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.
Are they any Radeon Pro 460 benchmarks out yet?
- Thread starter drvelocity
- Start date
- Sort by reaction score
A 2011 game - even in its 2016 Special Edition which is not a proper remake : it's still the same engine under the hood with a new light system and a set of new shaders - should not be the bench to judge gaming with the 460.
Last edited:
It isn't PROFESSIONAL grade anymore, its just a prosumer product like I said. I guess you can say its still Pro, just a grade lower now since it hasn't kept up with the times.
I do network engineering and the cloud software/packages/etc for cloud controllers and such can take up a ton of RAM - production will of course not be on a laptop but during initial testing and setup there's a need to spin up various VMs quickly so 32GB of RAM is one of those major items. And a lot of typing - so the new keyboard just isn't optimal. If they had made it just a little thicker, the butterfly mechanism is fine, it would allow for a bit more key travel.
Stupid things like that in the new design - compromises so they can be anal and meet stupid targets like that exactly 4lb weight target and make it as thin as possible? Why? Would 1 mm thicker and 4.2 lbs and a few minutes less battery life without low power RAM destroy the machine? Absolutely not. And could it just be made a bit thicker and heavier and max out with a bigger battery to compensate more powerful options in the machine?
Prosumer grade machine - that's what you get now. In some ways it still fails to meet this new audience they're trying to cater to - you have people here asking how the machine deals with games. Compared to the Nvidia offerings, the GPU is a weakling. So even for the prosumer crowd, they missed the mark in some regards.
They should have differentiated the 15" at least, its ok if the 13" is aimed at the prosumer market. Just make the 15" really really high-end and NO compromises professional grade material.
Well about your need for 32gb RAM maybe ask intel about that cause the intel processors used only support up to 16gb of ram. Next year I would assume apple will use the latest intel generation which then supports up to 32gb and then maybe the will make 16gb standard and an upgrade to 32gb.
That's not true. All the relevant Skylake chips support 64GB RAM. However, Apple uses LPDDR3 (low-power DDR3) and not DDR4. They did it - as was clarified by them - to conserve energy/meet the battery life.
And at this moment the 7th generation chips needed for the MacBook Pro line are not out.
There is nothing here that Apple has done that isn't keeping up with the times, or in keeping with previous design guidelines of their laptops. They are demonstrably closer in overall performance to a desktop computer than ever before, and they are using the most powerful video cards available in anything close to the power envelope that fits those design goals.
What laptops would you suggest get the overall balance better than the new MacBook Pro's? You're an engineer - give us some specifics.
See this just makes no sense!
There has never been a Macbook with 32 gb of RAM. Look it up on Everymac.
And here you're claiming it's "not pro anymore".
By your standards, it has been "not pro ever". Which is fine, but for many, many professions, it has never been up to scuff, until now. We can finally use a Macbook for what we do.
You're just trolling, period, end of story.
Btw is there some comparison around like a benchmark chart for all the different cpus used + gpus?
I am currently indecisive what to get. Untill now I always went with the 13" but I might pull the trigger on a 15" this time so I guess best value for money is the 15" with i7 2,6 and the 460 and for the 13" probably i5 3,1.
But is there some benchmark incoporating all the different cpus etc? I have only found bits and pieces so far.
I am currently indecisive what to get. Untill now I always went with the 13" but I might pull the trigger on a 15" this time so I guess best value for money is the 15" with i7 2,6 and the 460 and for the 13" probably i5 3,1.
But is there some benchmark incoporating all the different cpus etc? I have only found bits and pieces so far.
He is voicing out his own preferences. That is not trolling.
Be careful, though - it's obviously easier to make thinner laptops with better performance as technology evolves in all the sectors. You are right, if you put the caveat "within the power envelope/design goals" then your argument is fairly strong. But, I think at the price - which went up for all models essentially - should be a little more convincing especially when we're talking about the $3-4K mark.
