AMD Radeon R9 M295X Core Clock Throttling, Heat, and Performance

[WilliamG]

So my iMac is broken because it's cooler, mhh... ok... like the other who voted <100 degrees in the valley benchmark's thread.

Happy to have a faulty machine then

It really feels like you're not listening. Your system is NOT cooler. You already proved that with a video! Your GPU temperature is dropping MASSIVELY while your screen is BLACK. Do you not see that? Your video even showed a drop from 99C to 92C in less than half a second after your desktop showed up. What temperature do you think your GPU was at before the screen went black for far, far too long? That's right - 105C, just like the rest of us.

I don't know why we're continuing to go back and forth about this. You already showed us your GPU hits 105C.

 

[andy9l]

AMD Radeon R9 M295X Core Clock Throttling, Heat, and Performance

I'd also add that absolutely no-one from the <100 category has posted any kind of evidence.

The evidence in this thread suggests to me that all M295X 5K iMacs suffer >100C GPU temperatures, minor GPU throttling and increased fan utilisation (thus increased operation noise).

Unless I've missed something, I don't think there's any evidence to the contrary of those findings.

Whether or not these issues bother or even affect a given user is completely subjective - hence the disputes.

 

[driftless]

Guys, give it a rest. This thread is the worst part of MacRumors. Folks who do not have any real or no experience on the 5K iMac, pulling stuff off the internet, speculating, and talking out of their ass, etc., and dissing a machine that is fantastic. It is not gaming machine, sorry if you were delusional. It is a great machine for content production and works well Adobe CC, FCPX (yeah, I know, scaling), etc.

The 5K iMac is fantastic machine and just because you do not own it, can't afford, or that it does not match with your preferred counsel/Windows gaming PC, - move on.

- David

 

[WilliamG]

I'd also add that absolutely no-one from the <100 category has posted any kind of evidence.

The evidence in this thread suggests to me that all M295X 5K iMacs suffer >100C GPU temperatures, minor GPU throttling and increased fan utilisation (thus increased operation noise).

Unless I've missed something, I don't think there's any evidence to the contrary of those findings.

Whether or not these issues bother or even affect a given user is completely subjective - hence the disputes.

You are absolutely, 100% correct.

 

[WilliamG]

Guys, give it a rest. This thread is the worst part of MacRumors. Folks who do not have any real or no experience on the 5K iMac, pulling stuff off the internet, speculating, and talking out of their ass, etc., and dissing a machine that is fantastic. It is not gaming machine, sorry if you were delusional. It is a great machine for content production and works well Adobe CC, FCPX (yeah, I know, scaling), etc.

The 5K iMac is fantastic machine and just because you do not own it, can't afford, or that it does not match with your preferred counsel/Windows gaming PC, - move on.

- David

No, the problem with this thread is people who actually DO have the 5K iMac spreading nonsense because they're unable to interpret their own system stats etc.

And yes, the 5K iMac is a (somewhat relatively disappointing - compared to the 2012/2013 iMacs) fantastic machine.

 

[driftless]

No, the problem with this thread is people who actually DO have the 5K iMac spreading nonsense because they're unable to interpret their own system stats etc.

And yes, the 5K iMac is a (somewhat relatively disappointing - compared to the 2012/2013 iMacs) fantastic machine.

I see your point, but I am not convinced of that. I suggest that the 5K iMac is a new paradigm, as were other Apple products that were over-examined upon release. My Opinion Only.

 

[WilliamG]

I see your point, but I am not convinced of that. I suggest that the 5K iMac is a new paradigm, as were other Apple products that were over-examined upon release. My Opinion Only.

I've tried to be pretty realistic about the 5K iMac, really. It does run very, very hot - and it's unusual coming from the 2012/2013 models that in general ran very cool. It would be one thing if GPU performance were leaps and bounds ahead of the previous generation systems, but it's not.

In any case, long-term effects of this heat will/will not be observed, in time.

 

[driftless]

Yes, it will be interesting to see what experiences we have long-term. AppleCare is the friend of early adopters.

 

[dyn]

I'd also add that absolutely no-one from the <100 category has posted any kind of evidence.

