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.
2016 nMP
- Thread starter ManuelGomes
- Start date
-
- Tags
- nmp
- Sort by reaction score
- Status
Hard to tell, quick search on my phone I found this page at Apple.com.
https://support.apple.com/en-us/HT202823
Doesn't seem to have been updated in awhile...
Amazing then that they couldn't be bothered to mention current shipping OS, isn't it?
"The tables below list the OpenCL and OpenGL versions that are supported by Macs with OS X Mavericks or OS X Yosemite."
I imagine the web guy had more important stuff to do like update watch colors or emoji status.
Or maybe, just maybe, OpenCl doesn't matter anymore to them. (The sly reader will notice that this was the original point)
I thought the original point was that the current version of OpenCL is 2.2 - and Apple's best is 1.2.
Yes, when in future Metal is mature enough Apple might sack both openCL AND openGL. Those can be emulated on top of Metal without a big speed penalty. (Ok, so they wont be totally away, but run on Metal). Apple already chose to use Metal for UI. There's a reason why openGL is still at version 4.1.
At the moment Metal is lacking important features to replace old openGL/-CL. If Apple's going to change OS X --> macOS 11, this could be one of the big changes. Metal will take over the GPU.
Sadly it could mean that macOS is compatible only with Mac's running on Metal.
Same will happen to iOS too, of course.
At the moment Metal is lacking important features to replace old openGL/-CL. If Apple's going to change OS X --> macOS 11, this could be one of the big changes. Metal will take over the GPU.
Sadly it could mean that macOS is compatible only with Mac's running on Metal.
Same will happen to iOS too, of course.
Last edited:
they do stuff like that with Ruby & Python as well. probably other languages too. (fwiw)
Well, what year is it? Seems like this issue has been batted around on this site since the nMP came out.
Not a lot of budging.
I don't think a lot of business decisions are made because people are "too lazy", although on this site, it's a common excuse. I think people who have never run a business are projecting.
So has the topic of FCPX, the 2013 Mac Pro, Apples doomed and the list goes on. It does not seem to stop anyone else, so why should this?
Was Adobe not adopting Metal?
If some here claim that they don't want to vendor locked-in by Nvidia with cuda, would they like to be lockedin with Metal?
Might want to re-think that.
http://www.pcworld.com/article/3071...adass-graphics-card-ever-created.html?page=10
Here we see AMD's "best & brightest" burning up 200+ Watts MORE than a 1080 (I am NOT making this up) while simultaneously delivering WORSE performance.
200 EXTRA Watts...guess where that goes, HEAT PRODUCTION. Heats up the computer, which expels it into your room, where it either remains to heat your room (like a 200 Watt Space Heater on your desk to keep your hands warm) , or you pay to use more electricity to move it outside via Air Conditioning.
So, half of a nMP Power and Heat budget spent on...nothing useful. Unless you live somewhere cold, where this would be useful as an additional...space heater.
So, which GPU manufacturer makes cards better suited to skinny jean wearing computers?
And, I might offer that the "space heater joke" is an AMD Fury X or 390X.
It's a pretty consistent and broad pattern for Apple. Do something and make a big deal of it at WWDC (or elsewhere) that it's the next big thing, and then send it out to the pasture of neglect and deprecation.
Aperture, OpenCL, OpenGL, mDP are a few of the recent examples. How many Ipod/Iphone docks do you have for the various docking ports?
Last edited:
If Apple just wants to sell a big FCPX dongle, I'm good with that.
It just should be labeled as such, and it probably shouldn't cost ten grand.
I wasn't comparing GPUs at all. The immediate tangent was about a new lower-end Mac desktop, and I was replying to someone suggesting Xeon v5 systems.
I merely quoted the price of a Dell Xeon v5 system with both entry-level Quadro and entry-level FirePro cards. The post that I was replying to clearly stated that the higher end GPUs would be CTO options.
Yes - my thought is that the topic is "2016 nMP update", but you want it to be about Pascal because you found (or your handlers found) a single benchmark on which AMD does well.
For which you post one image with no context and no link to the full report. Perhaps you've noticed that I usually include links to the full story on my posts - so that you can read the details, and if I've misinterpreted something you can call me out for a correction. (And if the report has Nvidia losing on 9 out of 10 tests - I wouldn't dare post the 10th test where Nvidia wins.)
