Sidebar Sidebar
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

tillsbury said:
Old 2014 maxed-out rimac, (i7 4GHz/295x). Ran it through a couple of times to get the GPU up to temperature and then benchmarked when it was stable, with the fans very slightly on and the temperature steady at 99. Surprised you haven't had one of these on here yet. I only bothered doing this because I was surprised the 295X beat the 395X at Cinebench just slightly. Same results with Heaven:

View attachment 594890
Click to expand...

Nirurin said:
weirdly, your 775 unigine score is higher than the 780 scores??

and it cant be due to any throttling/heat issue, cos my 780 never got over 80c during the test.
Click to expand...

It's this sort of thing that makes it difficult to decide on which card to get!
I am still of the opinion that the difference between them is not all that great, probably not worth the money, but still waiting on some Unigine Heaven benchmarks for new riMacs.
 
fruitpunch.ben said:
It's this sort of thing that makes it difficult to decide on which card to get!
I am still of the opinion that the difference between them is not all that great, probably not worth the money, but still waiting on some Unigine Heaven benchmarks for new riMacs.
Click to expand...

The results all over the place, but at least its more clear. The thread is doing its work. Maybe we should try different benchmarks, or find a game a lot of people have and benchmark that?

All the cards are scoring within 2-3 frames of each other on unigine. The cinebench scores seems to make much more sense.
 
Indeed. Will be interesting to see the lower spec cards' results. But the 395X looks seriously underwhelming from here, especially given the price increase (the current top of the range imac in NZ is $1000 more than it was a year ago).

I can confirm I get the same Cinebench results as the french website review, too.
 
TechZeke said:
All the cards are scoring within 2-3 frames of each other on unigine. The cinebench scores seems to make much more sense.
Click to expand...

They might be close, but the percentage differences are the same between the two tests for the results that are in so far. There's a big difference between 17fps and 21fps. The exception is the 680 score on unigine, which seems very strange indeed given its Cinebench score.
 
fruitpunch.ben said:
It's this sort of thing that makes it difficult to decide on which card to get!
I am still of the opinion that the difference between them is not all that great, probably not worth the money, but still waiting on some Unigine Heaven benchmarks for new riMacs.
Click to expand...


Looks like we will get the information that we want tomorrow. http://barefeats.com/imac5k12.html Tomorrow I will know if I order a custom rig or if I just buy the stock high configuration.
 
Melodeath said:
How will this tell you what you want to know? Seems like they're just going to be testing the 395X, as opposed to the 395, but maybe not
Click to expand...


For me that is what I want to know. I already know these GPUs are fairly mobile/non-gamey so this will tell me what the practical affects are. Basically what I probably already know—the M395X isn't worth the extra coin for most games today.
 
sn0man1 said:
For me that is what I want to know. I already know these GPUs are fairly mobile/non-gamey so this will tell me what the practical affects are. Basically what I probably already know—the M395X isn't worth the extra coin for most games today.
Click to expand...
I guess my question is how will you know the difference between the GPUs if they only test one? I'm trying to decide myself. I'd rather spend $250 for the upgraded GPU if it makes games decent. If it makes barely any difference, I'd rather save my money and build a separate PC
 
Melodeath said:
I guess my question is how will you know the difference between the GPUs if they only test one? I'm trying to decide myself. I'd rather spend $250 for the upgraded GPU if it makes games decent. If it makes barely any difference, I'd rather save my money and build a separate PC
Click to expand...
From the article I linked to "Tomorrow we will will post results showing how the 'late 2015' iMac 5K with the AMD R9 395X (4G) compares in graphics intensive performance to the 'late 2014' iMac 5K with the AMD R9 295X (4G) as well as the 'late 2013' Mac Pro with FirePro D300s (2x2G). And we're just getting started."

I do see your point. I bet that site will get to testing the regular m395 at some point.
 
Even though the results may show the 395X to be underwhelming. Can someone clarify if the 4GB will help overall performance with 5K resolution?

I'm not planning to game with the 395X however I assumed the 4GB would help the overall handling with the 5K resolution and editing videos and photographs on the monitor.
 
