I really hope mine comes early. I was planning on being out of town and had it sent to the Apple Store. It looks like it's shipping now but I was pissed when I went to the store and they have stock of it but wouldn't give it to me. If my original phone calls were true in that they said they didn't have stock when they did, I would never had it ordered the way I did.
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.
CINEBENCH! AMD Radeon R9 M370X
- Thread starter admanne
- Start date
- Sort by reaction score
How? The Mac App Store version 1.5.6 tells me it is out of date and I shall refresh it. But how? The website only seems to offer the Windows version to download directly.
Or am I lost due to still using Yosemite on Haswell hardware and should have switched to San Francisco Golden Gate Park on Skylake hardware?
The link below can give a nice overview of which FPS to expect on a wide range of games as the performance of the M370X is almost the same as the 8870M:
http://www.notebookcheck.net/AMD-Radeon-HD-8870M.86798.0.html
M370X vs HD8870M
3DMark (2013) - Fire Strike Graphics: 2322 vs 2168
3DMark (2013) - Fire Strike Score: 2207 vs 2091
3DMark (2013) - Cloud Gate Graphics: 15409 vs 15738
3DMark (2013) - Cloud Gate Score: 11842 vs 11331
3DMark (2013) - Ice Storm Graphics: 134074 vs 57618
As the 8870 has the same amount of Stream Processors (640) , Memory Bus Width (128bit), and uses GDDR5, the slight difference in favour of the M370X is due to the slightly higher clock (800Mhz vs 725Mhz)
http://www.notebookcheck.net/AMD-Radeon-HD-8870M.86798.0.html
M370X vs HD8870M
3DMark (2013) - Fire Strike Graphics: 2322 vs 2168
3DMark (2013) - Fire Strike Score: 2207 vs 2091
3DMark (2013) - Cloud Gate Graphics: 15409 vs 15738
3DMark (2013) - Cloud Gate Score: 11842 vs 11331
3DMark (2013) - Ice Storm Graphics: 134074 vs 57618
As the 8870 has the same amount of Stream Processors (640) , Memory Bus Width (128bit), and uses GDDR5, the slight difference in favour of the M370X is due to the slightly higher clock (800Mhz vs 725Mhz)
I think it scores quite well compared to the 750m. Its pixel fill rate is double but it has some constraints on shading.
Overall it's an improvement. Keep in mind the 960m has a 75 Watt TDP. Making it a bit too hot to put in these notebooks. The 650m and 750m were both 50 Watt chips.
They could have gone with the 950m but looking at its specifications it doesn't seem all that different to the 650m/750m and I don't think it would provide higher gains than the R9 M370X has.
Of course we'd have CUDA but Apple is backing OpenCL, they authored it after all.
Overall it's an improvement. Keep in mind the 960m has a 75 Watt TDP. Making it a bit too hot to put in these notebooks. The 650m and 750m were both 50 Watt chips.
They could have gone with the 950m but looking at its specifications it doesn't seem all that different to the 650m/750m and I don't think it would provide higher gains than the R9 M370X has.
Of course we'd have CUDA but Apple is backing OpenCL, they authored it after all.
Its because most of benchmarks you find for 950M are for the DDR3 version. Overall, I'd say that the 850M/950M with GDDR5 is around 25% faster then the m370x in real world applications (the gap is bigger in synthetics). But having an equivalent of a desktop HD 7770 in the 2kg laptop is definitively not awful. It will be sufficient for my needs for the next 2 years.
Power draw for GTX970M is 65W alone. Power supply for MBP is 85W. How the ***** they would be able to fit it inside? Especially if MBP CPU is Quad core with over 45W power draw.
Leman: GTX850M should get 4.5K points in 3dMark11. M370X gets 4000.
In games the difference will be bigger, untill... games will move to Vulkan API.
The MacBook Pro is a productivity machine, not a Windows gaming laptop. The new GPU is fit for purpose and has excellent OpenCL performance is OS X.
You want to primarily game, then go buy a Razer Blade, MSI Ghost etc.
You want to primarily game, then go buy a Razer Blade, MSI Ghost etc.
True, but some models have proven to be a decent gaming machine in the past. Apple even advertised Diablo 3 with the rMBP 2012 to show the good performance with the 650m. This move will limit the audience group again to professionals. Games are ubiquitous and a huge market, but Apple somehow thinks they should be played on iOS not on macs.
Don't get me wrong, I've taken my 2006, 2011 and the rMBP to a LAN bash for a bit of FPS gaming.
They're pretty capable, but I game 99% of the time on a dedicated Windows desktop. 99% of everything else gets done on the rMBP. Use the right tool for the job.
They're pretty capable, but I game 99% of the time on a dedicated Windows desktop. 99% of everything else gets done on the rMBP. Use the right tool for the job.
Don't get me wrong, I've taken my 2006, 2011 and the rMBP to a LAN bash for a bit of FPS gaming.
They're pretty capable, but I game 99% of the time on a dedicated Windows desktop. 99% of everything else gets done on the rMBP. Use the right tool for the job.
Perhaps due to cost? Upgrading a Mac and gaming rig every year or two is an expensive proposition.
