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.
The Mac Pro Criticism
- Thread starter Simplybetter
- Start date
- Sort by reaction score
Yeah, I'm just trying to remember who asked for 6 TB ports instead of useable PCIE slots.
I think the new Mac Pro is interesting. I will reserve judgement till we see pricing and specs.
A bit of an exaggeration . 1 hour of 4K won't exceed 6TBs at 16bit RGB.
That said, we will still need spinning HDDs for years to come.
I'm not sure if you are aware, but it isn't out yet!!
----------
Right - but in the tower? No need, they will be better served in a centralized networked storage unit.
Unn4 said:
wonderspark said:
I totally agree with you guys!!! But I seem to immediately ask myself the questions:
- I (we) have no place, power, or position to "justify" Apple's design of a given product or not then so why would I care if someone tries to? We either support Apple by purchasing their product(s) or we don't.
- As I see it this forum has two central themes: Apple and Mac rumor sharing, and supporting each other in the use of those products. So where's the Apple Design Suggestion and Complaint Box here? I believe it's been overwhelmingly established that Apple doesn't listen to us at all. They respond only to RMA claims and sales. So what's the real result of airing such complaints? I guess only to prolong or stifle sales of the new machine. Above one guy says:
criticism for the new mac pro on the fact of expandability and upgradability is silly.and another replies:
Not criticizing the new mac pro on the fact of expandability and upgradability based on current and future needs is silly.But I kinda think either is silly. Neither way helps anyone figure out how to fit the new machine into their workflow - which IMO is the only concern. If it can fit in your workflow maybe we can help by researching or sharing how best to do so. But if it can't then what, it's useful to complain and attempt to convince others how bad it is calling them Kool-Aid drinkers and such? I suppose if you're selling illegal cards then it is - for the obvious reasons... otherwise, I don't get it... Why?
- I'm not sure these complaints are even legitimate. What can't we do with the MP6,1 that we could do with the previous versions? We don't know if the CPU is easily replaceable yet. Drives are the same deal but in a different box than the main system unit. Memory is the same deal although fewer slots if not compared to the 6-core models. GPUs are the same deal too when you think about it - the box to house them are too costly right now - oh well... RAM was over $2,000 for 16GB when the MP1,1 released... and?
For GPUs the bulk of the complaints seem to come from one guy who sells hacked ROMs with his cards (claiming he writes EFI code
) and then charges copious amounts of money for them. Typically and in all other communities I've ever been involved in such illegal hacks are not supplied in any way shape or form! It's illegal to do so and the fact that MacRumors knowingly condones this (by supplying such with "Vendor" status) places them in legal jeopardy as well! It's illegal and unethical. So as honest individuals why would we want to participate in or support some guys unethical business? Just because Apple's cards are closed developments? Typically in other communities even the lowest of hackers don't do such things - pirates do tho! Instead, the ethical hacker points to or teaches where/how YOU can get the ROM files and then tutorializes how YOU can modify and burn them - and does all this for free in order to benefit a sub-community or some elite friends. Huge difference both legally and ethically! For GPU upgrading I don't see much difference between the old and new MacPros myself - just the addition of a (currently expensive) enclosure. I still get to tinker in the same way don't I? Actually It may be even better because now I can get something like up to 18 of the GTX570 for $75ea. used (I paid $90 for my 3-fan GIGABYTE GTX570 2 months ago), order just the TB2 three Card Edge circuit-boards, and build my own case. That's potentially like nine times the amount of tinkering! But certain people can't profit because those cards don't need EFI - because the new MP comes with the GPUs needed for booting and display already and the PC cards likely to be connected via TB2 work fine as compute devices without hacked or stolen ROMs.
- Like I say, I agree that tinkering, self repair, and self upkeep of any system is highly desirable! I dig it. And right now it does look a little expensive when one applies that to the new MP. But besides the price which will likely be half in just 10 or 12 months, I'm really not seeing much difference. It's external tinkering instead of internal but does that matter? Are we just fearing being called Kool-Aid drinkers, or is it really important to you whether you open the side panel of the main system unit to install the part or do the same by opening an external box?
There's even a cool DIY case that's unique to go with it too.
That ought to keep me busy of awhile.
The custom case isn't tubular though.
http://rog.asus.com/137662012/news/armorsuit-sabertooth-z77-mod-by-paul-tan/
.
Last edited:
I did as well. About a month or two prior to WWDC. But I didn't really specify "instead of" - just that I wanted lots of TB2 ports.
----------
Thanks.
True, I don't doubt the enclosures will eventually drop in price. The problem is that unless they are much more affordable on day 1, then it's seen as an additional expense should you upgrade to the new machine. The upgrade to 16gb of RAM might have cost you a small fortune at the time, but it didn't cost you any more to do a 1:1 upgrade.
Both are probably silly, but I think they're equally valid opinions in the eyes of an internet forum. We're talking about an unreleased product with minimal verified information on it. So nothing can really be said with absolute certainty.
It's the sycophants and trolls residing on the far reaches of the spectrum that ruin any rational discussion around here.
