Isn't windowed mode a pretty poor way to benchmark? Games almost always run more slowly when in a window, as the GPU is still rendering all of the operating system stuff (which is being redrawn, etc). Furthermore, if different OS versions are used then it's possible that one uses the GPU more extensively than the other; for example, Yosemite appears to be more intensive on the GPU compared to Mavericks.
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.
iMac 5K GPU Performance
- Thread starter robertojorge
- Start date
- Sort by reaction score
I pretty much agree but theSeb wanted windowed for an apples-to-apples comparison with FireWall0400's numbers so I did it both ways.
And the difference is negligible anyway between windowed and full-screen!
Actually the difference between the two GPUs is not very much here either, maybe 15%?
And the difference is negligible anyway between windowed and full-screen!
Actually the difference between the two GPUs is not very much here either, maybe 15%?
I find the numbers surprising, but it just highlights the issues with synthetic benchmarks and trying to make any valid conclusions based on them.
Having said that, can I be cheeky and ask you to do one final benchmark using Unigine Valley, but this time change the preset to ExtremeHD (1920x1080 8XAA) and do not touch anything else? We already have a bunch of scores for that. It's the best thing to go on in the absence of having real world comparisons right now.
Attachments
What you are posting there is nonsense. Or let's say it is an enormous misunderstanding of your consumer rights.
Apple doesn't have to make free repairs for any failures for six years. The seller (which may or may not be Apple) has to fix problems that are due to a defective product being sold. Even after two years, you can expect that some things just break without the product being defective when it was sold, and after six months it is up to the customer to _prove_ that the problem was caused by a defective product.
The "six years" is the point when all your rights seize. Even if you bought a bronze statue for your garden that should last for hundred years, if it breaks after six years the seller doesn't need to help you anymore. But before the end of the six years, you still have to prove that the problem is caused by a defect that was present when the product was sold to you.
----------
The UK doesn't have exactly the laws that the poster claims it has.
----------
(If you are waiting for your batteries to give up to have an excuse to buy a new iPad, don't read further)
If your battery stops working, Apple will replace it for a reasonable price, much cheaper than buying a new iPad.
This video has been posted here ages ago, but most people seem to have missed that thread
https://forums.macrumors.com/threads/1807472/
So I've been playing with my 5K iMac over the weekend now. Totally love it. Coming from a mid-2011 iMac with 4GB ram and a Radeon 6770m with 512MB, this is like night and day...
Mine is the top of the line... everything upgraded except I went with 3TB fusion instead of SSD.
I'm not a designer, I don't edit videos, I don't program... basically I don't tax the system at all. The old iMac was fine for my browsing, amateur photography, spreadsheets and light gaming duty using bootcamp.
4.0GHz QC i7 Turboboost 4.4GHz
32GB 1600MHz DDR3 SDRAM-4x8GB
3TB Fusion Dr (3TB+128GBFlash)
AMD Radeon R9 M295X 4GB GDDR5
AppleCare
That config is overkill for my use. But due to a quirky tax break, I get back 60% of whatever I spend back as cash bonus PLUS 100% of a qualifying expense. So I get back more if I spec it up more? yipee!
on Yosemite, for everyday use, its super smooth. No lag on mission control, with parallels, or with the many many browser tabs and spreadsheets open.
Gaming on OSX with steam is ok for Counter strike and Dota 2 but I find the fan working overtime.
Same games installed on a Win7 with bootcamp doesn't tax the temp that much. Fans are mostly quiet. Has to be openGL vs DirectX here.
Anyway I digress... what do you guys need for benchmarks? Are there any suits I can run for you lot?
----------
Yes, on my 2012 rMBP, Mission Control was choppy as well. I figured it was because I only took 8GB RAM. My bro has the same model but he spec'd it to 16GB and I haven't heard him complain about the choppiness... but its still there in my system.
Mine is the top of the line... everything upgraded except I went with 3TB fusion instead of SSD.
I'm not a designer, I don't edit videos, I don't program... basically I don't tax the system at all. The old iMac was fine for my browsing, amateur photography, spreadsheets and light gaming duty using bootcamp.
4.0GHz QC i7 Turboboost 4.4GHz
32GB 1600MHz DDR3 SDRAM-4x8GB
3TB Fusion Dr (3TB+128GBFlash)
AMD Radeon R9 M295X 4GB GDDR5
AppleCare
That config is overkill for my use. But due to a quirky tax break, I get back 60% of whatever I spend back as cash bonus PLUS 100% of a qualifying expense. So I get back more if I spec it up more? yipee!
on Yosemite, for everyday use, its super smooth. No lag on mission control, with parallels, or with the many many browser tabs and spreadsheets open.
Gaming on OSX with steam is ok for Counter strike and Dota 2 but I find the fan working overtime.
Same games installed on a Win7 with bootcamp doesn't tax the temp that much. Fans are mostly quiet. Has to be openGL vs DirectX here.
Anyway I digress... what do you guys need for benchmarks? Are there any suits I can run for you lot?
