The Ethics of Buying Both to Test

[iRun26.2]

My long-term approach was to get rid of Windows programs that I had to run and I eventually got it down to two and then worked really hard to find macOS equivalents for those two programs.

That is definitely not an option. I have Windows software development tools that have been obsolete for 15 years that I still must use. There will never be anything that replaces it. I could name other examples. Despite disliking Windows I have no choice but too use it at my job.

 

[iRun26.2]

I once installed OSX in a VM and compared performance in the VM with base OS performance. The VM was much slower. You could possibly try to run a synthetic benchmark in the VM to get some data.

There is not so many thing you can tune. Start by looking at performance in the VM using windows tools, Win 10 runs a lot of stuff in the background. Especially if the VM has been suspended, Windows start by phoning home, checking for updates and running anti virus. Even on my 16” the VM is almost unusable for 15 minutes, with up to 100% CPU usage from the ”background” processes. I start up the VM and then take a coffe or two. Shut of all non-essential stuff in the VM, give the machine more virtual cores and all the memory it needs. Personally I really hate this behaviour of Win 10. It feels much slower than Win 7 for same tasks on the same HW.

Then check on the Mac side that you are not starving the OS, I make sure that OSX has at leats half of memory and cores available.

I have said earlier in this thread that your application looks more single threaded and that I think you would benefit more from the 16” sustained single thread performance. You could try to change the number of virtual cores in your VM to test this hypothesis.

The thing I don’t understand is how the reference computer (the 2019 Dell) can have its fans running (working hard) while my MacBook Pro is doing little (with the fans not turning on, or at least not loud). It’s like it doesn’t think it needs to go fast.

 

[pshufd]

The thing I don’t understand is how the reference computer (the 2019 Dell) can have its fans running (working hard) while my MacBook Pro is doing little (with the fans not turning on, or at least not loud). It’s like it doesn’t think it needs to go fast.

I built a macOS Mojave VM on VirtualBox today and I'll just say that it's usable compared to the host system. I wasn't expecting much and that exactly what I got. It's usable for just typing in text but I wouldn't try web browsing or anything with graphics in it. Parallels is a lot better than VirtualBox because they use the system graphics efficiently.

I've used Parallels with a Windows VM and it's not bad but it's not Boot Camp. One of the other approaches that I've considered is to leave a mac running in the house to VNC into it and that's still on the table. If you really need Windows, I'd say buy a PC or run Boot Camp. Parallels is a solution for a lot of people but certainly not everyone.

I'd never built a macOS VM before. Done it a gazillion times for Windows and Linux though.

I look forward to the day when we have macOS Cloud Servers. I understand that there's a company that does that today. The biggest problem with such systems is bandwidth to get your files in and out. If we can solve that, then we won't really need super-powerful laptops and desktops. And the sharing aspect should mean overall lower costs. At the moment, that's what I'm thinking of doing at home.

I keep thinking of buying a Mac Pro 5,1 with 12 cores at 3 Ghz and 48 GB of RAM for $750 for sale locally. And then I think that it's already vintage, the boot screen doesn't show up because newer video cards don't have Apple's secret sauce, and Apple could kill off new operating systems at any time. I could buy a Mac Pro but I don't need that amount of power - I just want the expand-ability. Figuring out the Apple hardware work can certainly be a challenge.

If I were you, I'd just have multiple machines. Right now I have three MacBook Pros and two Windows Desktops, and, between the five systems, I have plenty of compute power.

 

[pshufd]

The thing I don’t understand is how the reference computer (the 2019 Dell) can have its fans running (working hard) while my MacBook Pro is doing little (with the fans not turning on, or at least not loud). It’s like it doesn’t think it needs to go fast.

There may be a knob on the VM limiting resource usage. I had to set those today building the VM. It allows you to control how much RAM, storage, number of CPUs and the percentage of total CPU.

 

[KPOM]

That is definitely not an option. I have Windows software development tools that have been obsolete for 15 years that I still must use. There will never be anything that replaces it. I could name other examples. Despite disliking Windows I have no choice but too use it at my job.

Genuine question (i.e. I’m not trying to be flippant), but if the tools have been “obsolete” for 15 years, then wouldn’t they run on a relatively cheap Windows PC (e.g. a Surface Go) that you could use? Other than the convenience of using a single device, is there a particular reason why using 2 devices is undesirable?

 

[matram]

The thing I don’t understand is how the reference computer (the 2019 Dell) can have its fans running (working hard) while my MacBook Pro is doing little (with the fans not turning on, or at least not loud). It’s like it doesn’t think it needs to go fast.

