MacBook Pro CPU getting slow. Looking to upgrade, please help

[jdon]

Hi, i have a Retina MacBook Pro Mid 2015 Edition. It has been sitting on my desk; shut and connected to my Samsung UltraWide 4k monitor. I haven't really had many issues with it up until recently. Basically, the CPU is almost always above 80%, sometimes 98% or even 100%. It's making my charting software slow down and i'm experiencing lag. I'm aware that charting software such as MetaTrader or NinjaTrader can me quite heavy on the CPU, but come on... i mean, a machine for this price? I think that i should have no problems running multiple charts and still be able to do stuff in the background. I think that the issue is my external monitor. Its native resolution is 3440x1440 which is higher than the retina's res.

I was looking around and i found a laptop i could buy, here it is:

[image deleted, competitors' links are not allowed. The spec was basically Intel i7-6700HQ, nVIDIA GTX 1060 6GB,
16GB 2133 DDR4 RAM, 15.6" HD IPS screen]

Would this new laptop be significantly faster or would we be looking at around the same speeds? I know that the GPU on that Asus will absolutely kill the mac but i'm not sure about the CPU, which is of course what i need the most. Any help guys? Thanks a lot in advance!


Here are the specs of my MacBook:

 

[ubuysa]

I'm afraid I've removed the Asus laptop image you posted, and the attachment because these fora belong to PC Specialist who sell custom built gaming PCs and laptops. I've left the basic laptop spec in so that one of our more knowledgeable people can spec you up a PCS laptop that will meet or exceed those specs.

 

[Oussebon]

In terms of PC Specialist options for something with an i7 6700HQ etc you may be looking at one of the Defiance III models with the new i7 7700HQ processors, which may offer a noticeable boost over older generations. These are some benchmarks to give a rough idea of how much faster it might be in CPU intensive tasks: http://tinyurl.com/h3rh38c

When you say your CPU is at 100% I take it you mean that literally - 100% load on all 8 threads?

In which case you may be better off looking at a desktop PC. While the 7700HQ should be a bit of an upgrade over what you already have it may not have a huge amount of headroom. Desktop CPUs are simply more powerful. If a desktop PC is viable you may even consider a 6-core Intel CPU, assuming your software does indeed benefit for multiple cores over higher frequencies and IPC. If space is an issue there are mini PCs that can have CPUs like the i7 7700k.

If it must be a laptop, there is the Octane III family of PCS laptops has desktop processors upto the i7 7700k available.

If you want to suggest your budget and what kind of system would be okay (Desktop > mini PC > Octane laptop > normal laptops) users can suggest some options and you can go from there.

 

[jdon]

Hi, thanks for your response. As soon as i open up all the software i need, this happens

The issue really becomes an issue when i try to screen share while hosting a webinar or even have too many charts and a win10 vm running in coherence mode in the background.

In terms of PC Specialist options for something with an i7 6700HQ etc you may be looking at one of the Defiance III models with the new i7 7700HQ processors, which may offer a noticeable boost over older generations. These are some benchmarks to give a rough idea of how much faster it might be in CPU intensive tasks: http://tinyurl.com/h3rh38c

When you say your CPU is at 100% I take it you mean that literally - 100% load on all 8 threads?

In which case you may be better off looking at a desktop PC. While the 7700HQ should be a bit of an upgrade over what you already have it may not have a huge amount of headroom. Desktop CPUs are simply more powerful. If a desktop PC is viable you may even consider a 6-core Intel CPU, assuming your software does indeed benefit for multiple cores over higher frequencies and IPC. If space is an issue there are mini PCs that can have CPUs like the i7 7700k.

If it must be a laptop, there is the Octane III family of PCS laptops has desktop processors upto the i7 7700k available.

If you want to suggest your budget and what kind of system would be okay (Desktop > mini PC > Octane laptop > normal laptops) users can suggest some options and you can go from there.

I see. Thank you very much for clearing that up and as to your question, i am not quite sure. Let me show you what i mean.

