Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
This drivel again.

You're accusing someone of drivel for telling users to use a faster computer with more modern upgradability. Try insulting people less in life and you might learn to be more constructive.

A laptop with eGPU isn't a solution for many people - what if your workflow requires MacOS and GPUs running at full speed?

macOS and GPU in a cheese grater with CPU and PCIE bottlenecks will not perform as well as a new Mac (MBP, iMac Pro or Mac Mini) with eGPU or other upgrades.


As for Thunderbolt RAID, that's a solution looking for a problem the cMP doesn't have.

The cMP's drives are stuck on SATA2. With upgrades you're stuck with PCIE 2 bandwidth limits.

TB3 RAID will outperform both of these.
 
TB3 RAID will outperform both of these.

How?

If I understand correctly, the 8950HK (latest MacBook Pro), or 8700B (latest Mac mini), or 7700k (latest iMac) only has 16 PCIe lane in total. Even that's PCIe 3.0. The total available bandwidth still less than a single W3690 can provide.

AFAIK, iMac Pro (48 PCIe lane) is the only exception that can really provide more bandwidth (than cMP) for multi GPU setup.

However, the block diagram suggest that only 8 lane (in total) for all four TB ports. Therefore, the combined bandwidth for all four TB3 ports just equals to a single PCIe 2.0 x16 slot.
upload_2017-12-21_16-16-15.png


Of course, if you put the iMac Pro's onboard Vega's PCIe 3.0 x16 bandwidth into the calculation, then the combined bandwidth for all GPU will become PCIe 3.0 x24, that's definitely more than the max possible that a single cMP can achieve (PCIe 2.0 x36).

However, it's very clear that the onboard Vega is downclocked to fit inside the iMac Pro's thermal / power envelope. No matter give it how much bandwidth. It still can't perform better than the Vega on cMP. Most of the bandwidth at there is simply wasted.

So, if we look at the real possible GPU performance comparison.

A Vega FE in cMP PCIe slot 1 can do better than the Vega on iMac Pro. Luxmark proved this.

The PCIe slot 2 can provide same bandwidth as all four TB3 (combined) on the iMac Pro. So, if we assume the GPU performance is bandwidth limiting, then they should perform the same in this area.

However, the cMP still has a slot 3 and 4 to accommodate more GPU. iMac Pro? Already run out of expansion bandwidth (for GPU).

If the workflow is really GPU limiting, but not CPU limiting. I am afraid that none of the newer Mac can out perform a 5,1.
 
You're accusing someone of drivel for telling users to use a faster computer with more modern upgradability. Try insulting people less in life and you might learn to be more constructive.



macOS and GPU in a cheese grater with CPU and PCIE bottlenecks will not perform as well as a new Mac (MBP, iMac Pro or Mac Mini) with eGPU or other upgrades.




The cMP's drives are stuck on SATA2. With upgrades you're stuck with PCIE 2 bandwidth limits.

TB3 RAID will outperform both of these.
OK, that's what we're dealing with. Cheers.
 
Any updated or new GPU drivers offered in the 2018 Mac Mini ?

I hate the idea that Apple has largely 100% given up on PRE-BOOT menu options . .. GPU-wise.
If they have then I've had enough of the Post Steve Jobs era. I'm going to make a dazzling Hi Sierra or Mojave ( whichever ? ) Hackintosh.

h9826790 has done this but in my case with the time left to me I'll be forever done with Apple ( I'm 69 years old ).

I'm tired of 'almost developed ' yearly OS updates = some serious OS bugs just ignored. Eg : try re-naming a Bluetooth headset in Sierra.

I and many of you paid exorbitant prices for our new cMP 1,1, 2,1 3,1 4,1 & 5,1's not one of which included SATA III nor USB,3.x. which was already available for most of the cMP varients.

It took a Tsunami of Bug Reports ( . . . . perhaps . . . just myme. . also that some Apple OS developers sympathise with us ) + plus the undeniable fact that WE ( Gilles, Dosdude2, TsiAlex + dedicated & 200% stubborn cMP owners etc. ) proved that our MODULAR Mac pros could boot NVMe M.2 or SSD's.

Tim wants the nMP to be a big iphone with everything soldered to the motherboard.