I think the 16GB RAM decision which was made to use the LPDDR3 memory to fit the thin profile wasn't the best move for a professional machine. The graphics card isn't terrible, but there are tons of options, especially with NVIDIA currently the favored by most and the unknown nature of the AMD chips before the launch of the MBP's.
Those two things I think annoyed people the most by far - I don't think it's the change to four TB3 ports, or the Touch Bar - simply not having the option on the highest end machine in the MBP line is annoying to many.
I personally use a MBP for all kinds of intensive tasks - video editing and rendering, dual monitors, many inputs, a ton of programs, etc. I've never had an issue with RAM. However, again, I see the need to future-proof at the price level. Many people were waiting a long time for these machines, and the Touch Bar/Skylake chips were of course expected.
I think people expected the higher-end of the Polaris-GPU line for laptops given the ton of other options (albeit at higher power) - there's just so many that it's easy to see why people are upset.
If you buy an iPhone every year, you're basically spending close to - if not $1,000 on a phone that you'll be replacing likely much sooner than a laptop. That's the premium you pay for Apple. It's nothing really new.
At the end of the day people here are likely much more power users than Apple's market as a whole. A college student wants something thin and sleek. While we can argue that it's unfair, Tim Cook was the COO of Apple for the majority of his time there and is a businessman. If profits are good then that's what they'll follow. I am sure they sold a ton of these and their projections for revenue next quarter are high.
Will be interesting to see what they do with the iMac and/or Mac Pro as those don't really cater to the same market and have more room to expand.
I second this, I want to get the 15" with the 460, and would like to go for the 512GB SSD, but I think it might get too expensive if I also get the 2.9GHz processor (damn Spanish taxes)... Would be nice to know how much I'd be missing out by not upgrading the processor
But those expectations are pie in the sky unless you completely ignore Apple's consistent design goals for their laptops.
Right, that's why I took a neutral stance. I understand both sides. People will say that they didn't have to make it thinner. Not many people here will care. I personally don't - the original Retina MBP 15" was fine for me when it comes to profile.
I am also aware that Apple's goal is to do precisely that - and they have a market of many non-power users who buy MacBooks. Many are college students, for example. Also, Apple is clearly trying to make one line of MacBooks: the Pro, and then have their much smaller MacBook line. That way most people who want a computer from Apple will end up buying something from the MacBook Pro line.
It's a double-edged sword. Nothing will make anyone happy; Apple is a large company that cares about profit. I expect it to. But I see why people are upset here as well.
Last edited:
Here's why you're wrong.
GTX 1060 power consumption -- 120W, as quoted by Nvidia: http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-1060/specifications
GTX 1060m Power consumption -- While Nvidia doesn't disclose these numbers officially, laptop manufacturers place it around 80W. So, 30% less power consumption than the desktop card, and 10% less performance, According to this review: http://www.pcgamer.com/msi-gs73vr-reviewgtx-1060-goes-mobile/
The GTX 1050 should follow suit: 75W in desktop card mode, and 50W for mobile. It's also pegged at 1.8 Tflops in the desktop card, and the mobile version will get you around 10% less, so around 1.62 Tflops.
The Radeon Pro WX 4100 (Radeon Pro 460's bigger brother) gets you **2.4 Tflops at 50W** power consumption in the desktop version of the card. The Radeon Pro 460 gets 1.86Tflops at 35W in a laptop.
So... just looking at Tflops because that's about the only way to compare this... **For the First Time** AMD has a higher Perf/Watt than Nvidia.
No, I understand people saying they hate the laptop. That's fair. But this person is insisting the MBP line suddenly became no longer a professional grade product, when it is clearly superior in virtually all aspects (arguably keyboard aside) then all previous machines. So to say, I need 32gb of RAM therefore this product is no longer "pro", well that makes no sense because the old ones don't have 32gb of RAM. In fact their RAM is considerably slower than the current model.
I can totally buy the argument "these products have never been pro in any way", which was my personal view about MBPs until this model - i.e. I never for a second would have even considered a laptop an option at all. But to say, the other ones that came before were pro, but this isn't, it's just plain silly on the face of it.