 

[rt1707]

Hi Jdon,

The setup you have listed will likely give you a worse performance than your Macbook Pro (e.g. look at these benchmarks: http://cpuboss.com/cpus/Intel-Core-i7-6700HQ-vs-Intel-Core-i7-4870HQ ). The graphics hardware in your Mac can easily run a 4k monitor, so that is also not the issue either. I'm a big fan of PCSpecialist but the recent Macbook Pros are really great, and if you have one I see no value in buying another laptop from PCS unless you specifically want a portable gaming system.

As for your performance issue, this is likely to be either memory usage or CPU issue (I'm unfamiliar with forex). In either case you should get a desktop as suggested above. It will be cheaper and you will get much better performance.

 

[BlessedSquirrel]

Hi Jdon,

The setup you have listed will likely give you a worse performance than your Macbook Pro (e.g. look at these benchmarks: http://cpuboss.com/cpus/Intel-Core-i7-6700HQ-vs-Intel-Core-i7-4870HQ ). The graphics hardware in your Mac can easily run a 4k monitor, so that is also not the issue either. I'm a big fan of PCSpecialist but the recent Macbook Pros are really great, and if you have one I see no value in buying another laptop from PCS unless you specifically want a portable gaming system.

As for your performance issue, this is likely to be either memory usage or CPU issue (I'm unfamiliar with forex). In either case you should get a desktop as suggested above. It will be cheaper and you will get much better performance.

He says he has a 2015 MBP which has a 4th gen i7 (i.e. 4 years old) slower ram, and much poorer graphics. The GTX 1060 would walk all over the current MBP graphics let along the 2015 model.

 

[Oussebon]

(e.g. look at these benchmarks: http://cpuboss.com/cpus/Intel-Core-i...Core-i7-4870HQ ).

Don't use CPU Boss and GPU Boss. It compares junk numbers to give frequently bogus information.

Like the time GPU Boss said an R9 380 was better than an R9 380x for gaming. For pete's sake...


The notebook check benches are more reliable tbh, even if they are affected to a degree by the performance of different laptop chassis at least that info is indicated. How a Crystal Well (Haswell) CPU at 3.7GHz boost is supposed to outperform a Skylake CPU at 3.5GHz in single threaded tasks by 25% - which is what that CPU Boss page implies - is entirely beyond me.

Turning back to the OP - I'm not familiar with MACs so am not sure I'm reading the info you've displayed right but the screenshots look like they're saying your GPU is 84% idle? In which case I'd have pointed the finger at RAM as well. If the red zigzag like is something you've drawn in to show what it's like when really under load then perhaps it is the CPU and a desktop with a powerful processor is the way to go.

 

[rt1707]

Don't use CPU Boss and GPU Boss. It compares junk numbers to give frequently bogus information.

Like the time GPU Boss said an R9 380 was better than an R9 380x for gaming. For pete's sake...
View attachment 9602

The notebook check benches are more reliable tbh, even if they are affected to a degree by the performance of different laptop chassis at least that info is indicated. How a Crystal Well (Haswell) CPU at 3.7GHz boost is supposed to outperform a Skylake CPU at 3.5GHz in single threaded tasks by 25% - which is what that CPU Boss page implies - is entirely beyond me.

Turning back to the OP - I'm not familiar with MACs so am not sure I'm reading the info you've displayed right but the screenshots look like they're saying your GPU is 84% idle? In which case I'd have pointed the finger at RAM as well. If the red zigzag like is something you've drawn in to show what it's like when really under load then perhaps it is the CPU and a desktop with a powerful processor is the way to go.

Yes, there probably isn't a 25% difference but CPUboss page basically just shows what other pages show... that for an older architecture, the 4870HQ ****es all over the 6700HQ e.g. http://www.hardwareluxx.com/index.p...6666-tested-intel-skylake-core-i7-6700hq.html http://www.cpubenchmark.net/compare.php?cmp[]=2314&cmp[]=2586 http://cpu.userbenchmark.com/Compare/Intel-Core-i7-6700HQ-vs-Intel-Core-i7-4870HQ/m34954vsm14314

My point is that there is no advantage to switching over to the 6700HQ. He is using his Mac as a desktop replacement it seems, so he is much better off actually getting a desktop. And his Mac is almost certainly a nicer laptop than anything he could buy anywhere else (but also disproportionately expensive for what it is).

