I doubt that, SSD is way more important than the gpu for Lightroom. I have a more powerful GPU and LR barely use it (I use the latest version). Unless the gpu is saturated there is no gain in getting a more powerful one.
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 Definitive riMac graphics benchmark thread
- Thread starter fruitpunch.ben
- Start date
- Sort by reaction score
I doubt that, SSD is way more important than the gpu for Lightroom. I have a more powerful GPU and LR barely use it (I use the latest version). Unless the gpu is saturated there is no gain in getting a more powerful one.
I have the 3.5Ghz i5 model, Fusion drive with M295X graphics driving 2 displays at work for 10 months and I don't hear any fan noise at all. The Intel Power Gadget shows ~46ºC. Meanwhile, I ordered the 3.3Ghz i5, Fusion drive with M395X graphics last week for my home use.
That's the million dollar question for us photographers! I'll have my i7 395 this afternoon and will see how UI and LR perform!
They are doing this for multiple reasons but if they beefed up the GPU on the imacs I'm guess that would cannibalize their Mac Pro sales quite a bit.
I'm just not convinced how much it could ever really cannabalize beyond what it does right now.
The Mac Pro is totally geared towards multi threaded crunching. There's definitely room for the iMac to thicken up a touch (at least in a large screen model) and get better GPU performance (thermal and otherwise).
The "thin" thing is so ridiculous also - It's just the edge! It's plenty thick in the middle. There's no harm in going up a few MM to solve some of these things.
Oh please report and if you could please quote me so I can get the notification and see what you thought of it
OK - so I did 8 Heaven benchmarks with the settings suggested in the first post in this thread. I have enclosed the results of the 1st and the 8th test - can you guess which is which? 
The results of all 8 test are pretty much the same. So as far as I can see there's no throttling. Of course I don't know if a lot more tests will give another picture.
As far as temperature goes, the temperature goes to about 90 after the first benchmark and reaches 97 in the second. After that it is 98-99 and did not once reach 100.
As far as the fan goes, I can hear it when the fan ramps up when the temperature reaches about 99 degrees. But it is not very much - the fan is still very quite. If the sound bothers you then you seriously have a super duper sensitive hearing.
The results of all 8 test are pretty much the same. So as far as I can see there's no throttling. Of course I don't know if a lot more tests will give another picture.
As far as temperature goes, the temperature goes to about 90 after the first benchmark and reaches 97 in the second. After that it is 98-99 and did not once reach 100.
As far as the fan goes, I can hear it when the fan ramps up when the temperature reaches about 99 degrees. But it is not very much - the fan is still very quite. If the sound bothers you then you seriously have a super duper sensitive hearing.
Attachments
Added. Thanks! That helps fill things out a bit more
When you get it, please run Unigine and Cinebench. There's been some strange results (see list) where the 390 outperforms the 395, I want to double-check
Last edited:
Hmm... he claims the M395X get just as hot as the M295X did, so over 100 oC. +10dB in fan noise, which is more than x2.
That should pretty much put it to rest. The M395X is a bit of a joke. We don't know how much of the speed increase (tiny as it is), isn't a combination of newer processor + faster memory, either.
I was disappointed with Apple last year when I got my i7/295X 5K iMac. I did it for the screen, but... aside from that, it was a louder, throttling-happy machine than my 2012 iMac with i7/GTX 680MX.
This year? There's really no excuse for the M395X. None at all. I remember how my jaw dropped with that GTX 680MX. It was night/day better than the 2009 iMac which came with the AMD 4850. It was a revelation for gaming in the Mac space.
Now, 3 years on from the 2012, we get the M395X.
One word: Disappointment.
Hmmm. That's odd, the title is, "The Definitive riMac graphics benchmark thread". I don't see hackintosh/2013 and prior in the thread...
Nope, it´s the i7 that handle that stuff.
GPU is still used mostly for filters and effects in photoshop. GPU processin is the future but in the pace adobe works that future might still be the future when you upgrade next time after this one
Oh it could. People routinely consider top spec iMacs over Mac Pros. I've seen the threads on it all the time mainly because Mac Pros are updated so little that people start to consider how close cutting edge 4 core intels stack up against older the xeons. I've seen people ask "why doesn't the iMac have 6 core processors available?" as well.
There is no business reason for Apple to go backwards by making the iMac thicker. The next version is likely to be thinner as well. If you want a cool workhorse Apple has 1 solution for you. Mac Pro. Take it or leave it. People continue to beat their heads on this all-in-one computer. It's going to get worse.
Either get on Windows and have options on what hardware you are using or get a Mac Pro.(or hackintosh but thats another story and a tad more risky when used for critical work).
Ha really you dont see test results for 2012 and 2013 imacs in the first post? If you dont see that, then that's somewhat troubling
There's no reason that the iMac couldn't have better graphics cards. It wouldn't infringe on Mac Pro territory anymore than it already does. The main purchase point on a Mac Pro is CPU performance.
I'm sure whatever iMac we choose it'll be ALRIGHT for whatever application. Just my 2 cents 'cos I'm getting crazy…
Seems like everything I read about graphics cards lately is all geared towards gamers, but is that the whole story? Some of us are interested in driving lots of pixels (eg. 5k plus two 4ks). We're not as interested in 3-D and other capabilities as we are in ensuring smooth 2-D across lots of real estate.
Could it be that this is more of a concern to Apple, than gaming? Are today's video cards equally adept at all areas, or are some slanted more towards creative professionals than gamers? Just a question: I'm not pretending to have the scoop. But I often read these GPU reviews and wonder how XXX frames per second of some first person shooter will correlate to pushing creative software out across lots of high-res panels?
I think I remember reading somewhere about AMD working with Apple on 5k, which has me wondering if that's part of the story?
Could it be that this is more of a concern to Apple, than gaming? Are today's video cards equally adept at all areas, or are some slanted more towards creative professionals than gamers? Just a question: I'm not pretending to have the scoop. But I often read these GPU reviews and wonder how XXX frames per second of some first person shooter will correlate to pushing creative software out across lots of high-res panels?
I think I remember reading somewhere about AMD working with Apple on 5k, which has me wondering if that's part of the story?
I believe that knowing how well a GPU performs in games helps to understand how it would handle other stuff such as photoshop etc…but that's only my opinion. That said I think that for non-gaming purposes any GPU works well. Perhaps the only advantage would be having more than 2GB, but that's m395X territory only.
I see them all over the place. The point I am making is that based on the title and thread info, OP wants to compare retina iMac benchmarks, not retina iMacs compared to older iMacs and Hackintoshs...
Looking at this old set of barefeats tests, the http://barefeats.com/imac12g6.html, the 680mx posts better scores than the r9 cards on heaven, but on valley, it falls behind. So sometimes benchmarks obscure rather than clarify.
Actually, a $5000 card made with PRO-apps in mind usually don't perform any better than a $500 gaming card once you start up a game.. PRO apps use the GPU´s in a different way than most games do.
So, FPS in gaming won't tell you much of anything regarding the performance in pro apps.
kind of depends if you need Cuda, doesn't it? I mean, the r9 cards blow the nVidia offerings away in OpenCL, but if the app you use prefers CUDA to OpenCL, you're out of luck.
Last edited: