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.
2016 Macbook Pro 15" will be a lagfest in Parallels
- Thread starter dark_mark
- Start date
- Sort by reaction score
More like 4 times as many pixels since you are multiplying by 2 on X and Y.
It's like what happened from 1080p to UHD 4K.
[doublepost=1478402425][/doublepost]
Yes, but Apple could have used a different Skylake Core i7 with an Iris Pro 580 instead of the HD 530.
The Iris Pro 580 has about double the performance of the HD 530.
It's ironic that the Macbook Pro doesn't have an Iris Pro, isn't it?
I'm not sure what you mean here. I used to work in a creative field. Not all creative fields used Macs, but Macs tended to remain the standard standard if they were first to infiltrate a given area. Illustrators and prepress professionals tended to be almost unanimously Mac users for a number of reasons. Apple's early system level color management was somewhat better, and the hardware was sufficiently fast. Applications like Shake and Final Cut Pro were also Mac territory. Linux seems like a nice choice for a lot of those users, but it's not great for a desktop environment.
You're right that Apple looked for users with ample budgets, but that doesn't mean cost is a non-issue. They ship several million Macs per year. It's not really a niche product. If you increase the price sufficiently, sales will drop. If it was a complete non-issue, they could have doubled the price.
You're overthinking this. They could have gone the same route as last year, which would have been a 580. Those are more expensive, which means this was most likely a cost cutting measure. Notice how the non-touch 13" is still using a 128GB and even at $2400, the lowest touchbar 15" is still at 256? These are obvious cost controls. 256 fills up fast even for a normal user. Normal users are the ones who don't like to migrate things or clean up their drives.
Not picking on you specifically -- I don't get the hyper-vigilant "money grab" anger prevalent on the forums. Apple has always charged too much for low-spec. The iPhone 7 spec sheet does not compete with Samsung, and neither do Apple's laptop spec sheets compare with Dell or HP. Like, ever. Lately, there seems to be particular hostility toward the capitalistic element within Apple's decision-making algorithms. But they've always been there. In the most vocal cases, I think people just aren't getting their dream gaming laptop. As if Apple owes them something.
Was it really cheaper and more profitable to design a 2-gpu system? That doesn't make any sense, especially since the dGPU upgrades are relatively cheap. I mean, people need to make up their minds -- if Apple was out to gouge them with cheap, aging, underperforming tech, why have a dGPU at all? I have to believe this setup foresaw an MBP life cycle that includes progressively powerful partnerships between the iGPU and dGPU. "Anyone who needs the latest and greatest already has a 2015 MacBook Pro. The best pieces will be cheaper and more available in 2017, but we have to release something before Christmas 2016." I think this configuration (unpalatable to pros) is a subtle communication of exactly this.
I bet 2017 will see a stronger spec sheet. 2016 is an off-year release for the new-and-pretty crowd. Not saying I agree entirely, just saying this makes more sense, and is more consistent for Apple, than "They just want to squeeze another $50 profit out of each laptop. They don't care about performance." They do. It's just 12 months too early to see the potential realized.
I'm buying anyway -- because I'm not really a Pro. Just need something that runs Logic and synth emulators smoothly.
It was a long wait followed by a sharp price increase, 20% in the US and much more elsewhere. Combine that with disappointing specs. The base 15" at $2400 is still saddled with a 256GB ssd, and fixing that is at least $200 more. The only thing cheaper is last year's model, which retained its original price.
Now let's say you're a Windows user. You can jump to another brand or line. With Macs you have whatever Apple produces that round. It can be quite annoying with their recent flippancy. As for me personally, I hate that they added things that will become likely points of failure. Their discrete gpu implementations haven't been very good. Too many of them suffered very high failure rates including the 2008, 2010, 2011, and 2012 macbook pros, the 2011 imacs, and the 2013 mac pros. I've personally experienced these problems, and the refurbished boards don't hold up. The repairs last 6-10 months before failing again. I won't touch one with discrete graphics ever again, which is why I was hoping to see an update on the $2k model. It's still using Haswell.
