Late 2013 MacBook Pro review: CPU performance

▼ No more bottlenecks: the late 2013 MacBook Pro reviewed

Over the next few days, I'll publish my review of the new 13" MacBook Pro here in several installments. This is the second one, looking at CPU performance.

Late 2013 MacBook Pro review: CPU performance

Every time I bought a new computer, it was faster than the previous one. Sometimes by a lot, sometimes by a little. What I'll be doing here is compare the new MacBook Pro to my old mid-2011 MacBook Air. I paid extra for the upgrade to a 1.8 GHz Core i7-2677M CPU. This CPU has 4 MB L3 cache, and although it's advertised as a 1.8 GHz part, it actually runs at 2.6 GHz with both cores active, and turbo boosts to 2.9 GHz when only one core is running.

The new MacBook Pro has a 2.4 GHz Core i5-4258U CPU with 3 MB L3 cache. I haven't been able to find any information on which turbo ratios apply to the 2.4 GHz frequency with two cores in use, but it looks like this is also 2.6 GHz. The maximum CPU frequency (when using one core) is 2.9 GHz, same as the Air.

You can monitor the CPU frequency and power use of your Intel CPU with the Intel® Power Gadget. (Update 2019-7-1: the link doesn't work anymore, but find new links and more background info on the Powergadget in this PCGuide article.)

But let's have a look at some benchmarks first.

Geekbench 2 32-bit

MBA: 5861
MBP: 7110 +21%

Geekbench 3.1.2 32-bit single core

MBA: 2291
MBP: 2703 +18%

These numbers are not particularly exciting, especially considering that I went from a two-year-old "Air" to a brand new "Pro" machine. However, things are better for certain benchmarks and real-world tasks:

Geekbench 3.1.2 32-bit multicore

MBA: 4316
MBP: 5640 +31%

Handbrake 0.9.9 SD video conversion

MBA: 134 seconds
MBP: 96 seconds, 40% faster

Handbrake 0.9.9 HD video conversion

MBA: 270 seconds
MBP: 157 seconds, 72% faster

QuickTime QuickSync HD video conversion

MBA: 131 seconds
MBP: 48 seconds, 2.7x faster

Unfortunately, using QuickTime to convert to 480p, 720p or 1080p using the respective presets, which uses QuickSync, does result in really large files, so in practice it's not all that useful.

Cinebench 15 GPU

MBA: can't run the test
MBP: 20.02

When running the Cinebench benchmark with 4, 2 or 1 CPU threads, it becomes clear that the MacBook Pro is much better at multithreaded tasks:

Cinebench 15 CPU, 1 thread

MBA: 98
MBP: 106 +8%

For some reason both computers stick to a 2.6 GHz CPU frequency most of the time even under single threaded loads. It looks like MacOS doesn't make the second core sufficiently idle to fully benefit from the turbo boost.

Cinebench 15 CPU, 2 threads

MBA: 137
MBP: 195 +42%

Cinebench 15 CPU, default number of threads (4)

MBA: 187
MBP: 262 +40%

The MacBook Pro can do 1.9 times as much work when using both cores, but the MacBook Air only sees a 40% improvement when going from using one to using two cores. The Power Gadget shows why:

The MacBook Air runs into its power limits very quickly when using both cores fully, so the CPU clock has to be throttled back significantly below the maximum 2.6 GHz. Interestingly, the MacBook Air and the MacBook Pro use roughly the same amount of power with the CPU at full utilization: the MBA a little over 20 W, the MBP a little under. However, for the Air this is above its 17 W Thermal Design Power (TDP), so it quickly throttles back the CPU frequency. The MacBook Pro's TDP is 28 W, so it can run at full blast with thermal room to spare.

The fan speeds reflect all of this. When using 400% CPU time (such as when encoding video with Handbrake), the MBA's fan quickly spins up from its regular 2000 RPM, eventually reaching 6500 RPM, which is quite noisy. The MBP's fan runs at 1300 RPM under normal circumstances, and didn't go beyond 2200 RPM with the CPU at 400% for some time, which is barely audible.

Last but not least, let's have a look at OpenCL (not to be confused with OpenGL) performance. Apparently Apple is leveraging OpenCL for more and more tasks. One of the GPUs in the new Mac Pro doesn't even drive any displays and is just there for number crunching.

Unlike most programs, which run on the CPU, or OpenGL, which runs on the GPU, OpenCL programs can run on the CPU and most GPUs, or even on both at the same time. The Luxmark OpenCL benchmark is a bit (14%) faster on the MBP's CPU than on the MBA's CPU.

Luxmark 2.1beta2 OpenCL benchmark, CPU

MBA: 181
MBP: 207 +14%

The MacBook Air has Intel HD 3000 graphics on board, which doesn't support OpenCL. The HD 5100 Iris GPU that's part of the the MacBook Pro's CPU die does, and this is 78% (!) faster than using the CPU.

Luxmark 2.1beta2 OpenCL benchmark, GPU

MBA: -
MBP: 368

Unlike the CPU, the GPU has no problem pushing the i5-4258U past its 28 W TDP, so it has to throttle back after 20 seconds or so. Note that this is with the CPU mostly idle.

Using the CPU and GPU together increases performance another 9%, for a total of 121% over the maximum the MacBook Air can do.

Luxmark 2.1beta2 OpenCL benchmark, CPU+GPU

MBA: 181 (CPU only)
MBP: 401 +121%

I have to be honest: this comparison started from a place of disappointment. Intel didn't move the needle much regarding single core performance on an identically clocked system. (However, two years ago I had to pay extra to get the 1.8 - 2.9 GHz i7, while the 2.4 - 2.9 GHz i5 is now in the base configuration.) But the extra thermal headroom, and probably the faster memory (but less level 3 cache), allows the new Haswell CPU to push ahead in multithreaded performance, which should benefit most number crunching type applications, such as video encoding. And applications that make use of OpenCL and thus the much beefier GPU should benefit a lot. So I'm still reasonably happy with the performance of this computer.

