New MacBook Pro (13" & 15")

[quote=395994:@Christoph De Vocht]I just read the keyboard issues (sticking keys) isn’t resolved. The keyboard is quieter though.
My MBP2017 has already got 3 new keyboards (all replaced for free). The sticky keys issues is not what you expect from a +3200€ machine. :-)[/quote]
top of the line 15" comes to 8000 euros … last one was only 5000 euros … come on Apple ?
they will soon be more than a Lisa …

That’s one of the reason I like Xojo, I’m a Mac user since Apple ][e (1982 I think, bought by my father) but if it was out of thinking some years ago to use Windows, I think about it more and more. Not for Windows itself, but for the computer prices.
And if I do the jump, I will be able to continue my development without any problem.
It will be the same if I jump to Linux I hope (I have absolutely no idea of how Xojo runs under Linux).
What I read here (on Xojo forum) let me think :

  • Xojo’s developers satisfaction under Mac : 7/10 or 8/10.
  • Xojo’s developers satisfaction under Windows : 6/10.
  • Xojo’s developers satisfaction under Linux : ?/10.

I never build for Linux, I build some of my applications to Windows and they run well. I can only say I’m satisfy with Xojo for Mac.

I’ve been an Apple fan for 10+ years, but my next laptop might not necessarily be an Apple. The MacBook Pros are too expensive and the lower end ones either don’t have enough ports (really, apple just a single USB-C port on some models???) or not enough of the right ones.

Apple has been so obsessed with making them so thin they have sacrificed other features.

You can stop now Apple, your laptops are now more than thin enough.

I think that is an unfair statement. The previous top of the line was 16GB memory and 2TB drive. If you take the current top of the line 6 core i9 and price with 16GB memory with 2TB drive it is $4299 USD. I don’t know for sure what that is is Euros, but since the top of the line in USD is $6799, I’ll assume a 8000/6799 conversion which puts the same capacity machine about 5058 Euros.

So roughly the price but with spec bumps like faster CPU, v3 keyboard (time will tell…), True Tone display, BT 5, T2 security chip, etc.

The real price killer is the SSD drives since they are NVMe drives with a 3200Mbs read rate. Going from 2TB to 4TB adds $2000 USD. For those who can get by with a slower external portable drive (even if SSD but not NVMe) then you can drastically reduce the price. In fact, you can get top of line 32GB i9 with 1TB for $3899 USD (4587 Euros?).

I had long thought I’d end up getting a 2TB model (before this refresh) but ended up deciding that I could not justify $800 USD to move from 1TB to 2TB. Instead, I’d take the 1TB drive out of my 2012 (which I kept upgrading over the years) and move to a small external enclosure and buy just a 1TB model. I really don’t need over 1TB when not docked in my office, if I have a drive with some other VM’s etc available.

For those who can really benefit from 4TB of NVMe speeds, they probably welcome the chance to have the option and can justify the price differential. I’m certainly not in that camp myself. But I don’t see this 2018 refresh cycle as a price increase but rather the same prices for a spec bump. They just moved the bar of the top end.

If anything it now gets CHEAPER for me because a 13" can now do quad core and replace my 15". I suspect benchmarks of various models will start showing up very soon, at least for stock models instead of BTO variations.

I’m looking forward to seeing the benchmarks. @Christian Schmitz could you share your Geekbench score from old and new MBP when you receive it ?

  • 1

Expensive is a word you use when your income does not allows you to buy the object whose cost is expensive.

Years ago, a current priced mid-range laptop was not expensive for me. Now my situation have changed, and they are expensive for me. The result will be no more Apple hardware for me until prices drops or income raise.

In the above statement, there is no attack on anybody just facts from my open door on a sunny summer. :wink:

@Jeremie Leroy. I think some others will get theirs quicker and publish it…

While I agree with the above, the latest refresh isn’t any thinner. Nor even thinner display bezels. What is interesting is that they still got about 15% more battery capacity without making it thicker or increasing the weight. (Not that it makes it last longer, just helps make up for DDR4 memory draw. So they increased the battery capacity to keep the hours usage basically the same. )

What I don’t get is why the 13" – which still seems to use LDDR3 memory – still has the same battery life quoted while the capacity rating went from under 50 to 58 watt hours. But regardless, it will be a vast improvement over the battery life in my 2012 MBP. And it got a new battery when my motherboard was replaced for the GPU issue and only has 57 charge cycles on it.