You need to leave out "from the <100 category". The truth is that no-one is posting any actual evidence. There is only very inaccurate evidence (running a benchmark once or twice on a single machine doesn't make it an accurate reading!) to draw the conclusion that all iMac Retinas will reach a temp. higher than 100C for the GPU. Conclusions are drawn way too quickly.

The evidence in this thread..

...is not present and that's exactly the problem with it.

Unless I've missed something, I don't think there's any evidence to the contrary of those findings.

And that's the second problem: people living in their own worlds seeing all sorts of things that aren't there.

Whether or not these issues bother or even affect a given user is completely subjective - hence the disputes.

It's only a problem because people do not accept someone else's experience. Like I said, they live in their own world. Any other experience than their owns does not exist to them.

Just accept the fact that there are people with a positive experience as well as people who have a negative experience. Don't draw quick conclusions like most are doing here. I just wish people here could actually discuss the matter objectively instead of highly subjective.

 

[andy9l]

All evidence - yes, real evidence - so far has shown 100C temperatures or greater, objectively speaking. Not just on this forum - think bigger.

We are not talking about users who have bought the iMac and do not push the GPU.

 

[dyn]

Then where are those tests? You know, the ones where the poster shows them what he tested and how, did multiple runs on multiple machines with real measuring equipment and thus being able to actually draw the conclusion the average iMac Retina will reach 100+ temps when it comes to the gpu.

The only thing we have in these topics are people posting pics of iStat Menus after saying that they ran some benchmark/game/video/etc. That only says a tiny bit about the iMac it is tested on but absolutely nothing about iMacs in general.

You have a really poor feel for quality if you think a few measurements on 1 machine is enough to speak for the productline in general. It only says *something* about that particular machine, nothing more, nothing less.

Evidence alone is more than just showing off temps. It's the entire reasoning and interpretation behind it. Not many websites/users are capable of that. Like with diabetes: a high sugar level doesn't mean you are a diabetic, it only means that you sugar level is high. If you have no other symptoms it means you are not a diabetic but if you have them then there is a very high chance you are diabetic. Just because you measure something doesn't mean it's good or bad, you need more information, more context in order to say it's good or bad. People here simply ignore that fact and that is rather sad.

 

[andy9l]

AMD Radeon R9 M295X Core Clock Throttling, Heat, and Performance

You're too detailed. Think higher level. It doesn't matter why it happens, it happens. Watch the videos.

You find me a 5K M295X iMac that can run a benchmark at sub-100C, I'll eat my words. You won't, though.

As I've said before, macro trends don't just happen based on 5 screenshots. There is a huge amount of noise online about the issue - excuse the pun.

To be clear NO-ONE has said it's bad. We don't know that. It's noisier and some find it concerning, that's all.

 

[AlifTheUnseen]

Guys, give it a rest. This thread is the worst part of MacRumors. Folks who do not have any real or no experience on the 5K iMac, pulling stuff off the internet, speculating, and talking out of their ass, etc., and dissing a machine that is fantastic. It is not gaming machine, sorry if you were delusional. It is a great machine for content production and works well Adobe CC, FCPX (yeah, I know, scaling), etc.

The 5K iMac is fantastic machine and just because you do not own it, can't afford, or that it does not match with your preferred counsel/Windows gaming PC, - move on.

- David

Many of the posters here have or had (see Andy9l's signature) a 5K iMac, so what are you talking about?

This thread is the very best MacRumors forum thread in recent times, hands down. While everybody else was enchanted by "OMG 5K in such a low-price machine", and the reality distortion field was in turbo-mode, you could find info that counts in this thread. At the beginning, it was a sparkle of doubt, recently, a clearer picture emerges. Of course it's not convenient, of course questions are still open, but it's on a higher level than the whole "Whoa, this screen is just fantastic, I could look at the desktop pic for hours…"-type of reviews and posts.

I thank the thread starter and all those who added to the discussion. I especially give my hat to those who measured stuff and tried to bring light into this issue. This thread is the info that contributed the most to my decision that the 5K iMac is not for me (at least right now, I'm eagerly awaiting rev2).

To those who own it and are happy, that's fine, you have a great machine. But plz accept that it's not the best machine for everyone.