The MBP will not slow by itself. Something is limiting it: CPU, memory or other processes. Use the tools in Win10 / OSX to identify what and play around with VM settings.

My guess would be virtual cores. Your MBP had four hyperthreaded cores, which would be eight possible virtual cores. A default VM will get one virtual core, i.e. 12% max CPU load on you MBP, that is not a lot . If that is the case give the VM more cores and see if it makes a change.

 

[iRun26.2]

Genuine question (i.e. I’m not trying to be flippant), but if the tools have been “obsolete” for 15 years, then wouldn’t they run on a relatively cheap Windows PC (e.g. a Surface Go) that you could use? Other than the convenience of using a single device, is there a particular reason why using 2 devices is undesirable?

Because I also have other software (like the FPGA I described earlier) that is very challenging for the PC).

Having two devices means that I have to carry around two computers. I’d rather, then, have the 16” beast (price is not an obstacle for me).

I must say, too, that I really love using my new computer. It is a joy to type on and I love the screen (even though it isn’t the 14” mini-led I was hoping for). I also love the touch-bar. I haven’t gotten used to all of the things it can do yet but it is kind of fun to use.

 

[iRun26.2]

There may be a knob on the VM limiting resource usage. I had to set those today building the VM. It allows you to control how much RAM, storage, number of CPUs and the percentage of total CPU.

It would nice if it was as that easy. It seems, intuitively, that it should be but I’m certainly not seeing anything that suggests that it is.
[automerge]1592126756[/automerge]
I’m not 100% certain that it is possible but I would like the Boot Camp Windows option to allow me to access the same system that the Parallels Windows uses.

Is that possible? It sure seems like it might be.

 

[matram]

I’m not 100% certain that it is possible but I would like the Boot Camp Windows option to allow me to access the same system that the Parallels Windows uses.

Not sure what you mean, bootcamp assistant creates a new partition on your disk, installs windows to it and then you select which disk to boot from. The VM is a virtual disk as a set of files in your home folder. You can create a new VM from a bootcamp partition but these are two different systems. To my knowledge you can not go the other way.

I believe you need to be clear on your priorties and as others also have said play around with settings for the VM, there is no magic solution that can be had through this forum, you need to do the work!

Today 2% use bootcamp, most people are OK with a VM but you will need to tweak it to get good performance.

 

[iRun26.2]

Not sure what you mean, bootcamp assistant creates a new partition on your disk, installs windows to it and then you select which disk to boot from. The VM is a virtual disk as a set of files in your home folder. You can create a new VM from a bootcamp partition but these are two different systems. To my knowledge you can not go the other way.

I believe you need to be clear on your priorties and as others also have said play around with settings for the VM, there is no magic solution that can be had through this forum, you need to do the work!

Today 2% use bootcamp, most people are OK with a VM but you will need to tweak it to get good performance.

What I am referring to is described under the "How to set Parallels Desktop to run Windows from the Boot Camp partition" in the following article:

Run Windows on Mac with a virtual machine | Parallels Desktop

Download Parallels Desktop virtual machine to run Windows on Mac without rebooting or slowing down your Mac, plus get over 200,000 Windows apps.

www.parallels.com

If this does what I think it does, it will be the best of both worlds. I think this will allow me to use Boot Camp alone (without Parallels) when I need the optimal speed/control but still let me use the same partition in a virtual machine when I want the convenience of Parallels.

I have tried / experimented with every kind of setting I can see under Parallels. Nothing seems to make it go as fast as the Dell computer. Under "Optimization" I've set the Resource Usage to "No limit". In Parallel 15 this is the only setting now available under "Optimization". I have tried playing around with the various operating scenarios offered by Parallels with no improvement. I think that it is not possible to get the Virtual Machine to match what the Dell on it own can do.

I'm now going to try and figure out how to set up Boot Camp. I think it would be good to get a baseline to know how the machine perform without any possible hindrance of Parallels.

Edit: I guess you were talking about the same thing. I do agree I will need to reinstall Windows to create a Boot Camp partition. I didn't know before that I could run the Partition from Parallels or boot too it.

 

[matram]

Thank you for sharing. I have no experience with running a Bootcamp partition as a VM but it seems you would just be accessing the files from the bootcamp partition but performance wise it would be similar to a normal VM. I do believe botcamp will provide you better performance if you boot into it, the question is how much.

I struggled a lot with getting good performance from Visual Studio Enterprise in Win 10. In the end it took a lot of tuning of Win 10, Visual Studio, the VM plus a 16” MBP to reach my target.

If you provide a bit more detail about what you are trying with some objective results you may get more and better suggestions.

