The Mac Pro Review
anandtech.com
anandtech.com
Not the only place. I have found the German site NotebookCheck.net to be more thorough than Anandtech when it comes to laptop reviews.
Not only do they measure color space and color accuracy, but they also measure backlighting levels across the screen, comment on backlight bleed, test for viewing angles, check the reflectiveness of screens, etc.
In addition, they also test decibel levels of the fan, temperature of the laptop surface, etc. at various loads. They point out situations where a certain laptop may ship in two configurations. They point out the pros/cons of the nearest products from competing vendors, which have also been tested with equal thoroughness.
In other words, they actually test the machines against a checklist, to assess the performance of each element. In contrast, when an Anandtech reviewer claims that a machine is quiet or that the screen is matte, you don't actually get all the information. Maybe the ventilation system was masking some of the sound. Maybe the screen is only semi-matte.
The only problem is that NotebookCheck is based in Germany, so some of the machines they review are not available in the United States. Still, I'm happy restricting my choices a bit to avoid being surprised with a laptop purchase.
Which, instantly, should be reason enough to understand that said criticism is BS.
Apple products are among the best in the industry, period. Not just from the industrial design part of it, but overall: coherence of product vision, attention to important characteristics for the target market (battery time, portability, weight), quality machining and materials, attention to small details (from multitouch touchpad to magsafe adaptor and from backlit keyboard to magnetic, non protruding, lid hinge).
These people think that because they are not speced and designed like gaming PCs they are not worthy ("I can have a better GPU for less money in my custom box, and with xeon lights on the sides too).
And they attribute their popularity to some BS "reality distortion" effect, ignoring the fact that hardcore hackers, prominent programmers and old school neckerbeards, from Rob Pike, DHH, and Duncan Davidson to Jamie Jawinsky and Miguel De Icaza (the frigging founder of the Gnome desktop) down to Linus Torvalds, who waxes poetically about his MacBook Air as the best in the market.
So, "fawning over Apple" justs translates to "did some favorable reviews of products, instead of making up BS reasons to dislike them".
http://commandcenter.blogspot.com/2012/09/thank-you-apple.ht...
Linus Torvalds is said to be using the ChromeBook Pixel as his primary machine.
http://www.geek.com/chips/linus-torvalds-is-making-the-chrom...
As for Miguel De Icaza, he's eating his own dog food, as he now makes his money selling an IDE for iOS development, so of course he's going to choose the Mac because that's where the toolchain is.
Half of it it's about how it's Apple's fault that he didn't have a USB to boot off, so he tried to boot of a (camera) CF card (unsupported) and then an old Mac that wasn't up to running the latest OS version. Because Apple should consider what olders machines an OS supports not on hardware specs needed by the new OS, but on the needs of people upgrading their OS without a USB that want to use their older machine as a firewire devices (huh?). It's like the kind of complaints you read on Tripadvisor ("the bell boy didn't smile enough to me", "the bedsheets where not the exact Pantone blue they had on the hotel website" etc).
>http://www.geek.com/chips/linus-torvalds-is-making-the-chrom...
Yes, but he said a MBA just until then, of which he writes on his Google+ page. Also, if you read the article, what he solely likes about the Pixel is the screen resolution. And he dislikes its weight. He looks like a perfect candidate for the inevitable retina MBA.
>As for Miguel De Icaza, he's eating his own dog food, as he now makes his money selling an IDE for iOS development, so of course he's going to choose the Mac because that's where the toolchain is.*
Well, it's not just that. He also wrote a post about why he moved to OS X, and how he got dissillusioned with the Linux desktop prospects.
EDIT: rsynnott, below, has an excellent point.
I've had other parts crap on me though, e.g a battery after 2-3 years of use that had to be replaced. Also had an iMac (sold now), which had a faulty DVD (also replaced).
The thing is, those things happen to ALL production runs, there are some % of defective units. You can be Apple, IBM, Dell or BMW, and you still get this. I've had "upmarket" IBM hard disks die on me for example ( http://en.wikipedia.org/wiki/HGST_Deskstar ).
And, for tens of millions of machines sold, you only get to read about the far fewer faulty ones on such problem forums (well, duh!) -- so it's not much to get an accurate picture on.
I like to do extensive research before I buy something (I buy lots of tech gear, from stuff like DSLRs to Audio interfaces), and if I gave much promimence to the occasional forum complaints, I wouldn't have bought anything at all, because there are always people that have issues with any product you search. I prefer to stick to reviews, seeing units in action from friends and in the store, etc. Case in point, my latest buy, a Focusrite Scarlett interface. Pages of complaints about strange audio glitches with Mountain Lion / iMacs etc in audio forums. Have been working 100% fine for me.