 

[Astelith]

This discussion is half useless, some users put efforts in testing and to collect more data (still nothing to talk about averages or to speak generally) but a lot of others are just trolling and refusing to accept part of the reality (variety of assembly quality between macs) even after video tests, demonstrating a very sad lack of trust in who spent time to have a better understanding of a common big question mark (temperatures).

To be honest I'm truly disappointed

 

[Astelith]

You're too detailed. Think higher level. It doesn't matter why it happens, it happens. Watch the videos.

You find me a 5K M295X iMac that can run a benchmark at sub-100C, I'll eat my words. You won't, though.

As I've said before, macro trends don't just happen based on 5 screenshots. There is a huge amount of noise online about the issue - excuse the pun.

To be clear NO-ONE has said it's bad. We don't know that. It's noisier and some find it concerning, that's all.

https://www.youtube.com/watch?v=lyXbjRJzGeo&feature=youtu.be

I can't be the only one, I'm sure the average is in between and as you just wrote, there's no bad things yet... and maybe never will happen.

The golden rule is : You like it ? Keep it and enjoy, you don't, get it back and peace

----------

It really feels like you're not listening. Your system is NOT cooler. You already proved that with a video! Your GPU temperature is dropping MASSIVELY while your screen is BLACK. Do you not see that? Your video even showed a drop from 99C to 92C in less than half a second after your desktop showed up. What temperature do you think your GPU was at before the screen went black for far, far too long? That's right - 105C, just like the rest of us.

I don't know why we're continuing to go back and forth about this. You already showed us your GPU hits 105C.

The video above is also for you, it seems valley benchmark doesn't have the issue when alt-tab

 

[dyn]

You're too detailed. Think higher level. It doesn't matter why it happens, it happens. Watch the videos.

I am thinking higher level, having the right context and knowing how somebody is doing his tests is crucial. Having a video or screenshot that shows some measurement being taken is just absolutely useless. What are you measuring and how accurate is it? If you can't answer that question the data is useless due to lack of context.

You find me a 5K M295X iMac that can run a benchmark at sub-100C, I'll eat my words. You won't, though.

Since someone here already provided you with such a video...you want ketchup with that?

As I've said before, macro trends don't just happen based on 5 screenshots. There is a huge amount of noise online about the issue - excuse the pun.

Because people are franticly trying to proof they are right instead of trying to find out what is going on. I find the discussion in the Valley benchmark thread here a lot better since people there are actually trying to do that: gather data.

To be clear NO-ONE has said it's bad. We don't know that. It's noisier and some find it concerning, that's all.

Clearly you haven't been reading this thread nor any of the others. There are many people here that claim it's bad. There is a minority which doesn't and that minority (like a lot of minorities) is actively being attacked for seeing things differently. Instead of continuously demanding proof/evidence (and thus acting like a troll) how about asking for more data and information on how they tested and measured things? The only way to cut down that noise you're talking about.

 

[WilliamG]

https://www.youtube.com/watch?v=lyXbjRJzGeo&feature=youtu.be

I can't be the only one, I'm sure the average is in between and as you just wrote, there's no bad things yet... and maybe never will happen.

The golden rule is : You like it ? Keep it and enjoy, you don't, get it back and peace

----------



The video above is also for you, it seems valley benchmark doesn't have the issue when alt-tab

That's a more useful video. I ran the same, and still peaked at 105C. Actually today I saw a peak of 107C in iStat after playing some pretty high-level, busy Diablo 3 (Greater Rift 49, for those wondering, where things can get very graphically intensive).

107C. Good grief...

Can you post some information about your system. From System Report:

Chipset Model: AMD Radeon R9 M295X
Type: GPU
Bus: PCIe
PCIe Lane Width: x16
VRAM (Total): 4096 MB
Vendor: ATI (0x1002)
Device ID: 0x6938
Revision ID: 0x0000
ROM Revision: 113-C773AA-731
EFI Driver Version: 01.00.731

That's mine.

Just curious if there are any differences between iMacs.

Including iStat showing a max temp hit of 107C. MAD.