 

[rt1707]

He says he has a 2015 MBP which has a 4th gen i7 (i.e. 4 years old) slower ram, and much poorer graphics. The GTX 1060 would walk all over the current MBP graphics let along the 2015 model.

Yes, the 4870HQ is 22nm based rather than 14nm but... so what? Unless he plans to install 64Gb of RAM into his laptop there is really no advantage to the 6700K. There appears to be no performance advantage. For the OP's purposes, the GTX1060 does not walk all over the MBP graphics because both will run his single monitor equally well. His problem is clearly not related to his graphics card. He isn't gaming or video editing, and it is probably unlikely that his forex software is using the GPU for parallel processing purposes. Why should he waste his money on an expensive graphics card? It would make sense if he wants to game a bit on this side, but he doesn't say this.

 

[Oussebon]

6700K

I take it you mean HQ. Best to be clear to avoid confusion. Because if you're saying a 6700K would be no better than a mobile CPU from 2013 you need to get your facts straight. (well, 2014 release but afaik Crystal Well was basically Haswell with a better IGPU).

There appears to be no performance advantage.

For the 6700HQ that's only true if you use junk websites like CPUboss. See my post on that. And the notebookcheck benches I linked earlier.

 

[rt1707]

I take it you mean HQ. Best to be clear to avoid confusion. Because if you're saying a 6700K would be no better than a mobile CPU from 2013 you need to get your facts straight. (well, 2014 release but afaik Crystal Well was basically Haswell with a better IGPU).

For the 6700HQ that's only true if you use junk websites like CPUboss. See my post on that. And the notebookcheck benches I linked earlier.

Yes, that is a typo. The 6700K is rather better

I would be interested to see benchmarks that put the 6700HQ as much or any better than the 4870HQ. The only ones I have seem put them on a par, or the 4870HQ on top.

The point I am trying to make is that there is no point to the OP upgrading his processor to the 6700HQ, and should consider a desktop (in which case the aforementioned 6700K may be good option!)

 

[Oussebon]

I would be interested to see benchmarks that put the 6700HQ as much or any better than the 4870HQ. The only ones I have seem put them on a par, or the 4870HQ on top.

Like I said repeatedly, I already posted a link to such benches.

In terms of PC Specialist options for something with an i7 6700HQ etc you may be looking at one of the Defiance III models with the new i7 7700HQ processors, which may offer a noticeable boost over older generations. These are some benchmarks to give a rough idea of how much faster it might be in CPU intensive tasks: http://tinyurl.com/h3rh38c

And if the only benches you've seen are on CPUboss you're better of disregarding them as already explained. You can't trust those numbers to mean anything....

I certainly agree that the difference probably isn't worth the cash to upgrade to most people and that a desktop is a better solution, but it's easier to stick to that line of discussion rather than linking to really bad info.

I would be interested to see benchmarks that put the 6700HQ as much or any better than the 4870HQ. The only ones I have seem put them on a par, or the 4870HQ on top.

The point I am trying to make is that there is no point to the OP upgrading his processor to the 6700HQ, and should consider a desktop (in which case the aforementioned 6700K may be good option!)

Although both have been superseded by Kaby Lake, which would be a rather better option, especially at stock.

 

[rt1707]

Like I said repeatedly, I already posted a link to such benches.


And if the only benches you've seen are on CPUboss you're better of disregarding them as already explained. You can't trust those numbers to mean anything....

I certainly agree that the difference probably isn't worth the cash to upgrade to most people and that a desktop is a better solution, but it's easier to stick to that line of discussion rather than linking to really bad info.

Although both have been superseded by Kaby Lake, which would be a rather better option, especially at stock.

I can't see the links you put that show the 6700HQ is much or any better than the 4870HQ. The various benchmarks I have seen - real world and theoretical - all put the 4870HQ as even or better than the 6700HQ. I have not just seen the CPUboss benchmarks, I glanced at half a dozen site after you called into question what I said and they all support my opinion.

I'm enjoying chatting with you, but in terms of the OP's question, you can't seriously be suggesting that using a 6700HQ, or even 7700HQ (Kabylake is a minimal upgrade over Skylake if we are honest) is going to solve his issue. That would be terrible advice! It is fairly clear it is either a memory issue or he needs a substantial improvement to his processing power - which can only be provided by desktop level processors. He would be wasting his money getting another laptop.