- But the machine is slated to already come with two new and powerful cards from day-1. So the only rational purpose I can see for adding a TB2 connected GPU is as a third compute device. I dunno. I see your point and all but who going to want to add a third card like that right away? And from what I've seen other Brand Name Workstations which allow more than two GPU cards start around $16K to $18K so the MP6,1 might still be cheaper yet. I sure wish Apple would have mentioned a target ballpark price model at WWDC. It would eliminate a lot of the guesses and conjecture.
- Yup! Both good points! Being told I'm an unprofessional Kool-Aid drinking fanboy who should be placed on everyone's ignore list for what I think are common sense opinions or just trying to figure out how the machine is going to best be used, kinda makes me wanna question the concept and purpose of it all tho. (And MacVidCards supposedly a respected "Vendor" here said just exactly that - and in 3 or 4 different threads too). Irkfull!
Last edited:
Yeah, it really is going to come down to price.
As for the gpus, I think it's awesome that there are going to be some powerful options and 2 of them at that. However, developers still dictate which hardware will run software best. So while Apple has thrown its weight behind OpenCL, certain programs still perform much better with CUDA or even just NVIDIA cards versus their AMD counterparts. Of course that can change and I've seen that Adobe has promised OpenCL support for Premiere. But it could be a long long while before other programs do the same. And in the case of Adobe, I'll have to subscribe to CC in order to get the updates. Autodesk, Avid, and everyone else? Who knows?
Tru-dat! It kinda makes me wonder if there won't be a CUDA capable Nvidia option for MP6,1 with it's release or shortly after. It's a puzzling that the orator at WWDC said "for those that use OpenCL - and you know you should..." (did he actually shake his finger when he said that? Heh). When I head that I kinda cringed slightly. I hope he didn't mean anything stubbornly exclusive by that remark - as Apple has done in the past. It kinda reminds me of Bill Gate's remark: "You'll never need more than 4K of RAM" or whatever it was.
That remark makes perfect sense, an open language that isn't tied to specific graphics cards (or graphics cards specifically) is clearly better than a proprietary language owned by NVIDIA, tied to their hardware.
EDIT: And this being a developer conference, it makes even more sense since the all new Macs support OpenCL and the crowd is full of developers.
Last edited:
Tru-dat! It kinda makes me wonder if there won't be a CUDA capable Nvidia option for MP6,1 with it's release or shortly after. It's a puzzling that the orator at WWDC said "for those that use OpenCL - and you know you should..." (did he actually shake his finger when he said that? Heh). When I head that I kinda cringed slightly. I hope he didn't mean anything stubbornly exclusive by that remark - as Apple has done in the past. It kinda reminds me of Bill Gate's remark: "You'll never need more than 4K of RAM" or whatever it was.
I can't remember but, have they ever sold a model with both nvidia and amd/ati available for the same machine?
If an open language, not tied to a specific graphic card is good, what does that make a completely closed box that IS tied to specific graphics cards with no possibility for ugprades or expansions?
I don't see how we can pick and choose here. It's always good if Apple picks it?
Also, OpenCL isn't as good as CUDA right now. And NVidia supports OpenCL.
Yes. Actually all (MP Intel) machines so far if I'm not mistaken.
Last time i bought some, it only came with ati 5xxx cards. I had to buy card from 3rd party
I didn't mean driver support.
The difference is from a software vendor perspective, a language that can address all hardware have a larger market.
That's highly debatable, and probably comes down to specific implementations. I know that OpenCL works on NVidia cards, that's the point.
Yeah, I "cringed" for the simple fact that his remark sounded exclusionary. Like FU CUDA USERS! BAHAHAhahaha... <cringe>
I haven really looked into which is better or with the most potential or anything. I just think both should have a place in a hardware developer's rhetoric - at a show or whatever.----------
http://store.apple.com/us/mac/mac-accessories/displays
I see two NVidia cards and two ATI cards for the current MP and these current offerings are supported on some previous models. I can think of a time except the first 4 or 5 months of the MP1,1 where there weren't cards from both manufacturers offered.
The meaning probably wasn't "do not use anything else", but "use OpenCL". The fact that all Macs now support it, means that you know that your software will work and take advantage of it across the entire platform.
Yeah, I "cringed" for the simple fact that his remark sounded exclusionary. Like FU CUDA USERS! BAHAHAhahaha... <cringe>
----------
http://store.apple.com/us/mac/mac-accessories/displays
I see two NVidia cards and two ATI cards for the current MP and these current offerings are supported on some previous models. I can think of a time except the first 4 or 5 months of the MP1,1 where there weren't cards from both manufacturers offered.
Yes but they weren't listed in the bto when i ordered the last batch. I only had the choice of one or two ati 5xxx.
Ya, that's true. Often the on-line configurator doesn't list everything available. That's what you mean?
Through the Apple Store I guess you can get whichever you like initially. Or maybe in the worst case you'll have to sell or trade-in the old cards for the new maybe. But the options have always been available for every machine - with the noted exception.
When your product line serves a particular niche, it seems good practice to me to ask which has the greater ubiquity in terms of actual production. If you have one hardware vendor that supports both, why alienate a significant portion of your market by adopting hardware that does not support the language which may enjoy the broadest use? It seems to me that that would be definition of the larger market.
Incidentally, are there any numbers as which of the two (CUDA vs. OpenCL) is more commonly used?
Incidentally, are there any numbers as which of the two (CUDA vs. OpenCL) is more commonly used?