 

[andy9l]

Astelith's benchmark score was on par with that of a 2013 i5 iMac with a 775m 2GB GPU that peaked at 82°C and, as such, didn't raise the fan from 1200rpm.

https://forums.macrumors.com/posts/20572719/

Does that mean the 780m 2013 iMac outperforms the M295X RiMac?

 

[WilliamG]

Astelith's benchmark score was on par with that of a 2013 i5 iMac with a 775m 2GB GPU that peaked at 82°C and, as such, didn't raise the fan from 1200rpm.

https://forums.macrumors.com/posts/20572719/

Does that mean the 780m 2013 iMac outperforms the M295X RiMac?

Wouldn't surprise me. I've been saying this from the very beginning: The M295X is a huge disappointment, RELATIVE to the previous-year iMacs. That Apple actually charges a $250 upgrade from the 290 to the 295 is... to put it mildly... - hilarious.

I think, probably, that so many months after the 5K iMac release - most of us are still in disbelief about how poor the M295X is. I was just getting over it, too... *sigh*

 

[Astelith]

Here the detail:

Chipset Model: AMD Radeon R9 M295X
Type: GPU
Bus: PCIe
PCIe Lane Width: x16
VRAM (Total): 4096 MB
Vendor: ATI (0x1002)
Device ID: 0x6938
Revision ID: 0x0000
ROM Revision: 113-C773AA-731
EFI Driver Version: 01.00.731

Regarding the score I got variable results from 720 to 755, but I think this benchmark it's not her cup of tea, in the day to day use I cannot complain about performance, at least for the use I need (photo, productivity, video editing,world of warcraft)
And at least we have the same performance, the only variable it seems temps and fan speed, it makes me think about some differences of heat pipes or thermal paste used, just guessing

 

[h9826790]

Astelith's benchmark score was on par with that of a 2013 i5 iMac with a 775m 2GB GPU that peaked at 82°C and, as such, didn't raise the fan from 1200rpm.

https://forums.macrumors.com/posts/20572719/

Does that mean the 780m 2013 iMac outperforms the M295X RiMac?

It depends, for gaming, user experience usually defined by the minimum FPS, not the average.

In this aspect, the M295X is 35% better than the 775m

 

[WilliamG]

Here the detail:

Chipset Model: AMD Radeon R9 M295X
Type: GPU
Bus: PCIe
PCIe Lane Width: x16
VRAM (Total): 4096 MB
Vendor: ATI (0x1002)
Device ID: 0x6938
Revision ID: 0x0000
ROM Revision: 113-C773AA-731
EFI Driver Version: 01.00.731

Regarding the score I got variable results from 720 to 755, but I think this benchmark it's not her cup of tea, in the day to day use I cannot complain about performance, at least for the use I need (photo, productivity, video editing,world of warcraft)
And at least we have the same performance, the only variable it seems temps and fan speed, it makes me think about some differences of heat pipes or thermal paste used, just guessing

Perhaps. It's still interesting that not a single person can duplicate your results.

Either way. 99C or 105C. It's not ideal..

 

[Astelith]

Perhaps. It's still interesting that not a single person can duplicate your results.

Either way. 99C or 105C. It's not ideal..

Well, we are just 2/3 users testing here, hard to make a serious comparison, if I consider the valley bench poll thread the figures are different, assuming all the voters were honest.
For who is concern there's macs fan control with a custom fan profile, you can set it at 95 as maximum, should do the job for free

 

[ioxakoyba]

Well, we are just 2/3 users testing here, hard to make a serious comparison, if I consider the valley bench poll thread the figures are different, assuming all the voters were honest.
For who is concern there's macs fan control with a custom fan profile, you can set it at 95 as maximum, should do the job for free

The problem with Macs Fan Control is that with single fan macs it you can set the fan to be controlled only from single censor, like from CPU temp or GPU temp, but not both.

Even though setting it from GPU may work with the games, but this may cause overheating of CPU on some usage that stress only CPU (resulting that the fan wont increase rpm)

Also it was reported here that setting the fan on 2700rpm wont keep temp under 100 C if the load is constant on both cpu and gpu. Howevee I hope it helps with the normal use, like gaming.

 

[Astelith]

I really want to understand what difference we have between iMacs to have such a big fan difference, it must be something with the air conduct or a different heatsink / thermal paste