Not a huge difference. They're charging basically what Intel suggests. There's always a decent price hike in the highest chip Apple offers which is usually something like a 2.9 vs 2.7 thing. The suggested price difference from Intel is $200+
http://browser.primatelabs.com/v4/cpu/search?dir=desc&page=1&q=MacBookPro13,3&sort=score
The 6920HQ i7 @ 2.9 is there. If you scroll past page 10 or so, you'll start seeing the 6820HQ @ 2.7.
It's not a big difference at all. If you have the extra money, go for it. Otherwise, no big deal.
From Reddit
3DMark Time Spy
Overall: 1528
Graphics: 1374
CPU: 4210
http://imgur.com/RvDJ6Yg
3DMark 2011
Overall: 6973
Graphics: 6706
Physics: 9279
Combined: 6492
http://imgur.com/7sO0zSD
Geekbench4 CPU (Windows 10)
Single-Core: 4610
Multi-Core: 14217
http://imgur.com/kmiJzz7
Geekbench4 OpenCL (Windows 10)
Score: 52389
http://imgur.com/TzNzeRx
Games:
The only game I have is Fallout 4. I tried a few different settings, nothing scientific though. 1080p Ultra went between 22-32 FPS, averaged 27-28, running on an external 4K monitor while also powering the laptop's display. I purposefully found areas with tons of geometry to tax the system.
LINK: https://www.reddit.com/r/apple/comments/5d7xs3/2016_15_macbook_pro_i7_29ghz_pro_460_512gb/
3D mark 11 score is on par with AMD Radeon R9 M290X
3DMark Time Spy
Overall: 1528
Graphics: 1374
CPU: 4210
http://imgur.com/RvDJ6Yg
3DMark 2011
Overall: 6973
Graphics: 6706
Physics: 9279
Combined: 6492
http://imgur.com/7sO0zSD
Geekbench4 CPU (Windows 10)
Single-Core: 4610
Multi-Core: 14217
http://imgur.com/kmiJzz7
Geekbench4 OpenCL (Windows 10)
Score: 52389
http://imgur.com/TzNzeRx
Games:
The only game I have is Fallout 4. I tried a few different settings, nothing scientific though. 1080p Ultra went between 22-32 FPS, averaged 27-28, running on an external 4K monitor while also powering the laptop's display. I purposefully found areas with tons of geometry to tax the system.
LINK: https://www.reddit.com/r/apple/comments/5d7xs3/2016_15_macbook_pro_i7_29ghz_pro_460_512gb/
3D mark 11 score is on par with AMD Radeon R9 M290X
https://forums.macrumors.com/thread...nchmarks-out-yet.2014822/page-7#post-23924004
Read this post. 10% slower, you say?
One thing to note is that the reviews are subjective. NVIDIA and AMD have different topologies. Looking at OpenCL gets us a decent benchmark. If you look at CUDA cores NVIDIA will usually blow AMD out of the water. If you look at specific things one chip will do better because of the instruction sets.
That is not bad at all.
It all due to software. It has nothing to do with hardware, but software optimizations. You do not need to optimize your software for CUDA, you just put CUDA into your software for GPU acceleration. For openCL - you have to optimize your software for hardware.
If AMD GPUs would be able to run CUDA code(they cannot because it is proprietary API) they would run it in comparable way. Because there is nothing in compute in Nvidia hardware that would make it faster than AMD. Compute is just mathematical algorithms executed on hardware, and exposed in TFLOPs performance levels.
Thanks for the reply. In my case, it's a 350€ difference from the 2.6 to the 2.9... I wonder if in the long run (around 5 years from now) there will be a noticeable difference in performance between the two options
Not a big jump at all from 2.7 -> 2.9. Slightly more from 2.6 to 2.9 due to the slightly higher L3 cache on the 6820HQ/6920HQ's, but not that bad.
The 2.7 15" config might be a good middle ground.
Mmm, why am I wrong? I said basically the same what you did, with the only difference that you made a better argumented story