Even their known defects make Apple come out smelling like roses. Contrast that with obvious design defects in other, cheaper PC laptops, which get ignored or refused at the 3rd party retailers they come from, and it makes Apple a pretty simple recommendation for power users and casual users both. Nobody I've recommended Apple laptops to has been disappointed.
By "you", I don't mean you, yapcguy. Just in general towards those that are criticizing Anandtech.
They compared it's price to a few workstations to show that the price isn't unreasonable for similar hardware (i.e. Apple isn't adding a $1200 workstation tax).
For the rest of the review they compared it to other Macs, because chances are that's what buyers are going to compare it with. If you want a Mac, those are your choices. I really doubt too many people who are in the market for an HP or Dell workstation are going to consider a Mac Pro.
Plus there is the problem of benchmarks. The OS can make a big difference, so you'd either have to run every benchmark twice on each system (once on OS X, once on Windows or Linux), and then the non-Macs can't run OS X. It would be a ton of extra work, but I'm not sure how much gain it would give.
Again, I think the number of workstation shoppers who will consider this machine is small. I expect the vast majority of it's sales will be to Mac users who want something more powerful than an iMac or a MacBook Pro.
Especially with the gap in Mac Pro releases over the past few years, I have seen many people deciding between Mac Pro and a custom built hackintosh. For the stuff that really justifies a Mac Pro, these were the only two options for some time (short of moving off of an OSX stack). Sure, someone custom building a multi thousand dollar work station is not your typical consumer, but neither is your typical buyer of a spec'd out Mac Pro.
Anandtech fawns over a lot of products, but I suspect perhaps this is due to them reviewing products they are interested in, and not wasting time on stuff they aren't. I'd start to worry if people can't reproduce their lab numbers but at the moment, they're considered the most detailed and accurate of the lot.
Unless someone can actually call their conclusions and numbers into question using actual facts, then it's just anti-Apple whining, whether it's him or "some readers".
http://forums.anandtech.com/showthread.php?t=2270612
Pretty much every review of Apple gear has people moaning about bias in the comments.
I'm not saying there is any favoritism, but certainly the perception amongst Anandtech readers is that there might be.
At the end of the day, there's no reason why the writers at Anandtech would be any more immune to access journalism than other folk who have tried and failed.
You are actually saying there is favoritism. It's dishonest to claim otherwise. You've just suggested that Anand has succumbed to access journalism.
Why not just be honest about what you think, rather pretending to be disinterested reporting on the views of others?
Let's also note that the commenter accusing AnandTech of bias in that comment thread you linked is resoundingly and overwhelmingly rebutted by other readers.
Those kinds of complaints can be seen anywhere there is a positive review of any product. If someone likes Google Glass, four hundred comments by people calling you a glasshole and telling you that you must like Google stealing all your data. Positive reviews of the Samsung Galaxy Note got huge criticism among the Apple sphere, everyone agog over how anyone could possibly like something so big (must be payola). And on and on and on.
Apple gives special treatment to outlets and bloggers who treat them "well" (and punish those who don't). It's all the same inducement, and it's all just as grubby.
Having early access to products and personnel for Apple reviews is a huge coup for media outlets: They need Apple far more than Apple needs them (if Anandtech didn't get sent a day 0 $7000 review unit, AnandTech would have gone without a lot of views and a front-page of HN and many other sites. Apple would have lost nothing -- the target market for the Pro would have just read the reviews elsewhere, and are at no risk of buying a competing workstation).
Apple hand-selects who gets these early access units, and thus who gets the attention of an early review. There is a strong incentive for those reviews to gently understate negatives and to overstate positives, remaining on the list for the next go around. In this case a workstation that prioritizes the irrelevant (box size is never even a discussion point when talking about high power workstation, but suddenly it's the primary design point?), and has some astounding faults that most other companies would be eviscerated for: The dongle approach of expansion; Power draw that at times exceeds the power supply rating; Chipsets going to 100C+ because the "thermal prioritized" design was actually "being novel small" designed; Performance that even in CPU+GPU scenarios only marginally improves on the performance of a box from four years ago?
I would posit that had this box carried a Lenova or HP tag on it, the reviews would be extremely negative, if not mocking.
This is not unique to Apple: Exactly the same thing has happened over the years with various industry or namespace leaders, reviews veering towards the positive to assure that you get to the front of the list for the next wave, in a perpetual cycle. There was a time when getting early access to Microsoft inspired a whole industry of fawning and soft-gloving.
