Apple MacBook Pro 13-inch (2018) Review: Actually for Pros

bananaforscale

Commendable
Mar 14, 2016
46
0
1,540
2
It doesn't throttle because it doesn't have an i9, and even with the current CPU it's close. "The CPU ran at an average temperature of 87 degrees Celsius (188.6 degrees Fahrenheit)."
 

TJ Hooker

Champion
Ambassador
It'd be nice to know the hardware configurations of the other laptops tested. I only checked out the XPS 13, but it has a variety of options including CPU, RAM size/speed, SATA vs PCIe storage, etc.
 

Kelavarus

Distinguished
Sep 7, 2009
510
0
18,980
0


This is exactly what I was going to say. Makes it kind of confusing. I was going to say a lot more in specifics, but decided to do some digging, and holy cow. I was confused because the Surface Book 2 gets such a low score on Geekbench 4 compared to the MBP, and the SB2 is equipped with (judging by the score) an i7-8650u, while the MBP has an i7-8559u. I would have assumed the higher number i7 should be better, right? Especially as they both start with 8, so the same generation right? Buuuut apparently Intel makes no sense when it comes to naming conventions and the 8559u is Coffee Lake and has a higher TDP than the Kaby Lake 8650u. Ooookay, Intel. Not confusing at all.

I am still curious though about the very large gap. Seems unlikely it's all CPU, and the SB2 should be equipped with at least a GTX 1050, so is that T2 chip working some magical wizardy on the benchmark?
 

TJ Hooker

Champion
Ambassador
@Kelavarus yeah as a consumer there's little you can glean about an Intel CPU just based on its model number (other than generation), at least as far as I can tell. As you noticed, the 8559u has a higher TDP, and higher base/boost clocks compared to the 8650u. Also, the 8559u is coffee lake, whereas the 8650u is Kaby Lake refresh. Confused yet? :p

I'm guessing the geekbench scores are purely the CPU results (from what I can find it looks like Geekbench 4 has separate compute workload test to test GPU performance).
 
Aug 1, 2018
11
0
10
0
Something smells fishy. When I see scores like the one in your benchmarks, I cannot help but wonder how it was achieved. And then I look for explanations.

I love Tom's because most articles are somewhat analytical. Some articles, by certain contributors, are very analytical. They are interesting and usually answer most of my questions. This article, instead, avoids even asking the most obvious question: How was the storage performance achieved?

As I see it, there are three possible reasons for the very high score -

1. We have ourselves a huge breakthrough in storage technology that the rest of the industry has somehow missed (highly unlikely) or,

2. Apple is utilizing Intel Optane and hasn't told us (unlikely) or,

3. The true performance of the SSD is being hidden by Copy-on-Write (likely).


I would like to know which one it is. Most of all, I would like a review on Tom's to NOT just buy into the hype and instead investigate the issue.
 

TJ Hooker

Champion
Ambassador
@owhansson my best guess is that the storage benchmark used is basically reporting max sequential read speeds, in which case the results are about on par with other premium PCIe SSDs. It appears all the other laptops tested were configured with SATA SSDs, making the comparison a poor one.

It's also possible your option # 3 is coming into play, in the form of Apple's new "Instant Clone" file system feature.
 
Aug 1, 2018
1
0
10
0
I don't understand why you would recommend Huwaei products when the FBI, NSA, and CIA and other security agencies have all urged people not to buy their products. They're considered a national security risk.
 