Enough raving .. just had to get that off my chest

I wish a Happy New Year from an Aussie in Japan to all cMP stalwarts wherever you live tonight !

:D
 
So...my GTX 1080 is sitting a top my cMP patiently waiting for a web driver I fear may now never come. Speculative, but should I just offload this card while it still retains value? This whole situation is looking quite grim. :(
 
So...my GTX 1080 is sitting a top my cMP patiently waiting for a web driver I fear may now never come. Speculative, but should I just offload this card while it still retains value? This whole situation is looking quite grim. :(

Why don’t you hang on to it? Does Mojo offer you something that High Sierra doesn’t? I’ve been running Mojo since the 1st beta. It’s a decent enough OS, but not so much that it leaves HS in the dust.

High Sierra was for me a very reliable OS, and doesn’t strip you of FileVault2, or any of the options we’ve all come to know and love in macOS. I am running Mojo, but I didn’t ever use FileVault anyway.

Stick it out until you feel that you MUST upgrade. Just my 2¢
 
  • Like
Reactions: Synchro3
Why don’t you hang on to it? Does Mojo offer you something that High Sierra doesn’t? I’ve been running Mojo since the 1st beta. It’s a decent enough OS, but not so much that it leaves HS in the dust.

High Sierra was for me a very reliable OS, and doesn’t strip you of FileVault2, or any of the options we’ve all come to know and love in macOS. I am running Mojo, but I didn’t ever use FileVault anyway.

Stick it out until you feel that you MUST upgrade. Just my 2¢

I upgraded a while ago, for experimenting with the Titan Ridge PCIE TB-3 card. At the time, HS wasn't working with it. I've never needed file vault or boot screens but have grown fond of dark mode and some of the new stacking/sorting features in Mojave. I'm just worried that Nvidia Web Drivers/Cuda have been sunset and will not see another day. I suppose that's why I'm thinking about it.
 
So...my GTX 1080 is sitting a top my cMP patiently waiting for a web driver I fear may now never come. Speculative, but should I just offload this card while it still retains value? This whole situation is looking quite grim. :(

Mine is still in my machine, waiting for the web drivers. HS aint so bad, but I wish I was on Mojave13879dunno.gif

Lou
 
  • Like
Reactions: crjackson2134
That has now been posted to at least 3 threads. I posted one, though not sure I believe it to be authentic13879dunno.gif

Lou
 
Commander, tear the internet apart until you’ve found those drivers! And bring me the developers, I want them alive!

We have no choice General, our cruisers can't repel firepower of that magnitude! We've got to give those fighters more time. Concentrate all fire on that Super Apple Destroyer.
 
  • Like
Reactions: AlexMaximus and w1z
Hopeful, but you never know what to believe with tier 1 support staff.
I wish Nvidia would just come out and make an official statement, even a "We're working on it" or "Not going to happen" would suffice.

If we can infer anything from their Windows drivers:
Support for the 580 series (release date: Q4 2010) was dropped from driver support in Q2 2018.
The K5000 and 680 Mac Edition were released Q3 2012 and Q2 2013, respectively.
7.5 years of support. So we're on the inside edge of that if they apply the same support principles to those products.
 
If we can infer anything from their Windows drivers....
You can't, though.

Windows has much more elegant opaque driver APIs without extreme version dependencies. Nvidia also has an elegant driver architecture where older drivers work fine - just without exposing newer features.

On the other hand, OSX SUXS.
 
  • Like
Reactions: ssgbryan
You can't, though.

Windows has much more elegant opaque driver APIs without extreme version dependencies. Nvidia also has an elegant driver architecture where older drivers work fine - just without exposing newer features.

On the other hand, OSX SUXS.
I should say would their product lifecycle support for Mac products be any different than Windows?
Especially for their professional products.
The Quadro K5000 for Windows still receives Windows 10 driver updates.
 
If you guys will think about it. Last time when we have had Nvidia GPU in any Apple computer was... over 3 years ago. Apple could have thought that there is no reason for them to support new Nvidia web drivers, and can let this platform rot with old ones.

If it is Apple who is blocking Nvidia to post new web drivers - wouldn't this be reason why?

If so - no Nvidia GPUs on Macs for foreseable future. Even as a eGPUs.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.