 

[majkom]

I have very good reasons to want both a 2019 16” MacBook Pro and a 2020 13.3” MacBook Pro to compare:

1) I want to see the impressive performance of the 16”. A certain simulation (no GPU required) I need for my job currently takes 30 minutes to complete (on a 4-core Dell Inspiron 3493). I also like the bigger display of the 16” but hate the bulky package.

2) I want to experience the portability of the 13.3” that would allow me to use my computer more comfortably in many different environments (especially when not working on the project described above). This would make the computer far more useful when away from my 32” 4K monitor setup.

I’m very tempted to order both (from Apple), try them out with my specific software/setup, and send back the one I decide against before the 15-day return window closes.

While there is nothing stopping me from doing this, I feel it is bordering on being unethical. It obviously is not the intended method of purchase as one of the computers would be ‘open box’, costing Apple extra money.
.

Has anyone ever done a ‘trial run’ like this?

nothing unethical, actually, apple got this policy calculated into prices (and boasts with customer friendly return policies), so why not... if that helps you, do that.. after all, apple will win when you buy apple device that would suit you best - happy customer spreads the word - more sales...

 

[meady100]

You won't be the first person to do this, and you can be guaranteed Apple has built this kind of purchasing into their bottom line. Hell, Asos clothing actually has a returns tickbox on their paperwork listing 'I bought more than one size'.
These returns policies make more money than they cost.

 

[naturalstar]

nvm

 

[pshufd]

I think that Parallels allows you to run a Bootcamp partition in a VM. But you have to install Bootcamp and Windows. I don't think that it works in the other direction. I had Parallels for a few years and only ran it as a VM and would have had to purchase another Windows license to test it. I don't know if Microsoft charges for Windows 10 these days. I'm pretty sure that you could try installing it and they give you 30 some time before you need to activate it if they still charge for it.

 

[iRun26.2]

Well, it sounds like your usage has some specific challenges. Not sure those were mentioned earlier, or at least not in this detail; perhaps others here smarter than I regarding the demands of running Parallels/Windows 10 alongside Mac OS would have anticipated this disadvantage for the 13" model and advised you more towards the more powerful 16". If in fact the 16" with 8-cores will in fact be more suitable?

At this point I can fully understand the desire to evaluate the 16", keeping in mind that the clock is running on how much time you can take to try the 13" before you'll either have to return it, or live with it. With the further complication that the announcements at WWDC could cause you further angst!

Your OP didn't say that your current computer was broken or that you required a replacement on any particular timeframe. It is a 2019 Dell? If your work must be done on software that can't be operated on the Mac OS and requires Windows, then why not get a more powerful Windows laptop? If most of your laptop time is spent running that program then that is probably the most cost effective solution.

I know, we are all Mac fans here, but is trying to make a MacBook Pro perform (via Parallels) as well as a Windows laptop with a more powerful processor than what your Dell currently has, really the best solution?

At this point, if I was in your shoes, I might return the 13" and then keep my eye on what is announced and what is foreseen within the next 6 months or so. You have concerns about the current 16" Pro, so why not wait? The rumor mill says that mini-LED MacBook Pros at both 14-inch and 16-inch are coming in late 2020 or early 2021.

1) I am willing to stick with the 13.3” unit if it can match the speed of my Dell (10th generation i5) computer running my FPGA development tools in Boot Camp. I’d definitely prefer to run in Parallels but that hasn’t been impressive. The bigger Dell might have an advantage over my smaller computer because of better thermals (just like the 16” MBP). That would be disappointing to lose out to such a cheap Windows laptop.

I would like to know how much faster the 16” would be. I could order it thinking I wouldn’t get it but like it enough to keep it if it blows the Dell away. It’s just too big for my tastes.

2) I have heard that I can sort of extend the time I have to evaluate the 13” MBP by saying I want to return it at day 14 and then I have 14 more days to send it back. I really do want to keep this computer. I’m willing to sacrifice some performance. It’s compact compared to the Dell but large compared to my 12” MacBook.

I am curious what is revealed at WWDC. I’m curious if the new ARM MacBook will get the 14” screen. I can’t imagine switching over to ARM and I can’t imagine they would release something that would outdo the 13” MBP that released just last month.

3) My old computer is not broken. It’s 512GB SSD is getting full. Additionally I realized how much time I was wasting waiting for my FPGA development tool software to compile and simulate. This was made acutely aware to me when I borrowed a Dell laptop for comparison. My dual- core 12” MacBook didn’t have a chance against a 4-core computer.

I guess I could get a more powerful windows laptop. I hate the thought of that and having to carry around my other computer just so I can write Xcode. As I also said. I’m really liking this 13” MBP except for its poor performance in Parallels compared to the Dell.