TEAMSWITCHER

Distinguished
Aug 7, 2008
203
4
18,685
0


The answer to your question is #1. With High Sierra, Apple introduced a new file system that is dramatically faster at certain operations (i.e. copying files) than the previous HFS+ file system, or even Microsoft NTFS. Files are not actually copied, new file descriptors are added with back pointers to the original file and modifications are stored as deltas. That is not to say that Apple SSD controllers are not fast .. they are. But in this test, it's AppleFS that is doing most of the "magic" here.
 

bortao

Distinguished
Jul 10, 2015
4
1
18,515
0
Why there's no specs of the comparing products?
Why there's no comparison to last year's mbp? or to a 15" model?
File transfer benchmark is invalid, because what AppleFS do is a simple clone, not sector by sector copy, so this do not measure the hard drive speed.
This processor's turbo is 4.5Ghz. In your "real-world video editing test" you just mention the base 3.0Ghz. So I assume it is indeed throttled down (to the base clock), and couldn't achieve turbo speed. So your headling "Throttling? Not a Problem" is incorrect.
 

lew_zealand

Commendable
Aug 1, 2018
3
0
1,510
0
4.5 GHz is a single core Turbo number and video editing is a multi-core workload so you'll need to see what the all core Turbo mode is.

Additionally, all laptops throttle below maximum turbo when all cores are used, thin and light laptops more so. The Macbooks are no exceptions and neither are Dell, Lenovo, etc. The differences are in the cooling hardware for each. I'd love to see a comparo of the same CPUs across different manufacturers to see who is doing the best job of cooling in a thin and light laptop.
 

TJ Hooker

Champion
Ambassador
@lew_zealand all core turbo is 4.1 GHz. And there are some laptops that can maintain max turbo indefinitely. But they're usually much bigger, you're right that it would be very exceptional for an Ultrabook class laptop to be able to maintain max turbo indefinitely.
 

studmoose

Prominent
Feb 27, 2017
75
0
660
15
So, essentially, the performance test is pretty bogus as it doesn't test the actual SSD performance.

What we are seeing here is a test of Apple's "copy" function, which really just creates an alias pointer to the same file, which only applies when you are copying files within the same drive, the actual rewrite of the new file will occur when it is altered.

The proper SSD test would be to perform writes and reads from separate media, which would test the entire I/O subsystem. Copying the same files back and forth on the same drive is kind of a scam test.

Now, one issue with this is in the event of data reclamation. Having files copied to separate locations means that if an area on the SSD fries or becomes unusable, the ability to access an actual file copy allows recovery more likely. Here, if the actual file area gets corrupted, the base pointer and its new file alias will both point to the damaged file location--hence, loss of data. Also, by making actual copies, the index area is spread out, possibly to multiple tracks, allowing for easier access to a copied file. Now, some will say, that this is where backups come into play, but the exposure exists to lose files more readily when one is not physically copying files to multiple locations within the drive.
 
Aug 1, 2018
11
0
10
0


It is not. The APFS isn't that efficient and it's hardly news. A software implementation cannot fix a fundamentally hardware problem (NAND is slow, controllers are slow). It can, however, mask the problem. The APFS factor is in my #3, Copy-on-Write, which is a decades old technique. That could mask the actual SSD performance, if the benchmark used doesn't bypass it.

On APFS: APFS performs differently than HFS+, but certainly not better in all situations.
 
Aug 1, 2018
11
0
10
0



These are the lines I've been thinking along as well.

Though I think a proper test of the SSD would be to revert to HFS+ and see how it compares to previous generations on the same benchmarks and file-system.

Thanks for the info on file-corruption! I hadn't thought about that.
 

studmoose

Prominent
Feb 27, 2017
75
0
660
15
Let's put this in perspective here:

This is a test of a $3,700 MacBook with just an i7 and a 13 inch monitor.

For her college STEM program, my daughter purchased a 6th Generation Lenovo X1, on sale, It has an i7-8650U, 16MB 2133Mhz memory, a 1TB Opal SSD, a 14-inch 500 nit 2650x1440 display, an Ultra Sleeve, a 27-inch Samsung monitor, a docking station, a mouse, a keyboard, an Ethernet dongle, an HP ENVY 7855, a year of printer ink, paper, a 5-year Next-Day On-Site Repair with accidental damage warranty, 3 years of LoJack for Laptops, CAT6 & HDMI cables for under $3,000--a savings of $700.
 
