug requirements it is exactly the same. The vast majority of devices/drivers can't see Thunderbolt. So TB going from v1 to v2 is completely opaque to them.
.
A new controller that has a new driver? That aspect of new is independent of TB v1 or v2. The stumbling block during v1 was far more so on drivers and bug fixes that crossed organizations in terms of responsibility. If everything else is kept largely constant and just moving same device from TB v1 to TB v2 at the controller level there isn't much change that ripples into other components the device/system builder doesn't control.
Nothing is free. For vendors who did a TBv1 product, it is also as much generating a higher return on what you have already invested as much as any new amount. Have already built the skillset, might as well spread the cost of doing that over a larger set of devices.
Chuckle. Pragmatically TBv2 will enable external USB 3.1 devices; at least the likely 1st general USB 3.1 discrete controllers. If want to throw those into the slightly slow board add on category go right ahead.
I think TB will catch on better when large groups of folks stop characterizing it as something it is not. It isn't a USB 'killer'. It isn't the 'one socket to rule them all'. It isn't a PCIe 'killer'.
Prices aren't going to crater. There won't be a race-to-the-bottom on pricing. But bigger drops probably going to come until TBv2 has been out in volume for a year and Intel isn't invoking an Obsorne effect with a TBv3 that is radically different in the next year. Every 2-3 years is going to spook most peripheral vendors.
TB is going to be stuck if it isn't largely applied to aggregation problems. Mimicking USB drives and dongles is problematical because it is largely a misapplication of the solution to a problem.
TB isn't just good for external drives or relatively extremely low volume external connections.
The difference of TBv1 versus TBv2 of something like
http://www.sonnettech.com/product/echo15thunderboltdock.html
could be a TBv2 variant with more drive sleds that aren't throttled. It isn't 'just a drive'.
Doesn't particularly have to be raid. larger JBOD collection also. The set of storage devices don't have to organized into a coherent set. Just the number of devices is larger than 1-2.
This is again largely trapped in the "it is just a USB killer" rathole. Doesn't particularly have to be USB replacement. If actually need more USB sockets Thunderbolt offers more bandwidth than diluting bandwidth with an external USB hub. Again aggregating multiple USB sockets down to one wire back to host system is a better fit then zeroing in on being a "killer".
Flawed in two ways. First , scarily enough, most organizations are sharing files through email and sneakernet (portable drives, ODDs, thumb drives , etc ) than centralized storage. Larger oranizations typically move to centralized storage because that gets choatic after get to more than several people, but most organizations are small. There are tons more mom-and-pop businesses than even medium sized ones.
Second, in order to get to network storage that reasonably faster than a single HDD if have more than a few folks and/or larger files you'll need a TB device on most Macs to get to that network. For example, Promise SanLink2:
http://www.promise.com/promotion_page/promotion_page.aspx?region=en-global&rsn=103
Not particularly a DAS (direct attached storage ) drive enclosure, but it is a TB device.
So the MBAir user who hooks their TB docking station/monitor is a niche set up? While it isn't mainstream that also has a bit to do with perhaps they don't need a $999 27" screen. A 21.5" screen still would be substantially bigger than the Mac laptop screen and far more affordable. Thunderbolt isn't the primary driver of that $999 price point. ( the old mini displayPort version at the exact same price point was quite illustrative of that. )
Time will tell but USB 3.1 seems more likely to fail at being a "USB 3.0 killer" than Thunderbolt did/will be. For classic USB devices it isn't going to offer much. Aggregating traffic is far easier path to build up enough to fill up that amount of bandwidth but USB is an backwards approach to that. (e.g., a PCIs/SATA Express SSD drive ...... why map that into USB protocol space to get back to a host???? )
USB 3.1 will run into many, if not more, brick walls that USB 3.0 ran into trying to displace USB 2.0. It isn't in the chipset so it will be more expensive. So USB 3.0 and 2.0 devices are going to be cheaper and even cheaper respectively.
USB 3.1 and Thunderbolt overlap. The competitive overlap will help drive down costs for both, but
If folks want it, TBv2 can deliver USB 3.1. Apple just going to ignore the 2 (maybe 3 ) year ramp up that USB 3.1 takes to perhaps get weaved into the core chipset and volume adoption. As much as the USBIF folks continue to primarily market USB 3.1 as a TB "killer" it will get the cold shoulder before chipset adoption.
I have to disagree - Apple is spending money on R&D, but hardly on components . The customer pays for those .
Also, TB has been included with many but the most important Mac (for that sort of technology) - the MP .
That's changing only now, a few years after TB's introduction ; how many MacBook and iMac owners might have gone further than TB dongles and display adapters to use their TB ports ?
That's not a clientele spending a premium on technology which yet has to proof its benefits, and justify its pricing .
Apple hasn't really pushed TB until now, I think , and I don't understand their approach .
They don't do peripherals, and certainly don't want to support them across platforms, but how else could Apple widen the appeal of TB ?
As for pricing (sorry, dead horse), where is the point in introducing a data transfer technology that is much more expensive than FW, eSATA and USB, and make its use quasi mandatory even for mainstream MP owners ?
Let's face it, the nMP only makes sense if one embraces TB - performance critical data has to go external, obviously, and the 4 USB ports will not cover that, even if they were fast enough .
That multibay HDD enclosure will need to be well away from the desk, else the fan will drive you bonkers - that's 200+ bucks for the cable alone .
I understand external storage, mine is external .
But why should I pay top money just to have a scratch disk and a project files disk ?