C’mon folks - don’t complain about the Apple prices! Just buy it, My retirement depends on it :D.

We have done some benchmarking between my '13 15" rMBP i7 and the curent 15" i7 and the performance difference is around 3% except for direct memory accesses. I don’t see the value proposition here since my '13 rMBP will still run all of the Mac OS versions including 10.14 and I get to keep my tactile function keys.

Also, it’s obvious that Apple is trying to tease us away from wanting a true desktop solution. An i9-based laptop with 32GB RAM is not a laptop, it’s a desktop that trades expandability for portability. I mean, what more do we need than an Apple Watch?

Well, the 2 new cores on the 15" are probably useful for Xojo and Xcode.
Also VMs will benefit from more RAM and more cores.

… but only for compilation and apps using helpers for using more than 1 core if I understand correctly.

That’s actually one of the things I’d love to quantify. Isn’t it true that Xojo IDE/debugger won’t really make use of more than a single core except (maybe?) when using the LLVM compiler? Multiple cores obviously also help the more tasks or VM’s that are running, but how much does it help Xojo itself?

How much does better multi-core performance help when launching the debugger in 64-bit macOS projects? Builds?

Geekbench 4 on my 2012 rMBP is 3474 single core and 10617 multi core. While the LLVM scores are 3970 and 10195 respectfully.

The leaked 13" i7 is reported to be around 4448/16607 for the general Geekbench 4 score, with its LLVM scores at 7372/30239. The leaked 15" i7 running at 2.2GHz was reported as 4902/22316 but its LLVM scores were a whopping 8399/52673. And it was an i7 not i9, so probably a prototype model (this one was June 1, ahead of WWDC).

I have not found anything on the 2.9GHz version yet.

Many more scores should be published in the next few days/weeks, though I’d love to know if the LLVM scores have much correlation to the speed of launching debug (or build) of 64-bit macOs projects.

LLVM uses multiple cores.
Now I still wait for Xojo to use them for (pre)compiling.

Your Xojo app can get a lot of cores busy by using MT methods in MBS Xojo Plugins.

e.g. SHA512MBS.HashFile is optimized for threads.

[quote=396197:@Christian Schmitz]LLVM uses multiple cores.
Now I still wait for Xojo to use them for (pre)compiling.[/quote]

Are you saying that the LLVM compiler will use a single core when (pre)compiling Xojo 64-bit projects?

I’m not asking about how Xojo built apps can use multiple cores. Just if IDE debug cycles (and to a lesser extent builds) benefit from more cores and higher LLVM benchmarks.

While I expect lots of Geekbench scores to short appearing in the short-term, I’m still not sure how much the multicore speed (and presumably, specifically the LLVM score) will effect (pre)compile times in Xojo. It would be enlightening to know how much precompile times improved for you from your current i7 to your new i9 for a large project and if the ratio seems at all related to the Geekbench scores for each machine.

LLVM can use multiple cores.
Xojo may decide to use them or not.
Currently they only use multiple cores for linking as far as I know.
I’d love to see them doing the precompile as well as normal compile in preemptive threads.
e.g. pass the data just to the C plugin to compile with other core.

Bummer. But at least the Geekbench single core LLVM scores seem to be significantly higher than the general score by a much bigger margin then they are for my 2012 chip.

First benchmarks are insane: https://www.macrumors.com/2018/07/13/2018-macbook-pro-fastest-laptop-ssd-ever/?utm_source=feedly&utm_medium=webfeeds

And it seems the keyboard is “fixed”: https://www.macrumors.com/2018/07/13/ifixit-butterfly-keyboard-silicone-barrier/?utm_source=feedly&utm_medium=webfeeds

Another way to look at the benchmarks is a Geekbench limit to MacbookPro15 where I think you can compare the various models by looking at the chip used. I think they work out like this:

13" i5: 8259U
13" i7: 8559U
15" i7: 8750H
15" i9: 8950HK

Some of those are well over 50% higher single core than my MBP, and over 200% more multi core. And if you drill down to individual tests, the LLVM scores are even more impressive.