Apple likes to brag about battery life, so any of their laptops with discrete graphics is a 2 gpu laptop while running OSX (not under parallels or bootcamp). The graphics switching is unfortunately a little buggy. It can stutter while switching, which is ultimately just mildly annoying. If the discrete gpu starts to fail, these things will sometimes crash on gpu switches.
I actually think they'll just do a partial revert. The cpu with iris pro is more expensive. By not using it, they save a little on the touchbar models. They clearly required a high minimum sale there. Unlike 2012, they left the other model with older hardware. Sadly it's still on Haswell.
This isn't about $50 profit. Some of those cpus are quite expensive, and it doesn't make sense to encourage dongle sales as accessories or bar people from upgrading ram for the sole purpose of selling the upgrade. That stuff is merely incidental. They do however consider desired price points and margins. In this case I think they're being overly aggressive, because it's a price increase combined with some (from my perspective) regressive points.
Good point. To be completely honest, I was hoping that Apple would throw something like a E3-1575M mobile Xeon into the new Pro.
Windows 10 Pro flies in VMWare fusion on my 2012 MacBook Pro non-retina so I don't see any reason why a vm would be slower on the 2016 models.
The point of this article is for Retina displays. Essentially, non retina is easier to drive from less pixels. Past Retina models have apparently had problems.
Personally, that's always been my worry about the Retinas. They are putting very high resolution screens that need stronger components to push, yet they are making it thinner and thinner taking away space for the needed components.
Last edited:
But you can force the system to use just the dedicated gpu. I don't get the issue here. If iGPU is slow in Parallels and dGPU doesn't activate automatilcally (pretty sure it will, though) - just force it from Energy settings. What's the problem?
Ahh, ok! I was about to give you crap for something, but now I'm jealous.
I find this interesting too, did the performance in Parallels for the 2012 rMBP 15" continue to be terrible? Or was it just the teething problems at the start? I'm worried we're gonna see bad performance on the iGPU for Parallels or compromise battery life if it was running on dGPU only.
Booting parallels should automatically trigger the switch to dGPU, just like every other 15" macbook pro.
Why? We already know the performance of HD 530, since it's been widely available in a number of laptops for a while now. Same for Iris Pro 5200
I am completely confused by this post. First of all, Parallels uses OpenGL to emulate DirectX
From Parallels website:
"Guest OSes in Parallels Desktop have no access to physical graphics cards present in a Mac. Instead, Parallels Display Adapter driver (which is part of Parallels Tools installation) interfaces with virtual hardware and provides 3D acceleration features. The actual acceleration is achieved by translating DirectX commands from the guest to OpenGL API on OS X side."
This means, basically, the performance is not what you'd call "optimized" for the hardware. So, any number of factors can cause Parallels to lag. Now, when you run Parallels on a MacBook Pro with a dedicated GPU, the dGPU takes over. I am not sure why OP is talking about the integrated GPU. Parallels will use those Radeons, and they will work the same or better than the previous MBP GPUs, depending on the model.
Second, GFLOPS is not everything. Floating point performance is only one aspect of GPU performance. The Iris Pro 5200, for example, has more GFLOPS (832) than the GeForce 750M (722.7), which is the external GPU in 2013 and 2014 15" MacBook Pros. And yet, the GeForce is used in demanding apps, not the Iris Pro.
But it doesn't even matter, because in Parallels, the Radeon will be used. And if, for some crazy reason, it is not - you can always force the system to use only the dedicated GPU with a simple click/tap in the Energy Saver setting in System Preferences. The dGPU is also used exclusively in Bootcamp.
So, again, I don't understand this issue. The only time that HD 530 will be used is while browsing, using email, etc. In fact, based on my experience with the 2013 MacBook Pro, the dedicated GPU is the "default" for apps, unless the developer specifically says the system to use the low-power one. Almost anything I use that is not the most basic of tasks on my MPB activates the dGPU.
From Parallels website:
"Guest OSes in Parallels Desktop have no access to physical graphics cards present in a Mac. Instead, Parallels Display Adapter driver (which is part of Parallels Tools installation) interfaces with virtual hardware and provides 3D acceleration features. The actual acceleration is achieved by translating DirectX commands from the guest to OpenGL API on OS X side."