I dislike Apple for a whole variety of reasons... but his Macbook Pro is a great laptop and not the source of his problem here!

 

[Oussebon]

Try reading, then. There's a particular difference in multithreaded performance, which, if the OP's CPU is actually at 100% load, would be an asset. Again, as I've said before, I'm not claiming it will be a sufficient benefit to outweigh the cost, but I can't agree with your opinion that the 6700HQ is no better than the 4870HQ outside of possibly the iGPU.

Perhaps you'd like to link these websites. Instead of just using innuendo to back an opinion.

Some other websites like Game Debate are little or no better than CPUboss.

 

[rt1707]

Try reading, then. There's a particular difference in multithreaded performance, which, if the OP's CPU is actually at 100% load, would be an asset. Again, as I've said before, I'm not claiming it will be a sufficient benefit to outweigh the cost, but I can't agree with your opinion that the 6700HQ is no better than the 4870HQ outside of possibly the iGPU.

Perhaps you'd like to link these websites. Instead of just using innuendo to back an opinion.

Some other websites like Game Debate are little or no better than CPUboss.

I put three links in a previous post - one a theoretical look, one with practical benchmarks, and one hundreds of actual user benchmarks (which put the CPUs at 113th and 114th - basically even). You did not link to any sites with 6700HQ benchmarks as far as I can see.

Edit: Here they are again. There are plenty more benchmark results out there too, which you can find by googling.

http://cpu.userbenchmark.com/Compare/Intel-Core-i7-6700HQ-vs-Intel-Core-i7-4870HQ/m34954vsm14314
Puts 4870HQ as 113th and 6700HQ as 114th processor based on actual user systems, based on around 28,000 real world systems.

http://www.cpubenchmark.net/compare.php?cmp[]=2314&cmp[]=2586
Puts 4870HQ substantially ahead of 6700HQ based on benchmarking

http://www.hardwareluxx.com/index.p...ted-intel-skylake-core-i7-6700hq.html?start=1
Puts 4870HQ ahead of 6700HQ on most benchmarks again - Cinebench 11 and all three 3Dmark tests.

Pus the CPUboss, which again puts the 4870HQ ahead, but only on a theoretical basis that as you point out exaggerates any advantages.

Basically, the processors give pretty much the same performance in the scheme of things. There is no advantage to "upgrading to the 6700HQ over the 4870HQ other than having a "6" rather than "4" at the start of the number, which perhaps makes some people feel nice.

 

[Oussebon]

My apologies I hadn't see the links there. userbenchmark is in the same kind of category as CPUboss for me. The numbers aren't transparent and lump together all kinds of data including overclocked and thermal/power throttled systems.

Hardwareluxx is at least a legit benchmark, however their result is pretty clearly anomalous/flawed as their 6700HQ is apparently outperformed by a CPU with the same clock speed and a lower IPC - the 4710HQ. So I don't really buy that...

You did not link to any sites with 6700HQ benchmarks as far as I can see.

Now you're just messing with me right?

http://tinyurl.com/h3rh38c

Oh, and if we actually read the article, Hardwareluxx say as much themselves:

Interestingly, however, is that some Haswell CPUs also pass it. Here we see the i7-4710HQ achieve almost 6% more, the i7-4700HQ is around 8%. However, this applies only for the older version of Cinebench. In recent benchmarks the Core i7-6700HQ at least takes third place and beats its predecessor clearly. The situation is similar in single-threaded benchmarks.

Since Notebookcheck shows higher scores for the 6700HQ in Cinebench 11.5 I can only assume that something was rotten in the state of Benchmark at Hardwareluxx.

 

[rt1707]

My apologies I hadn't see the links there. userbenchmark is in the same kind of category as CPUboss for me. The numbers aren't transparent and lump together all kinds of data including overclocked and thermal/power throttled systems.

Hardwareluxx is at least a legit benchmark, however their result is pretty clearly anomalous/flawed as their 6700HQ is apparently outperformed by a CPU with the same clock speed and a lower IPC - the 4710HQ. So I don't really buy that...

So you admit the 4870HQ beats the 6700HQ here

Now you're just messing with me right?



View attachment 9603

This site puts the processors as virtually identical at 49th and 51st place in their rankings. Again, proving my point.