I consider selective cutting and pasting of items like this, without links to the original, to be intellectually dishonest.
[doublepost=1463616159][/doublepost]
And perhaps Apple should have a more flexible (PCIe slots and internal disks) system for the people who don't need yet another Apple dongle.
But if they did so, they could discontinue both at the same time, with justification.
I'd wager 67 to 33 that the Mac Pro line will be discontinued at WWDC. It's just no longer important for Apple.
The Imac is fine for making Iphone apps. No business case at Apple for anything better.
Last edited:
Aiden, I was amazed that you thought he might misrepresent something so I went and found the test he grabbed that image from.
I guess he forgot to mention a few things, like the fact that the testers decided that the Luxmark test & card had an issue:
The GeForce GTX 1080 didn't perform quite as well as expected here. With its high clocks and 9TFlops of peak compute performance, it should have performed better here. We mentioned the results to NVIDIA and a bug report was submitted, but we don't have any additional information to share just yet. The GTX 1080 outran the Titan X in the most-taxing Hotel Lobby scene, but trailed otherwise.
And that the 1080 (and 980Ti) beat the Fury X on his old fave, "Ashes of Singularity" the DX12 test. (whoops !)
And all the tested Nvidia cards eviserated the Fury in 3D Mark.
To be fair, there was one other place where the Fury had the highest numbers. Power use !
Surprise ! It used nearly 100 Watts more than 1080 while performing significantly worse !
I guess we can understand why he forgot the link.
It was also significantly faster than a Radeon R9 Fury X, in both DirectX 11 and DirectX 12 titles.
Throughout our entire suite of tests, the GeForce GTX 1080 outpaced all of the other GPUs we tested, in every application / game and at every resolution, with the sole exception being the OpenCL-based benchmark, LuxMark.
Last edited:
That sounds about right, but they may have forgotten that they actually still make it, there's little evidence to the contrary.
I'm 100% on the nMP getting an update this WWDC, while maybe not as good for someones and much better than expected for others, and mostly focused on VR content not compute, so I dont have big expectations on the nMP FP64 performance, it should be as poor as current, i didn't eat the rumours on Polaris having 1:3 FP64, and at minimun 150W TDP its unlikely the nMP would get any PASCAL GPU unless apple expands the nMP TDP, which actually is easy, but they'll do? Apple could expand the nMP tdp to 1KW and accept dual high end compute gpu as well dual 145W xeon, but this would need another product along the nMP name it nMP^2, tecnically possible, Apple unlikely they will build an mac which will cost upto 20-30 grand.
Only produc to be discontinued its the Macbook Air
[doublepost=1463619490][/doublepost]Posible Moves from Apple about HW at WWDC:
most Likely Updated: tcMacPro with Xeon E5v4+Polaris+TB3, rMBP15(16) with Polaris/skylake/tb3, and rMP14 & rMP16 with Skylake Iris Pro 340 upto 380.
most Apple-interest clever Updates for WWDC: retina iMac to Skylake/TB3/Polaris, rMBP15 w polaris/skylake/tb3, Apple Retina display Thunderbolt, Mac Mini witk Skylake/polaris/tb3, no Mac Pro Update until AMD-Vega gpu is available, maybe also paired with Zen CPUs.
Either I spect some VR-focused product, either Apple's Carboard Like kit or an iPod-VR or iVR on vr-dedicated HMD.
Last edited:
I hope they don't discontinue the nMP just so I can keep reading these threads.
I don't see the nMP going away any time soon as Apple is already have invested in FCPX and the nMP for the unseeable future.
they use them at apple HQ.
or- at least the design team uses them but i'd assume many of the other depts (developers/photo/ad/video/etc) use them as well. (source: recent tv program showing the inside of design sector.. mac pros were on the desks running Alias for mac)
idk, from my perspective, it's pretty cool that one of (or the?) the largest companies in the world are doing their industrial design work on mac with software tailored to os x.
..because ten years ago, this wouldn't really be possible.
anyway, i'm pretty sure apple completely remembers that they still have a mac pro in the lineup..
that said, they'd probably be fine if the imac was their highest end machine at HQ (assuming they had a cluster or two to for certain jobs.. but they probably use these clusters anyway.. even with a mac pro at the desk)
- Status