Customers don't pay Intel. After they have paid Apple for their Mac it is Apple's money; not the customers. Umpa Lumpas don't produce the money that pays for Apple R&D either. Short of the couple $100M or so Apple gets on interest on their cash hoard, that was all formerly customer money too.
With Thunderbolt deployed across the entire Mac line up, then when a Mac is bought money is going to Intel. Not just for the CPU and chipset but for TB also. Just like now with iWork bundled every Mac bought, every Mac purchases contributes to iWork . Something for the rest of the apps bundled with the OS and the OS itself.
In terms of revenue Intel received for TB controllers, that really is immaterial. It has been paid. How many customers have opened TextEdit or the calculator app ? Bundled things that were paid for too.
$10 is a premium? More than likely Mac customers have contributed several multiples of that to just making Apple cash pile taller. Taller just to be taller.
There are laptop uses who never plug anything into their socket for an external monitor but may use some other feature more heavily. There are other users that use the external monitor all the time and never use that other users feature thing. Both pay for the bundled and both get the bundle slightly cheaper.
With Apple approach to minimalistic sockets Thunderbolt doesn't really cost much in terms of edge space if don't use it as Thunderbolt. Some form of video out was likely going to there because it is useful for a sizable enough subset of the population. Thunderbolt piggybacks on that to expand the set of users who can get added value out of that socket.
Haven't pushed it? The chart that includes ExpressCard being stomped by Thunderbolt isn't sprinkled around Apple's website? Widen appeal? A customer base of 20M host system devices to sell into. Peripherals tend to get trapped in a "chicken or the egg" dilemma. Peripheral vendors don't want to do product until there is someone to sell to ( the peripheral usually can't work by itself and at least has to connect sometimes. ) and the system vendors don't want to jump in with both feet until there are devices.
Apples only does a very limited set of peripherals. That isn't specific to Thunderbolt. USB devices? Highly limited. Displays? None unless you count the Thunderbolt docking station display. Printers? none. etc. If these peripherals connect with standard connectors that it is 3rd party issue to make work on both OS X and Windows ( and Linux and ... whatever else). Apple's position is to set up the ecosystem. Not occupy all of the roles.
How could Apple widen TB appeal?
1. They could have gotten the bugs out of their system software faster. Part of the hang up with stuff that straddle between OS X , driver work of Apple, and driver work of 3rd parties. Apple isn't totally at fault but they certainly were a participant. ( This is part of Apple's chronic ''we didn't have enough developer resources to work on that" issue that is largely because they don't develop and hire the talent they need. )
2. Apple doesn't necessarly have to make the peripherals but it wouldn't hurt to do some relatively minor (for them) speculative investment in jumpstarting some 3rd parties. Apple will buy equipment so that some Chinese manufacturer can pump out units with low wage workers but do almost nothing to defray capital costs to doing the R&D to get a new TB device off the ground. Apple has have to had some ideas about what they perhaps do with TB devices that got rejected. Find a vendor who think might be more interested and generally ask what they are thinking of doing with TB. If there is a match throw some kickstarter money on it ( as loan , equity investment , etc. ). Just don't stare at the giant stack of money. Put a very small amount of it to work to work on something other than hedge fund hocus pocus. Sure some of those products might go bust and might loose some of the money, but frankly that money would be better spent on that than on lining Carl Ichan's pockets.
3. Apple has to lean on Intel to make sure they are opening access to vendors at a reasonable rate. They should just sit down with Intel to haggle over volume TB controller prices or request new TB features. Apple should be evaluating Intel's stewardship of Thunderbolt and making it clear that is an Apple concern. Right now things are highly skewed. Nobody but Apple buys at the same level of volumes. I'd bet even if aggregated the next 4 volume purchasers, that group doesn't buy at Apple volume levels either.
Because it is faster with lower latencies and can travel longer distances over reasonable expensive fiber. Why introduce PCIe SSDs when there are standard 2.5" SATA drives that are SSDs. Why develop Infiniband whether there is Ethernet. Different tools for different jobs.
It isn't necessarily a 'killer' of any of those. However, FW has hit a brick wall adoption. Essentially it is completely stagnant in general the PC market and in most of the electronic device market. eSATA is only good for doing SATA stuff. It can only get faster as SATA gets faster. USB is still trying to unwind from a legacy pf being attached to glitchy latency problems , overhead issue as increase throughput , and saddled with the inertia of all of the deployed legacy devices.
Thunderbolt does more and costs more. None of those other protocols are universally useful either.
So some groups of folks it already is external. It is not a "push" at all. It is actually following where those folks have been going over the last several years.
And there would be a non performance critical storage device in Mac Pro if Apple though that all performance critical data had to go external. It isn't.
Bulk data is certainly external. Relatively reasonable sized fast data isn't.
Not necessarily. The previous Mac Pros had 5-6 (or more ) fans and it didn't need a 200 cable to put distance between it and most users. Are there some noisy implementations out there now? Yes but there is not particular to Thunderbolt that is making them noisy.
If those are "top money" storage devices (plural) there is typically a difference with Thunderbolt over FW , USB , and mainstream eSATA.
Either can aggregated several 'mainstream' performance data streams or can target a small, but significantly faster set of devices whose steams aggregate to higher amount. For both of those increased performance that TB delivers has value. If performance doesn't matter then one of the cheaper solutions will suffice. Neither group is good for every problem.
Older Mac Pros have much nicer quality fans then most external units and now that they are required by the nMP what good does all that whisper quiet technology in the trash can do you if you still have a hurricane next to it?