4) I am fully expecting that I will buy the new (Intel) 14” MBP when it comes out (assuming my current one will suffice for now). Apple has me here. They could release a new one 6 months after the current one and I’d bite. I am a sucker for better displays. I love the 16” MBP screen but I will wait for better thermal regulation and the mini LED screen before I invest in the 16”. I want a small MBP with a 14” mini LED screen.

 

[burgman]

This is Apple, without any doubt that cost is put back to the customer one way or the other. It's like when people say OS X is free, it's absolutely not. The cost of the OS is baked into the price of every Mac Apple sells. In countries that have a sale and return policy have at it, equally don't think for one second there's not a cost index to it, Apple would charge to enter the store if it could...

Like all in life what you do and how you conduct yourself has repercussions. In this case Apple can act magnanimously, yet in reality applies the cost directly back to it's customers.

Q-6

Given the accounting tricks available to multinational companies and what Apple sells returned machines for on the refurb store, they make the same or more as the original sale. Also returns are used in the warranty exchange program.

 

[iRun26.2]

Given the accounting tricks available to multinational companies and what Apple sells returned machines for on the refurb store, they make the same or more as the original sale. Also returns are used in the warranty exchange program.

How could they possibly make more off of a refurb when they are discounted compared to the brand new machine?

 

[pers0n]

Apple is a billion dollar company that evades taxes in some counties, bows to Chinas will over free speech, is opposed to the right to repair, doesn’t believe in reuse or reduce, and hires third world workers for peanuts. Don’t worry about morality of buying both.

Just buy and return they have more money than any human could need. There is no morals to argue about.

 

[iRun26.2]

Apple is a billion dollar company that evades taxes in some counties, bows to Chinas will over free speech, is opposed to the right to repair, doesn’t believe in reuse or reduce, and hires third world workers for peanuts. Don’t worry about morality of buying both.

Just buy and return they have more money than any human could need. There is no morals to argue about.

Why is it that so many people think that doing something unethical to an unethical company is ok? I don’t even agree that Apple is as bad as you make them sound to be. But certainly two wrongs don’t make a right!

 

[iRun26.2]

Good test results ...

Del Inspiron 3493 (10th gen, i5) vs Apple MacBook Pro 13" (10th gen, i7):

Dell: Height = 0.8", Width =13.35", Depth = 9.52", Weight = 3.95 lbs
Apple: Height = 0.61", Width = 11.97", Depth = 8.36", Weight = 3.1 lbs

I set up Boot Camp on my new MacBook Pro and ran some tests using the Windows application software I plan to use (which is the primary challenging software I expect to be using the computer for):
1) Libero FPGA Compiler /Synthesizer
2) ModelSim FPGA Simulator

I set up Boot Camp such that it could either be run by itself or as a Parallels virtual machine. I ran the software and compared the times to complete tasks with those of my Dell reference computer:

Libero FPGA Synthesizer: (shorter times are better)
a) 2015 12" MacBook (previous computer) - 8 min, 34 sec
b) Dell - 2 min, 56 sec
c) New MBP (Parallels) - 3min, 20 sec
d) New MBP (Boot Camp) - 2 min, 50 sec

ModelSim FPGA Simulator: (shorter times are better)
a) Old 12" MacBook (previous computer) - Not tested
b) Dell - 40 min, 39 sec
c) New MBP (Parallels) - 44 min, 57 sec
d) New MBP (Boot Camp) - 37 min 3 sec
e) New MBP (Boot Camp while connected to 4K monitor) - 37 min 1 sec

I attached pictures of the Windows core activity (on the Dell) showing the typical load presented to the processors:

Operating Frequency: The Dell (with it's i5) was able to sustain a a 3.4GHz operating frequency throughout my testing. My new MacBook Pro (i7) was able to sustain an operating frequency close to 3.9GHz throughout the tests.

Thermal Throttling/Fans: I did not use an app to measure the temperature or fan speed but because both computer were able to maintain close to their turbo frequency throughout the test it is clear that the thermal throttling was minimal. This makes sense because the two attached plots show that the load presented to the processors is not optimal for multi-core processing (most of the time not all cores were active or required). Throughout the testing the fans speeds of the computers were definitely noticeable (the Dell was clearly louder).

It was nice to see that when I hooked up my MacBook Pro to my 4K display that the performance of the computer did not degrade (the time to complete the task presented did not take any longer when both screens were on, and the fan noise was not noticeably louder).