This means, basically, the performance is not what you'd call "optimized" for the hardware. So, any number of factors can cause Parallels to lag. Now, when you run Parallels on a MacBook Pro with a dedicated GPU, the dGPU takes over. I am not sure why OP is talking about the integrated GPU. Parallels will use those Radeons, and they will work the same or better than the previous MBP GPUs, depending on the model.
Second, GFLOPS is not everything. Floating point performance is only one aspect of GPU performance. The Iris Pro 5200, for example, has more GFLOPS (832) than the GeForce 750M (722.7), which is the external GPU in 2013 and 2014 15" MacBook Pros. And yet, the GeForce is used in demanding apps, not the Iris Pro.
But it doesn't even matter, because in Parallels, the Radeon will be used. And if, for some crazy reason, it is not - you can always force the system to use only the dedicated GPU with a simple click/tap in the Energy Saver setting in System Preferences. The dGPU is also used exclusively in Bootcamp.
So, again, I don't understand this issue. The only time that HD 530 will be used is while browsing, using email, etc. In fact, based on my experience with the 2013 MacBook Pro, the dedicated GPU is the "default" for apps, unless the developer specifically says the system to use the low-power one. Almost anything I use that is not the most basic of tasks on my MPB activates the dGPU.
Last edited:
So performance in Parallels shouldn't be a problem I guess. What do you think of the performance for animations within MacOS like minimising, opening mission control...will the HD530 hold up?
Well, first of all, I obviously don't know until the reviews come up, but I what I can do is give you my best guess.
My guess is that it will work just fine. And this is why:
1. MacBook Pro Retina from 2012 uses an Intel HD 4000 to run the same resolution and the UI doesn't lag. There was some lag here and there, but it was fixed in later OS updates. I checked a few videos on YouTube (one example:
2. You may disagree with this one, but for me this next one is also a good reason to believe it won't lag. I trust Apple to give us a good experience: Sure, there were Apple devices that did lag, I agree. But mostly it was due to either lack of software optimizations or hardware limitations of that time. I seriously doubt Apple would allow their flagship Mac to lag when doing something as basic as UI stuff - when there is more capable hardware available (in other words, they can decide to use less powerful components - as long as it doesn't take away from the experience, but not if it does).
Sure, there is this argument that they value portability over performance and that they would alow their computer to be underpowered to make it thinner and lighter. This is debatable to say the least, but even if it were true - they would never do it in an area where an average user would have a bad experience. And everyone would notice UI lagging.
For example, they will - for whatever reason, not going into it here - use a Skylake chip instead of Kaby Lake because, no matter how much people here whined about it - no one will notice the difference while using the computer. But the UI lagging - that will make the whole computer seem slow. And Apple is all about perception.
3. In my experience, and from what I've read - when the UI does lag, it's rarely, if ever, because of the GPU. Since we associate GPU with framerates and graphics we always attribute lag to it, but in reality, the UI usually lags because of some background process or data collecting.
Last edited:
I think we also have to consider that it is the first time the new dGPUs have less TDP than the CPU. 35W is a big improvement compared to the 50W from last year. Now that the dGPU uses a similar process node to the CPU, I think that the performance per watt is also similar or better than the iGPU at more heavy tasks. Maybe the Iris Pro performance per watt was less than the Radeon Pros and Apple decided to not include it?
I'm guessing something like that - the fact that HD 530 is quite enough for low-power stuff like UI and probably spends less power than Iris Pro. And since every MacBook Pro 15" comes with a dGPU, there is no need for additional power.
The argument for Iris Pro would be to have a version without a dGPU that would arguably spend less power, but as you said - perhaps the power spending difference between the new Iris Pro and the new Radeons is not that big.
While some may disagree with it, there is always logic behind Apple decisions. For example, the fact that every 15" MBP has a discreete GPU allows every 15" to connect to the same number of external displays. Apple can push this feature more easily, instead of explaining to customers that only the high-end versions support multiple 4K displays, for example. And this decision allows them to put less power-hungry integrated GPUs inside.
They can't the Intel roadmap looks like it's dropping their quad core iris pro chips!!