Aug 3, 2018
1
0
10
0
This computer is truly beautiful! Really happy to see this review from you guys! I'm a huge Apple lover and I feel like you guys dog on Apple products (specifically phones- hehe) sometimes, so I feel like this is really honest! Great write up and thorough review!
 

studmoose

Prominent
Feb 27, 2017
75
0
660
15


Hi Jaimelynn, We are a mixed house with an iMac, MacBook Pro, iPhones, etc. as well as MicroShaft and Linux There comes a point when a company charges $700 to fix their poorly engineered keyboards that really warrant a recall, as the key caps come off, and a company that charges twice the price of other high-end laptops, just because of some cult-like branding--that objectivity is needed. Sure, I like some of what Apple has to offer, and I am also critical when they are passing off junk or insignificant change as some must-have equipment. Perhaps, I'm a little more jaded and attuned to it after buying into that Nintendo GameBoy switcheroo I went through with my four kids. Every new device had to be purchased, for what?

What the new MacBook offers is nice, but they are fleecing their customers. Oh, when I mentioned all of the stuff my daughter bought, I forgot to mention that she also purchased an external DVD player too. And that was after tax, and it was still under $3,000.

So, I just bought her a Creative Labs speaker system, a vacuum cleaner, more supplies, a 30"x5' height-adjustable standing desk, a leather manager's chair and a rear-view mirror backup camera system for her Subaru Outback. There's still a little money to spend before we hit that $3,700 limit.

These are what detailed reviews looks like, including using thermal imaging to detect heat issues. Although the MacBook Pro 13 is an i5 system, it sheds a lot of light on this subject. There is also a MacBook Pro 15-inch review.

https://www.notebookcheck.net/Apple-MacBook-Pro-13-2018-Touch-Bar-i5-Laptop-Review.316648.0.html

https://www.notebookcheck.net/Apple-MacBook-Pro-15-2018-2-6-GHz-560X-Laptop-Review.317358.0.html

 

alysdexia

Distinguished
BANNED
Jul 4, 2009
6
0
18,510
0
fast -> swift; low travel -> small stroke; thicker -> broader, deeper; below -> before; above -> behind; nice < niais < nescius := not-skilled -> well, dumbass; APFS file system -> APFS; top -> hind; thin -> squat, fine, shallow; keyboard and -> keyboard, and; Upgradeability -> Upgradability; Software and -> Software, and; will -> shall; Numbers and -> Numbers, and

You forgot the MateBook and Surface Book only hav Bluetooth 4.1. Your review video is stupid; what does it matter if computers get 60 to 120 FPS if you can't see that many? Maybe if you had a handset on a ATV and there were a fan between you and the screen do the frames chop and tear.
 

alysdexia

Distinguished
BANNED
Jul 4, 2009
6
0
18,510
0

charges !-> their; 1 != 2.
Only the storage and memory upgrades are overpriced. The defect rate between brands of this tier's models either hasn't been documented or supports the opposite of your FUD. See my refutation of Louis Rossmann's "The horrible truth about Apple's repeated engineering failures": https://pastebin.com/F4JDZWvJ.


has !-> they; 1 != 2.
Are you exactly the target market for such gear so that you're qualified to criticize it?

I'm not sure what you mean with the switcheroo. The original Game Boy's screen was shit. The Game Boy Advance's too. Game Boy Pocket gave decent contrast and reflection and Advance SP 101 gave decent emission. GBA could play GB and GBC games so you could sell the GB and GBC. The most insignificant changes were the limited edition housing colors and prints.


It's not nice < niais < nescius := not-skilled; you are.


Check the upper review's September update.

https://www.crn.com/slide-shows/mobility/300106841/head-to-head-new-apple-macbook-pro-vs-lenovo-x1-carbon.htm

The specs between your and this laptop don't match. Start with the Apple's base configuration and make sure the other brand has NVMe (not SATA), Thunderbolt (not USB 3), Bluetooth 5 (not 4.1), retina display, creative and networking tools (not in Windows Home). Put your gear on Wi-Fi or Bluetooth. Or did you not realize that too many ports defeats a portable's purpose?
 

ASK THE COMMUNITY

TRENDING THREADS