Results: I was disappointed in my earlier tests which showed that the Dell computer was faster at running my FPGA applications when I was comparing it to my MacBook Pro using Parallels. It is now nice to see that my new MacBook is able edge out the Dell when running in Boot Camp mode. Note that the Parallels was running the Boot Camp partition for the Parallels tests. I assume that my i7 processor was what helped the improvement over the Dell's i5.

Comments: I am happy now because my MacBook Pro computer is able to match (and do slightly better) running the processes where I need the speed for may job. While it may seem silly to compare my expensive MacBook Pro to the sub-$1K Dell computer it still shows me that my new MacBook Pro is capable of replacing the Window's PC while still leaving the ability to run other OS X applications (like Xcode). I will not most likely keep my MacBook Pro. It allows me to have a portable computer that is capable of reasonable performance in both Windows and OS X on the same machine (I don't need to carry two computers around).

I still am tempted to order a 16" MacBook Pro to see how it would do. I would like to know the results but I think that whatever the results would be I would still keep my 13.3" version because of the improved portability of the system. It also should be noted that ordering the 16" MackBook Pro to test would be a lot of work (setting up Boot Camp and down loading my FPGA software). Additionally, because my software has licensing issues I would need to go through trouble of obtaining a temporary license to test it out (which I why I didn't perform both test on the my old 12" MacBook).

One thing to note is that the machines were both able to maintain a relatively fast operating speed (close to their turbo speed) because the load presented to them by my test software didn't overwhelm their cooling system. This makes me think that, though the 16" MacBook Pro would perform better (due to higher turbo speeds), its more effective thermal cooling system would probably not help it much for my applications. Additionally from the plots of the core activity it appears that the software that I testing is not optimal for multi-core processing (more cores don't necessarily help my software requirements...although it would probably remove the requirement of Boot Camp for a faster time). This would also tend to reduce the expect improvements to be gained by the 16" version.

 

[matram]

Thank you for sharing so much data it makes it easier to provide suggestion.

You confirm again that your FPGA simulator can only use two threads, which is 25% of the 13” and 12.5% of the 16” CPU capacity. As you say your app will not cause thermal throttling due to low CPU load.

You can get a 16” that will boost up to 5.0 GHz, this would be 28% more than the 13”, if you are able to sustain that frequency. I have said before in this thread that I believe the 16” provides the best performance for you. You should test the 16” if performance of this app is the most critical aspect for you.

Of course you may want to investigate other SW solutions that paralellize better.

 

[iRun26.2]

I've actually just realized something very critical. When using the 'parallelized' Boot Camp virtual machine, I can't copy and paste files line I can normally do in a normal Parallel virtual machine (even though I toggle between Boot Camp and OS X very easily). I do not like this at all! I don't think that the increased performance I see in Boot Camp is worth losing the luxury of being able to work nicely between the two operating systems.

I don't know what to do. I think I may just destroy the Boot Camp partition and learn to live with the slightly reduced performance of a Parallel virtual machine. Argh!
[automerge]1592319160[/automerge]

Thank you for sharing so much data it makes it easier to provide suggestion.

You confirm again that your FPGA simulator can only use two threads, which is 25% of the 13” and 12.5% of the 16” CPU capacity. As you say your app will not cause thermal throttling due to low CPU load.

You can get a 16” that will boost up to 5.0 GHz, this would be 28% more than the 13”, if you are able to sustain that frequency. I have said before in this thread that I believe the 16” provides the best performance for you. You should test the 16” if performance of this app is the most critical aspect for you.

Of course you may want to investigate other SW solutions that paralellize better.

The 16" would definitely provide the best performance for the apps that I currently use the most. I might order it to test it but in the back of my mind I still think I'd prefer the loss of performance of 13.3" because I gain a more portable computer. What I have now in Parallels is still much much faster than what I had in my dual-core machine.

I don't know if I'd feel right ordering the 16" machine to prove a point when I really don't think I want it. I bet the fans would come on when I connect up 4K monitor.

 

[iRun26.2]

While I do love my 13.3” 2020
MacBook Pro... I can’t get over the nagging question of how much faster would the 16” 8-core i9 version be. I sit here waiting for my FPGA simulator to complete on my 13.3” wondering if the 16” would be done by now.

I think I’m going to order a refurb (so I don’t ‘waste’ a brand new unit) 16” later today. I’ll pay the extra $8 to have it here by Tuesday so I can test both of them, side by side right away.

I’ve tried to decide with just the one but I just can’t. I’m sorry if that sounds unethical to you, I feel I need to do this to be sure.

 

[iRun26.2]

I just ordered a 16” MacBook Pro:

i9, 8-core, 32G RAM, 4TB SSD

I’ll see how much faster it is and how much bigger it is (is various use situations). Then it will come down to what quality overrides the other.