Dunno, about anyone else, but I buy for a five year lifespan. The only reason I'm rocking an rMBP is that the 2011 repeatedly failed and Apple replaced it.
Same for the gaming rig. I tend to upgrade the graphics card after a two/three years while the core hardware stays the same.
Same for the gaming rig. I tend to upgrade the graphics card after a two/three years while the core hardware stays the same.
Don't base the 950m performance just on specs alone. Maxwell is a very different architecture than Kepler that's used in the prior 750m chip. On real world gaming benchmarks, the 950m smokes the hd8870m that's basically the same chip as the rebadged m370x. In most gaming benchmarks it's 30-40% faster which is huge on the same 50 watt power envelope as the m370x. Even with compute, maxwell is much faster than Kepler and close to the performance of amd. So basically either AMD is giving these 3 year old chips for almost free or Apple is retaliating against that stupid nvidia lawsuit.
I think the saddest thing of all is in the last 6-7 years of MacBook pros, Apple has put in the best gpu they could based on heat/size constraints of the time. This time Apple went with the cheapest or most politically expedient choice.
Last edited:
I really think this is the answer, so when in october they launch broad well MBP with the NV960 they can market it.
As long as the profit margin is fine, shareholders are happy. Apple even got away with offering the dGPU only in the high end model. I think in the long run they will probably loose money on this (I explained this problem in more length in the news thread about european price increases). If they continue to overcharge for the "mediocre" GPU in the 15'' or spinning drives in a 2k iMac, customers with a practical attitude will move to different devices. E.g. a base 13 MBA/rMBP + windows desktop/notebook combination, which means Apple will get 1099-1449€ instead of 2799€. And customers, who aren't locked in the ecosystem anymore could consider windows alternatives for their next purchases.
Last edited:
In all fairness Maxwell is updated Kepler architecture. It has just bigger internal caches and lacks DP engines, thats why its so efficient.
Also, I believe Apple went with GCN architecture because its more mature in fact more efficient than any Nvidia Arch(Green500 list says that GCN is the most power efficient GPU arch. that has ever been on this planet, and lets look how big performance Apple got from 129W of power on Tahiti chip in Mac Pro) and will last longer compared to any Nvidia Architecture.
Lets look how fast Kepler arch gets old and outdated in games(Witcher 3, Project Cars). GCN with time gets more and more performance, which is not and will never be the case of any Nvidia arch(The way its meant to be milked...).
Also the factor of OpenCL performance is biggest thing and most important for Apple. Not to mention Mantle/Vulkan, which i believe soon will be flagship API for OSX.
It is so efficient because it rearranges the way how code is executed on the processor. With Kepler, you had to process large batches of data in order to get maximal utilisation of the ALUs. So when processing smaller tasks (which is very common both in graphics and compute tasks), Kepler was never able to perform as good as it should. Maxwell can combine small batches to utilise the computational resources much better. This is essentially the same approach AMD took with GCN as opposite to their earlier VLIW architecture and also what Intel has done with their GPUs (Intel's GPUs are probably the most efficient at utilising resources). You can compare it to hyperthreading (albeit on steroids) in Intel CPU's.
I don't understand how anyone can make such a conclusion. Maxwell cards have the same performance as AMD GCN card at 50% increased power draw. Just look up benchmarks and reviews. For example, a GTX 960 (~100W under load) is very similar in performance to r9 285 (~160W under load).
In games, yes, Maxwell is very efficient. Compute power draw is totally different.
4.6 TFLOPs GTX980 gets WITHOUT clock and power draw cap. Non-Reference cards for example while processing OpenCL can draw 277W which Tom's Hardware in their review have proven. If we have clock cap, and power draw with reference Bios in OpenCL you get 185W of power draw, but TFLOPs performance drops dramatically.
GPU is as it states: General Purpose Unit. If we will judge architectures by the way they can handle games we are getting to ridicoulus assumptions.
Well, Maxwell still performs better in these games. The GTX 960 is virtually identical to Tonga r9 285 again while drawing over 60W less of power. The reason why Kepler struggles (if this is even the case) is probably what I mentioned above it is not very good at parallel execution of multiple smaller tasks. And as games get more sophisticated and shaders get more complex, its the hardware flexibility that one needs. Maxwell and GCN have much better performance in complex scenarios, exactly because they have more internal flexibility.
----------
Could you link me the benchmark you base it on? Because all results I have seen show that 980M obliterates the AMD cards in compute, while drawing less power. Unless you are talking about FP64, it is true that nvidia cards were not really made for high-performance double-precision computations.
Yeah, I like to do both. Until about 2013 the MBP was a great way to do that.
So I ordered a Blade last night from MS. i7 quad, QHD screen, 16 GB, 256 GB SSD (standard interface easy to replace) with extended warranty for $2400 CDN taxes in.
If I like it and decide to keep it (heat, noise, battery life), I'll sell my Y50 and Surface Pro 3 - combining my business and gaming machine into one.
Ummm.... like no. GPU has never stood for General Purpose Unit. It's an acronym for Graphics Processing Unit.
Source:- http://en.wikipedia.org/wiki/Graphics_processing_unit
Seeing as my 750m iMac runs Crysis 2 at 60fps 1080p on very high, of course this will.