Bizarre that gress is so desperately trying to present the notion that I'm somehow defending Samsung and their pathetic attempts at astroturfing support (though such is the entire business model of the PR industry, which every business engages in, so pretending it's so unique is delightfully naive). Their tactics seem very trollish so I'll simply ignore their nonsense.
Samsung has been convicted of paying shills to make false forum postings and reviews.
You are accusing apple of choosing who to give review units and press invites to based on who they prefer.
Every company does this. What else do you expect them to do? Provide a review unit to every blogger who asks for one? If you claim their behavior is underhand, you should be able to explain an alternative.
Comparing giving out review units to paying for shills is plainly absurd.
As to your comments about the review itself, they certainly reveal things that you would like to complain about, but none of these things were whitewashed or concealed by AnandTech, in fact they were fully exposed in statements, charts, and numbers.
Your chief complaint seems to be that you would have liked the review to have a vitriolic tone.
Well, yes, but any time just about anyone anywhere reviews an Apple thing and doesn't say "This is worse than iHitler", there are cries of bias, evil conspiracy, etc. People are a bit funny about Apple.
However, according to this review, the new Mac Pro doesn't work with the new Dell 4K monitors (I don't consider 30Hz refresh as 'working'), and even with the 4K display that Apple sells, it only works at its native 3840 x 2160 at 60Hz. When choosing a 'Scaled' resolution, it renders blurry junk.
That is pretty disappointing (although I imagine it will be fixed at some point).
And that's probably generous - from the GPU analysis appears that the tested unit has D700's which bumps the price to $8299 - a configuration that isn't mentioned anywhere in the article. About the only thing left to upgrade on the test unit is the RAM to 64gb.
Since the article calls itself a review, it would be better if the review unit was accurately described. It seems to me there's a bit of bait and switch because the performance numbers presented are not for the $3000 or $4000 presented in the article's lead.
OTOH, it's a little surprising that Apple gave out 12 core review units, since the 8 core seems to be better at benchmarking.
In light of the fact that much less expensive machines often performed better, it seems obvious that the favorable conclusion of the review ought to provide some rational case for being so.
It's a bold bet on a possible trend, which I like a lot, since it would mean that non-gamers would profit from the GPUs that would otherwise bore themselves to death on their machines. Also, it would give AMD a better position, maybe averting x86 being a complete Intel monopoly.
For more conventional, but still pro workloads, most of us will be much better off with a $3000-4500 model.
It's going to be interesting to see how pro apps end up tailored for this architecture.
Or do you mean something more seamless and auto-magical?
OpenCL uses a special programming model so you can't use it for general application code. It's good for doing repetitive operations on large arrays - e.g. Image or signal processing, or machine learning. OpenCL code will run in the CPU if there is no GPU or if the overhead of shipping the data to the GPU is too high.
The line of video professionals happy to pay $10k/box to get their renders done faster screaming "TAKE MY MONEEY NAOW" might have something to do with it...
Also, since you missed all the charts showing the MP demolishing everything else by 2.5x+ on multithreaded workloads (ya know, the thing that people buy MPs for) you may want to verify your consumption of what the kids call "h8rade".
The MacOS market in general might be described niche, but it seems to have sustained itself over the years.
three-year Mon-Fri 8-5 next business day, parts, labor and 24x7 phone support,
They come to me.
Edit: to all the naysayers: I'm in the UK. HP here is pretty good. We have over 200 machines on next day and we've had only two (!) problems and they were relating to part supply resulting in a quick purchase on Misco that arrived next day.
I ended up torrenting an HP restore ISO and thankfully the BIOS's licensing key matched the ISO.
On desktops though, I'm willing to put more effort into settling software update issues/device conflicts, since I probably have to do that anyway to write performance-optimized code (depends on the exact purpose of the desktop though, but I do a lot of scientific computing). So a Linux/Windows split boot on a generic PC usually wins out. I used to do a lot of PC gaming, but that's really less of a factor now.
As the AnandTech review points out, you're looking at $2700ish if you opt for hardware that is apples to apples comparison... regardless of whether you went with hackintosh, Linux, or Windows.
I code with a focus on TDD and small unit tests, so even though I do mostly statistical computing work, my development tests use small amounts of data and low computing power. When I actually need to run big production stuff, it goes off to the Linux box or a Linux server cluster.
