tht

brian green said:

The question I have is a technical one, and one I don’t have any knowledge about, but I’m sure many members of this forum do.  We know that Grand Central Dispatch is a part of MacOS and that tasks can be divided up among multiple cores to speed up workloads.  The question I have is, is it possible to use this feature to have an older chip in the MBP with more cores that can complete the tasks faster than the newer chips because the older chip produces less heat?  

No, it’s not possible, for a multitude of reasons. In general, more performance means having more transistors, higher clocked circuits, and having tasks that can actually be made parallel. More cores means more transistors means more power is needed to drive them, and you’ll end up in the same spot.

A CPU vendor can notionally design a system that is more power efficient than today’s Intel x64 systems, like perhaps with Apple and ARM, but that introduces a whole new set of transition and compatibility issues that market may not like. And designing a system that has more performance than Intel’s is not a trivial exercise, especially on a year by year, decade by decade business.

Another thing I’ve wondered about is whether Apple could integrate the A10 Fusion (or whatever’s next) into the laptop in order to complete tasks in a cooler manner than the Intel chip?  We already know that the A10 Fusion is able to handle 4k video editing, because we can do it on the iPhone and iPad.  Wouldn’t it be possible to use the chips in tandem to complete work faster, creating less heat?

The notion of co-processors, targeted chips/ASICs is already present in Intel systems. The newer Kaby Lake and Coffee Lake hardware have special execution units in them that do video encode and decode such that the MBP with 4 CPU cores can be faster than the Skylake base iMac Pro with 8 cores. The Skylake CPUs in the iMac Pro have special execution units in them that are either fused off or not present in Coffee Lake CPUs that’ll make them faster at certain tasks. So, the technique of using specialize chips to do certain tasks more efficiently is done all over the place in computers today.

Lastly, more performance typically means more power is needed to drive them. There isn’t a free lunch. Apple may have a special chip that does this or that specific task faster or more efficiently, but there are not any low hanging fruit left.

I have to be honest in saying that I wish Jony Ive would be replaced by an Engineer that actually shoots video all the time, edits video all the time, and actually cares about that functionality rather than it looking “pretty” with chamfered edges.  The pendulum has swung so far toward “looks” that the “functionality” side seems to be ignored because it gets in the way of “pretty”.  Apparently, nothing was learned from the Mac Pro fiasco.

Jony Ive wasn’t the problem. The Mac Pro issue was in Apple’s marketing team not understanding where PC computing is going. For laptops and iMacs, the meat of their PC lineup, they are doing fine. All the machines are either class leading or class competitive. Apple is definitely generating “opinionated” designs, not safe designs as it were. With that, comes lots of complaining.

fastasleep said:

I'm still confused about this lineup. I don't understand how they're going to market the mid-sized LCD model in comparison to the other two. It appears that it'll have slightly thicker bezels, but what other differentiating factors will there be that separates it from the OLED models to the average consumer? I don't think most people are going to know or care about the difference between LCD and OLED. Plastic body like the 5c?

The LCD model is rumored to have a single back camera, no 3D Touch, aluminum frame instead of steel, the aforementioned larger bezels, and the usual segmentation in storage and maybe SoC. It will look less “pretty” for lack of a better term. 

No 3D Touch stings imo. That maybe implies no Taptic Engine, or a lessor version of it. 

tallest skil said:

I wonder what powering this thing directly from your laptop does to the battery life, though I guess if you’re somewhere that can fit both your laptop and this display in front of you, you’re also somewhere that your laptop can be plugged in.

Based on the MBP15, which advertises about 10 hrs with a 70 WHr battery, probably 5 W at medium brightness. Intel’s active idle enables the CPU to average less than 2 W for most lightweight tests.

The rest is just math. It would reduce the runtime of a MBP15 to about 70/12 ~= 5 to 6 hours, or 40 to 50%. Only gets worse with MBP13 and MB12.

It has to be one of those situations like being at a camp or conference type of thing. Maybe a quick one hours of work at a library or coffee shop, maybe airport?

cgWerks said:

tht said:
Comparing the current Core m3-7Y32 in the base model MacBook (Mid 2017) to the Core i5-5350U in the base model MacBook Air (Mid 2017). The MB is about a 10% faster. I expect the GPU to be a little faster as well.

Interesting... I'll have to look into that. That flies in the face of what I thought I knew. That isn't some kind of peak speed until the chip throttles back, is it?

My friend who is pretty on top of this stuff, and has and loves his MB recommended my wife get a MBA because the MB just isn't up to the heavier tasks. That was a year or so ago, but has it changed that much?

Yea, if the MB had more ports (and especially TB3) I'd consider it instead of a MBA if that's true. (Except for the keyboard... the MBA has a better keyboard.)

Turbo speeds between the two are about the same, but Kaby Lake is a little faster per clock than Haswell, and has hardware decode for HEVC and HEIF. For the vast majority of the people who buy a MB or MBA, the MB should perform a little better, and in the case of HEVC and HEIF, a whole lot better.

If some one was running a process that exercised the CPU, like a computational simulation, the Core i5 will be a little faster as its base clock is 1.8 GHz vs the 1.1 GHz of the Core m3. These folks usually buy the MBP though.

A similar thing is going to happen with the 4-core Coffee Lake. The other thread has the XPS 13 scoring some healthy multi-core GB4 points over the 2-core MBP due to its 4 core architecture. But if you run a long running computational simulation, the MBP may end up performing about the same because the base clock is at 1.8 GHz on the 4 core while it is 2.3 GHz on the 2 core, whatever the actual frequencies are. Any thing that heats up the CPU, will cause them to go down to their base clocks, and 4x1.8 GHz isn’t going to be much faster than 2x2.3 GHz.

Would be interesting to see some actual testing.

Our MacBook Pro is also thicker at 0.59 inches while the XPS is 0.46 inches at its thickest point, if you omit the feet like Dell does for its measurement

Did you actually measure the thickness of the Dell?

When I look at this picture:


The MBP isn’t 0.59/0.46 = 1.28x thicker than the XPS13 here. If anything, they look about the same thickness, with perhaps the Dell being thicker depending on the curvature. Since you say if don’t include the feet, the Dell XPS is thinner. This means that when laying on the table, the XPS13 is higher than the MBP? If so, how do you know the feet of the XPS13 are thicker or taller than the feet of the MBP?

After some basic photogrammetry, the 0.45” dimension looks to be the thickness on the very back face of the XPS13, and it’s not the actual thickness of the device. You can clearly see the XPS 13 is bulging thicker as you move away from the back Face.