Oh, and if we actually read the article, Hardwareluxx say as much themselves:



Since Notebookcheck shows higher scores for the 6700HQ in Cinebench 11.5 I can only assume that something was rotten in the state of Benchmark at Hardwareluxx.

This is selective quoting. The quote you gives comes in a secton called "The new CPU is not always better" and starts with the line "As suspected, the Skylake CPU can’t compete with its direct predecessor". This clearly proves my point, and the 4870HQ beats the 6700HQ in 4 of the 5 benchmarks in this test. Crushing it in the 3dMark tests actually, while there was only a 7% difference in the single test the 6700HQ won.

You also ignored the cpubenchmarking website, which base their results on benchmarking, which rates the 4870 at 9318 compared to the 6700's 8111. I'm struggling to find other direct comparisons, as the 4870HQ is a relatively rare processor. The 4th generation HQ model one above, which is used in the same model of macbook pro, easily beats the 6700HQ by some distance - http://cpu.userbenchmark.com/Compare/Intel-Core-i7-6700HQ-vs-Intel-Core-i7-4980HQ/m34954vsm21763 . Again - real world benchmarks from hundreds of real systems that that cannot really be argued with.

I am not sure why you are so surprised that the 6700HQ does not outperform the 4870HQ to any significant degree. The 6700HQ is the entry level HQ processor and I believe should really be compared to the 4700HQ if you want a direct comparison between generations. The 4870HQ is significantly more expensive than the entry level model for a reason - and it is hardly an "old" processor having only been released in late 2014. This is why you have still not been able to provide result that show the 6700HQ is better than the 4870HQ.

We are really talking about unhelpful trivialities now though. The conclusion has to be that "upgrading" to the 6700HQ will not help the OP in any way.

 

[Oussebon]

So you admit the 4780HQ beats the 6700HQ here

I also admitted it beat in in the CPUboss scores. What I'm saying is that it doesn't mean a great deal.

Crushing it in the 3dMark tests actually

You know that 3dmark is basically a graphics test and will mostly be influenced by the difference in iGPU right...

You also ignored the cpubenchmarking website, which base their results on benchmarking, which rates the 4870 at 9318 compared to the 6700's 8111. I'm struggling to find other direct comparisons, as the 4870HQ is a relatively rare processor. The 4th generation HQ model one above this absolutely crushes the 6700HQ - http://cpu.userbenchmark.com/Compare...m34954vsm21763 . Again - real world benchmarks from hundreds of real systems that that cannot really be argued with.

I didn't ignore it. I already explained why CPUbenchmark is no good for comparing A vs B. Read.

The 4870HQ is significantly more expensive than the entry level model for a reason

Yeah, the Iris Pro and the fact that it's in a mac

This is why you have still not been able to provide result that show the 6700HQ is better than the 4870K.

*HQ And I did, it's in the notebookcheck cinebench scores.

We are really talking about unhelpful trivialities now though. The conclusion has to be that "upgrading" to the 6700HQ will not help the OP in any way.

I'm afraid in good conscience I can't let you get away with posting junk info that is just incorrect, and then try to obfuscate by saying that it wasn't really the point anyway. In case someone else is weighting up whether to get Option A or Option B and is swayed by bogus info.

This is selective quoting.

Dude, it's context. Unlike you posting a link to the benches hoping people would look at the chart without reading the article.

The conclusion has to be that "upgrading" to the 6700HQ will not help the OP in any way.

I already said that on balance the improvement isn't worth the cost.

I'm done on this matter. You've posted mostly junk data and the one serious article itself explains that its own results are inconsistent. Admit your error instead of digging deeper, or don't admit it but please for goodness sake stop digging and ignoring what's posted. I just hope if some poor individual comes across this thread and makes it this far they won't be fooled by your poor data trying to support an opinion.

What a waste of time you are.

 

[rt1707]

What a waste of time you are.

And after that little gem, I shall bow out of the discussion. I don't have anything constructive left to add, and hopefully I have made my point to the OP.

 

[keynes]

I just hope if some poor individual comes across this thread and makes it this far they won't be fooled by your poor data trying to support an opinion.

That was a long read, I am still unsure whether to upgrade or not!