My beef with the imac is that it is not really that much faster than my MBPr, but has terrible heat management. When I am running big jobs on it, it gets a few degrees off from a toaster and heats up my office. A 2.5k linux box would not have this problem and would be way faster.
If I had to guess, it probably has mostly to do with the imac sitting closer to me while typing, and also that I tend towards using it for more graphics heavy tasks (since it has the largest screen). Also, I am in the southeast, so heat changes are unusually noticeable, especially in the summer.
You are right though, the imac probably is as efficient or even more efficient than my other machines, however it is more noticeable when the heat is in your face instead of blowing out off next to a baseboard.
Apple, your stuff is mostly nice. Fix your product warranty and maybe I'll open my wallet.
When you're spending $3k on a computer, I don't see how $250 will really affect the bottom line...
payoff * Probability of payoff = EV
3000 * Probability of payoff = 250
3000 * P / 250 = 250/250
3000 / 250 * P = 1
12 * P = 1
∴ P = 1 / 12
You require 8.3% of the computers to fail in years 2 and 3, and for the failure to cause $3000 worth of 'damage' avoidance, and for it to be something covered under the warranty terms, for your gamble to make sense.
Judging by the fact they offer the warranty, I would imagine the fail rate to be somewhat lower than this, or for the failures to be vastly less expensive than that - else they'd be losing money.
It's a necessary part of the package.
So your protection is really only 6 months long.
Older workstations these days are a lot more affordable and can easily be upgraded. Most MACS you're stuck with what you get.
Case in point, I just purchased an HP 8400 workstation for a friend. $320 for a dual proc 2.6GHz quad core Xeon, 16GB RAM, Two 320GB SAS drives in RAID config and ATI Fire V7350 1GB video card. Sure its a pig and isn't the quietest PC in the room, but it completely shreds anything I could find in a retail setting.
If this was true, it would really explain HP's financial situation /s
EDIT: that graphics card despite being old, still sells for $300 on Amazon: http://www.amazon.com/ATI-100-505143-FireGL-512-bit-Express/...
Note that the D700 specs match closely with the R9 280, 280X, or 290, cards which sell retail for $349 to $449 or so each (the Mac Pro has 2 of these in the 2x D700 config). See https://en.wikipedia.org/wiki/AMD_Radeon_Rx_200_Series . The 2048 shaders would match with the 2048 number for the 280X, I think (if I am reading the chart right).
The W7000 is the much more expensive "pro" version which has ECC RAM on the card and much less volume in terms of sales.
It's funny though, I remember as a kid, thinking how cool it was that you could have a backpack-able computer (the original Macintosh could be ordered with a padded backpack). Now at 11 lbs and not very large, you could almost do the same again!
He's basically dumbfounded by the current situation :-)
I haven't found any since the 10.9.1 update. They fixed the quicklook slowness bug which was my only complaint. Mavericks is one of the most polished OS X release I have used, behind Snow Leopard.
Quicklook is still slower compared to Mountain Lion for me, so I hope they make it faster still.
One has to wonder what all their engineers are actually working on.
1) my other comment https://news.ycombinator.com/item?id=6993850
Anybody running Linux on Macbooks or Mac Pro? Does it work well?
Techcrunch interview here: http://techcrunch.com/2012/04/19/an-interview-with-millenium...
*edited to add link
Which is complicated when you are trying to build hardware and realize "oh crap, the standard thermal layout leaves no room for us to actively cool <insert-x-part>, how are we gonna keep this thing from overheating?
The way we have our computers laid out today - vertical pcie cards, cpu sockets with either the amd snap-brace or intel screw in backplate heatsinks - is entirely arbitrary, but for posterity it persists as nobody wants to be the guy to throw out 10+ years of expansion card compatibility.
My PC had the same price three years ago, but it has 5GHz CPU, 32GB RAM, 1TB on SSDs and 5 TB on spin HDDs.
I remember when they said people wouldn't need ADB or SCSI. Or floppies. Or serial ports. It worked out.
If your 5 GHz CPU a Xeon with ECC? The RAM can be upgraded (those are options), the SSD can be upgraded (1 TB is an option) and it's faster than SATA, and you can attach as many disks as you want in Thunderbolt arrays. Did you have dual GPUs for that price?
It's a nice looking machine, very quiet, and very innovative. Maybe it will be a misstep, but I'm glad Apple is trying something interesting. I want to see what happens with this.
I am more inclined to say Apple doesn't want people to think they need them.
This may be another such case.
http://www.promise.com/news_room/news.aspx?m=23®ion=en-gl...
E.g.,
http://store.apple.com/us/product/HE153VC/A/promise-pegasus2...
