While I am with you here, I suggest installing a single workstation with the iMac and monitor arrangement with an external RAID over Thunderbolt. Let users use it who need next gen speed and see if they like it.
I predict Apple will kill the Pro soon in favor of a Mac-Mini form factor. A lot of minis fit on a rack, and there is no reason why it can't have 12-24 cores per box.
Rocketman
http://www.fcp.co/final-cut-pro/articles/380-the-top-ten-things-we-love-about-thunderbolt
The imac isn't really a great replacement or a suitable successor to the mac pro. The display really isn't so great, and it's too problematic of a machine. You can't even replace the hard drive with that stupid temperature sensor issue. What makes the Mac Pro great as a desktop is that it's an uncompromising solution. You can handle a huge amount of stuff within that single box. I wonder what imac growth looks like relative to laptops at the moment. It can't be that pretty. A few years ago they were incredibly popular, but now if you own a macbook pro, it doesn't always make sense to also own an imac.
With a mac pro or similar tower, you can add in any additional specialized hardware you need, extra drives, updated gpu, tons of ram, and if anything aside from a logic board goes bad, you can simply replace it yourself and be back up and running with minimal downtime.
iMacs are great for many, many people. Personally, I don't like the glossy screen or the fact that you can't get any airflow through it. My mother has the newest one in Hawai'i next to her old one. The old one was only four years old, and the screen took a dirtnap, forcing her to buy a new one. It makes me wonder if the humidity there is life-shortening for an iMac.Yeah, but bashing the iMac is not called for. It is a very popular machine serves a large and diverse audience. Me, included. The real issue is whether Apple will continue the long-in-the-tooth MP, revolutionize it; or, give it a final farewell. What gets me in the MP threads is how few posters talked about what will be new in both hardware and software technology in the future. Too many folks are fixated on the MP as it is and was. They do not sound like Steve Jobs devotees, Bill Gates devotees, maybe. Sigh ... Sorry, hit those negative icons at will.
Comments from macperformanceguide.com on this:
No New Mac Pro Yet, Which Means What?
Yeah, but bashing the iMac is not called for. It is a very popular machine serves a large and diverse audience. Me, included. The real issue is whether Apple will continue the long-in-the-tooth MP, revolutionize it; or, give it a final farewell. What gets me in the MP threads is how few posters talked about what will be new in both hardware and software technology in the future. Too many folks are fixated on the MP as it is and was. They do not sound like Steve Jobs devotees, Bill Gates devotees, maybe. Sigh ... Sorry, hit those negative icons at will.
Now days with iMacs coming with quad cores, 16 GB of ram, and terrabytes of hard drive space, and thunderbolts ability to add external storage, and an expansion slot chassis; I think this is an obvious move. Add a duel processor option to the iMac and there you go.
I am a science dude, and I need TRILLIONS of ops done real fast.
If the Mac Pro is dead, I am gone to linux (ubuntu, scientific linux, whatever's adequate).
I'm with the Mac because of ease of use & OpenCL. Kill that, and me & my students are moving the next day. It will open up a huge new space for some garbage PC company to move in. While I will probably have an iPad & some macbook for keynote, I, and all my students, will be gone.
I hope some of those retarded PC companies move into this space asap. We need cheap x86-64 hardware, we need python, opencl, matlab, C, C++, now. Yet anyone apart from Apple (& asus, perhaps) is simply a brain-dead slave to windows.
I can see Apple doing this move years down the line, if iCloud gets 10x faster than it is now, and we can use it just as Amazon CComputing. Onlive is an example of high-bandwidth computation over the internet. Even iOS could do massive work someday!
But not now, not in 2012, and not in 2013. Get icloud supercomputing first, let us move there, and change the line. Do it now and lose scientific work for good (to Amazon EC2 GPU instances).
If someone @Apple is hearing, please forward to Cook. There is a scientific community counting on you. Please don't let us down.
Well then why are you using Mac OSX and Mac Pros instead of GPGPU clusters?
Right - everything he needs is most likely available on CUDA with Nvidia hardware already. (OpenCL is a distant 4th in the 2 man race for GPGPU computing.)
Professor "science dude" should have dumped Apple a year or two ago - it's possible that he could have won the Nobel prize by now if he'd been using state-of-the-art hardware instead of Apples.
Hopefully implicitly parallel programming languages (Eg, JStar) will be in a state where they can be used commercially and scientifically soon.
Would make CUDA and OpenCL redundant.
Well then why are you using Mac OSX and Mac Pros instead of GPGPU clusters?
...but if it's not easily accessible from Fortran - it's irrelevant for much scientific programming.
![]()
...but if it's not easily accessible from Fortran - it's irrelevant for much scientific programming.
![]()
In my case, because the computations, while parallel, aren't terribly well suited for GPU-based computing. Believe it or not, not all parallel applications are, and while GPU clusters are *amazing*, they are not the silver bullet.
In my case, part of this is also that even if I am interfacing with a GPU-cluster, or one of the Linux clusters I have access to, I still need a desktop for everyday work. And I'd like it to be high-powered, as some projects I work on are particularly poorly suited to parallel processing.
I find OS X to be infinitely more productive than a Linux solution, while still being fairly painless to hop between the two. But, like the poster you're responding to, if Apple drops support for the Mac Pro, I'll be forced to move on. An iMac simply does not meet my needs.
----------
Truth.
The measures people will go too to avoid learning new things, Sadly.
Funnily enough, computers are better at deciding what is better suited for sequential or parallel computing than humans are.
I would say apples days are numbered in the high end production workflow and it's time for a new player to come in.