----------
Yes, on my 2012 rMBP, Mission Control was choppy as well. I figured it was because I only took 8GB RAM. My bro has the same model but he spec'd it to 16GB and I haven't heard him complain about the choppiness... but its still there in my system.
OK, here you go.
But I think you are overdoing the "synthetic benchmarks" position a bit here. Polygons, shaders, it's going to be pretty representative of the comparative performance on most games unless you're talking about really old ones that don't use the latest GPU features.
Attachments
Do you have any graphic demanding games installed on Windows? It would be nice to see some gaming permformance for a change.
Away from my iMac until tonight, but the next thing I'm interested in running is the same test(s) on a 1080p external display.
I'm thinking that once you get into the business of updating 14M physical pixels every 30 seconds or so some of the relative advantages of these GPUs are diminished.
I'm thinking that once you get into the business of updating 14M physical pixels every 30 seconds or so some of the relative advantages of these GPUs are diminished.
hmmm.... no Crysis. Though I do have Tomb Raider 2013. the rMBP had issues running that on high. I'll install on this and see.
But just bought Civ:Beyond Earth.... aargh... so much time wasting.
Other games won't be so taxing on this system. Counter Strike, Dota 2, Heroes of Newerth, L4D, Team Fortress... are all on high and plays smooth. These games are pretty old anyhow
One important thing to know if you want to game on the retina iMac - the GPU runs over 100 C on even the basic games I've played. I returned mine because of this (I'm not going to be stuck with a machine in two months that has a faulty GPU, no matter if it's covered by Applecare or not).
Yep, that's a high temperature. During multiple runs of the Unigine demo mine plateaued at 86 degrees...the fans were working hard too.
The iMac lets the i5 heat up to 85 degrees too (long Handbrake job), but the fans are still low since it's only pulling 59W according to Intel Power Gadget.
The iMac lets the i5 heat up to 85 degrees too (long Handbrake job), but the fans are still low since it's only pulling 59W according to Intel Power Gadget.
I have had my GPU temp as high as 104 C after running long strings of benchmark's, but it hangs around 100-102 during sustained tests.
I am going to power off my machine tonight and then boot up cold in the AM and rerun the benchmarks to see how much the heat is affecting them.
And that was my issue… it wasn't just running games at 5K resolution, it was running them at the old 2560 x 1440 resolution (and let me tell you, running them at that resolution looks *horrible* - compared to my 2012 iMac it's blurry and almost unplayable).
I tried posting here but no takers, so I did a lot of online research and it turns out that any sustained GPU temps of 90+ are pretty much going to wreck the machine. If you just have a fan app like SMCFanControl it's only showing you the internal system temp, whereas iStat showed the very dangerous GPU temperature.
I think Apple created a really nice machine - and yet failed on the graphics card. I would imagine there will be some horror stories in a few months when graphics cards start wrecking people's systems (and I'm an Apple fan, I'm not wishing for it to happen).
EDIT: By the way, running games at 2560 x 1440 still caused the GPU to run over 90 C. There was no difference in the 5K versus standard iMac resolution. I ran four games and all of them caused the CPU to overheat like this.
Last edited:
I don't have time right now to run all of them, but as a point of reference, a 2012 i7 iMac (mine) with 16GB ram and 680MX scores exactly 947 / 37.6 seconds with run with the same settings at 1st One above. Bizarre coincidence!
It seems basically most if not all of the extra power of the M295X is used to draw the extra pixels.
I don't necessarily agree with the temperature being an issue, as AMD responded to concerns around the R9-290X stating that the card was designed to run it's entire life at 95C. I realize this is a different card, but how can you be certain that the 295X is not designed to withstand the same temperatures? Here is the link for reference and scroll to the bottom for the quote. http://www.guru3d.com/articles_pages/radeon_r9_290x_review_benchmarks,11.html
I re-ran the Valley Benchmark after letting the GPU cool-down. Temp was sitting around 48C, jumped up to 60-70 range right at start of test, by end of test it was sitting at 100C
Benchmarks went up slightly with starting with a cooler GPU.
I did not buy my r5K for games, so this is not a major issue for me. I upgraded from a Late 2009 27" iMac so this was a huge performance upgrade all the way around for me.
Apple Care was a no brainier on this one.
Benchmarks went up slightly with starting with a cooler GPU.
I did not buy my r5K for games, so this is not a major issue for me. I upgraded from a Late 2009 27" iMac so this was a huge performance upgrade all the way around for me.
Apple Care was a no brainier on this one.
Attachments
That's fair, but as my purpose for using my iMac is mostly for gaming and design work, I'm not sure I'm comfortable with a system where the GPU runs 95-100C nearly all the time I'm using it.
If AMD comes out with a similar statement about the 295X, then I'll buy a retina iMac again (I returned mine to my local Apple store).(I know Applecare takes care of this stuff, but frankly I don't want my computer to be like owning a Jaguar - you know the old adage about needing two, one for the driveway and one that's always in the shop )