I have had the predecessor device, a 6bay Thunderbolt RAID 6, for two years and it has been quiet and reliable.
But I don't want two high-end compute cards, and I suspect that many who are trying to convince themselves that they'll benefit from it will gain no value from it.
For many, many workloads, modern compute still represents an iffy proposition (at the price levels being talked about, the Xeon Phi would almost certainly represent a better proposition). With unified memory things might get more workable, but as is it remains a relatively fringe benefit, and it seems odd that the entire value proposition of the machine relies upon it.
In which case, do you want a Xeon workstation of any sort? As mentioned later on in the review, you make significant sacrifices for Xeon (startlingly expensive, last-gen cores), and, besides the option for more cores than you can get on an iX, the main thing you get is extra PCIe lanes, which are not actually that useful for most things; one of the few things they _are_ useful for is dual hefty GPUs.
ECC memory is a big one. Usually getting a Xeon workstation comes with SMP, though not on the new Mac Pro. Big memory support. Lots of PCI lanes. Usually lots of space to drop in extra storage, a couple of 10GbE ports (the Mac Pro has just 1Gbps ports which is another oddity).
There are a lot of traditional reasons a so-called workstation features a Xeon.
Worth noting that there is a couple of Haswell (therefore AVX 2.0 supporting) Xeons -- the E3 v3s. Unfortunately they're the baby ones so they have ridiculous low max memory, no SMP, and max out at 4 cores. Hopefully the E5 v3s are out soon.
I honestly don't get how the Mac Pro hasn't gotten more mainstream criticism. It solves a problem that I don't remember anyone ever having (honestly a garbage can form factor seems like more of a nuisance than the flexible cubes we're all used to), while bringing a ton of problems to the table, and being a massive sunk cost for fixed hardware that is going to be outdated very, very quickly.
The Xeon Phi is significantly more expensive, draws significantly more power, and is significantly less useful for creative software loads.
Which is what? It's just anecdotal but most of the people I know with Pros got it as a high end development workstation, building iOS apps, etc. Final Cut Pro is pretty much the only app that benefits from the dual GPUs, and even then the gain is relatively marginal over a machine four years old. And that's paying for very high priced "workstation" GPUs (I called them compute cards because that is what they are geared for, though as with all compute cards they are derivatives of GPUs. They're really price ineffective as GPUs), when you can get almost all of the same advantages on a basic ATI card for a couple hundred dollars.
Virtually every review of the Pro seems to be giving it a very soft glove approach for some reason. It is an enormously expensive monument to the dual compute GPU, for marginal gains in most apps.
The Mac Pro is a reasonably priced dual compute GPU workstation. But it's enormously expensive when compared to what most people actually need.
I'd say that if you are doing 4k Final Cut Pro work or writing your own OpenCL software, this machine is for you. In the future it may be suitable for people using GPU-based 3D renderers (I don't think there's any great ones on OpenCL at the moment).
Aside from the above niche target market, the Mac Pro target market doesn't exist yet. Apple seems to be using this machine to push the development of OpenCL and to push the development of "Pro" software for Mac OS X in the direction they want to see it go.
Personally, I think it's a very exciting direction to take pro software. I hope many developers jump on board. Once that happens I think we'll start to see a larger target market for the machines.
Only a domain expert could be expected to automatically know whether double GPUs is effectively the same thing as double CPUs or not. Apple has put a lot of effort into making their GPU structure effective (in general, not just for a few special cases), and I didn't know if you were saying they'd failed or not. i know parallel processing is hard, but i also know apple has smart people who've worked on it. i don't know things like whether "lack of unified memory" is a problem for apple's design too, or not. it seems completely plausible to me, not knowing the domain that well, that apple could have had something that works well in general, or not – i don't really know and you didn't say, just assumed your reader would somehow know what you meant (which will basically only work for people who already know your point and have no need to read your comment at all).
He got antsy because from his perspective, your question definitely seems kinda out of the air. In his original comment, he's basically implying that any GPU based compute solution (so these can be plain gaming optimized GPUs, the "professional" GPUs, or compute optimized GPUs) aren't worth is for the majority of use cases.
The "second card" he talked about was the Xeon Phi, and he makes the differentiation between it and other "GPU based" compute solutions since Xeon Phi consists of a "large" (sub 100) number of relatively simple, but full blown CPU cores (for example, the current Xeon Phi is based on an old Pentium core, the next gen is supposed to be basically an Atom core). This should, in theory, make it easier to exploit parallelism.