  • Like
Reactions: sluggishadj
sn0man1 said:
Looks like we will get the information that we want tomorrow. http://barefeats.com/imac5k12.html Tomorrow I will know if I order a custom rig or if I just buy the stock high configuration.
Click to expand...
Not what I want to know. I am more interested in 380/390/395. I have already decided the $300 for the 395X is too much, just wandering which of the three is the sweet spot for performance/dollar
 
  • Like
Reactions: OddyOh, sluggishadj and AppleHater
On the blender test, it seems to freeze up my machine while working on the second tile.
 
Textureboy said:
Even though the results may show the 395X to be underwhelming. Can someone clarify if the 4GB will help overall performance with 5K resolution?

I'm not planning to game with the 395X however I assumed the 4GB would help the overall handling with the 5K resolution and editing videos and photographs on the monitor.
Click to expand...

This is what I want to know as well. I will do no more than casual gaming on my iMac, but I will be doing a lot of photography and video editing (including 4K from my iPhone and Panasonic LX100). Specifically, will I benefit from having the 395x and what apps will be able to take advantage of the extra power and memory?
 
  • Like
Reactions: sluggishadj
Can we please get back on subject here. Honestly if you have an older iMac or hackintosh, I can guarantee you that no one cares. Nobody in this thread cares how much better your older iMac handles graphics at your lower resolution, or how your hackintosh with a top of the line desktop graphics card performs better than the AMD 3xx series.

The goal here is to help undecided iMac buyers help decide which configuration to purchase.
 
  • Like
Reactions: Hezzaa, OddyOh, theosg and 6 others
vetruvian said:
The goal here is to help undecided iMac buyers help decide which configuration to purchase.
Click to expand...

No, it is not. Looking at the first post, the goal is to create a big database of iMac benchmarks, which includes current-gen iMacs as well as older iMacs.
Especially having the older iMacs in is interesting for undecided buyers, since it shows that you can easily save a few $1000 by going for an older used model (if you don't need the 5K screen). The performance progress over the last years was very little.
 
  • Like
Reactions: theosg
I agree, it helps to compare with old gen, you can have a global overview, you can see the difference between different gen or differences into each gen...
 
Can you all try benchmark of FFXIV (link) on bootcamp with this settings (1080 & 1440)?
Using the exact same setting with all cards (680mx/780m/m290/m295x/m380/m390/m395/m395x) should clarify the differences between them on gaming.
Thank you

ps. I don't think opengl will give us a good comparison since apple is ditching it for metal (but i could be wrong so don't be angry xD)
 
I would suggest we focus on the same (free benchmarks, e.g. Unigine but probably not cinebench) at the same settings and we can then compare the relative performance.

I for one want to know which of the GPUs in the new iMacs offer the best performance. If there is little difference between the 395 and the 395x then I'll save the money.
 
Tesselation should be enabled in unigine and iMac's GPU's are not for games...there is a good reason why apple did go with AMD! Metal and AMD are going to smoke Nvidia, just wait and see...


GPUtest- TessMark
This benchmark measures tessellation processing speed. The more levels, the more GPU processing power is required. We tested with 16 and 32 levels. (RED graph bar means FASTEST in FRAMES per SECOND.)
Click to expand...


mpt5k1_t16.png




http://barefeats.com/mpt5k1.html
 
Last edited:

Attachments

  • Senza titolo 3.jpg
    Senza titolo 3.jpg
    78.2 KB · Views: 173
Last edited:
Fl0r!an said:
Barefeats benchmarks are online: http://barefeats.com/imac5k13.html
Click to expand...

Yowch... look at that noise and temp! I am still wondering if the i7 is heating up the enclosure (display) and causing as much or more of an issue than the GPU. I may be delusional but I believe the best gaming setup would be the i5 + M395X (maybe even the M395 based on the results of those tests). So goes the inability to isolate the GPU from the CPU because most people are checking the boxes for both GPU&CPU if doing custom order.

REQUEST: Does anyone here on the forums have an i5+M395X? I would be willing to chip in a couple bucks to pay for games and benchmarks so we can see how it stacks up to the Barefeats tests and other users on the forums.

Thinking about it and looking at the benchmarks more, the M395X is most likely a rebrand of the M295X so the performance increase shown is likely all due to Skylake update over Broadwell (which is the chip I believe is in the 2014 5k iMac).
 
  • Like
Reactions: sluggishadj and jordanz
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back