• Battery life
• CPU performance
• Ports and networking
• Audio
• General impression and the screen(s)

Permalink - posted 2013-10-30

Late 2013 MacBook Pro review: battery life

▼ No more bottlenecks: the late 2013 MacBook Pro reviewed

Over the next few days, I'll publish my review of the new MacBook Pro¹ here in several installments. This is the first one, which looks at battery life.

¹ This is of course the Intel Haswell-powered MacBook Pro "with Retina® Display", but the latter pretty much goes without saying now. Which is a good thing, because I really dislike (mis-)using the word "retina" to indicate a high resolution screen, especially when capitalized. I got middle stock model: 8 GB RAM, 256 GB SDD, 2.4 GHz CPU.

Late 2013 MacBook Pro review: battery life

Many of a computer’s features reveal themselves in only a few minutes or even seconds of use. That is not exactly the case for battery life. The amount of time you can work on a battery charge also varies wildly with the type of use. So Apple's claim that this computer can work on battery power for 9 hours was one of the first things I wanted to test. Here are the results of three tests I did.

My first test is the "wait for mail" test. This is the lightest possible workload, so it should provide an upper limit on battery life you can expect. The computer has bluetooth and Wi-Fi on, the screen is set to the lowest non-off brightness, sleep and the screensaver are turned off, and only two applications are open: Mail and Messages. When I went to the System Preferences to turn off sleep, I was a bit confused. Did the Energy Saver pane always look like this?

No, it didn't. Mouse over the image to see the energy pane looks like on my mid-2011 MacBook Air that's also running OS 10.9. (Or click here for the one on the MacBook Pro or here for the one on the MacBook Air.)

(Turns out that when using the power adapter, you can keep the computer from sleeping when the display sleeps, but there's no setting for how long until the computer sleeps.)

So after removing the sleep light a few years ago, Apple took the next logical step and simply lets the entire computer sleep very soon after the display goes to sleep. Which of course helps you get even more battery life. However, this is not so great when it comes to mail and IM clients and the like that run in the background and monitor what happens over the network. But this can be remedied with tools like Wimoweh, that make sure the computer doesn't sleep when you don't want it to. Or maybe not... I was still able to ping the computer even though the caps lock light on the keyboard had turned off, presumably indicating that the computer had gone to sleep.

In this setup, clicking on the battery icon in the menu bar showed that the system expected to be able to run for 20 hours. I had prepared the test setup before I went to bed. I woke up early 6 hours later, but the battery still showed a 100% charge and 20 hours left. After running for 8 hours, this had dropped to 74% and 14 hours, 45 minutes. I checked every hour from then on, and in each case, the battery had depleted another 4%. After 12 hours, I ended the test with 58% charge left in the battery, because I wanted to actually use my brand new computer for something.

Wait for mail: 24 hours (!)

Verdict for the "wait for mail" test: it seems the battery charge percentage wasn't calibrated correctly when full. But this computer can easily run 12 hours on the battery if it's not really doing anything, and probably as much as 24 hours.

Watch video: 6 - 12 hours

The next test was watching video. I tried a few different types of video and looked at the battery percentage to evaluate the battery's performance to avoid having to spend a week running battery tests. For these tests I set the screen to half brightness and turned off Wi-Fi. (I think I forgot to turn off bluetooth.) Headphones were connected for the audio at a quarter volume.

First I wanted to see how this machine handles viewing DVDs. But... it doesn't have a DVD drive! So I hooked up Apple's external USB Superdrive. At first it seemed that this wasn't the best possible idea, as during the first few minutes of watching the DVD the drive was very noisy. But after that, it spun down and only spun up in a more modest fashion momentarily from time to time. Playing back a DVD for 1.5 hours used up 16% battery charge. So it should be possible to watch DVDs for a little over 6 hours.

But obviously this isn't a very realistic test. If you're going to watch a DVD on a computer with only an external optical drive, you'd rip the DVD beforehand and then play the ripped DVD from the internal SSD drive. This is what I did starting at 8% battery capacity left. 45 minutes later, the computer went to sleep. So this means the computer should be able to play back ripped DVDs for more than 9 hours on one battery charge.

Then I played back a movie in 1080p HD resolution in H.264 format. Two hours of this used up 17% battery capacity. So the computer should be able to play back HD video in GPU-accelerated H.264 format for nearly 12 hours on a single battery charge.

Normal use: 9 hours, 42 minutes

Last but not least: how long does the battery last during more typical use? Many professional reviewers have test setups where a script loads a website every 15 seconds or something like that. The good thing is that this makes for a consistent, repeatable test. But it's not real use. What I did is use the computer like I do normally, with two slight tweaks: I made sure I had the computer do something pretty much the entire time, and display sleep was set to "never". I played music most of the time, either using the built-in speakers or streaming over the network to my Apple TV. Most of the time I was writing or reading web pages in Safari, but I also turned on Filevault at some point, so the computer had to encrypt its entire "256" GB SSD.² The computer also slept for 12.5 hours, which used up 3% battery charge. Most of this test was done at night with the screen at low brightness.

When awake, the MacBook Pro used 10 - 12 % battery per hour. Ultimately, it went to sleep after 9 hours and 42 minutes of use. So Apple's battery life claim of 9 hours is completely justified. Obviously, you're not going to get this kind of battery life doing things like video encoding. On the other hand, you can make the battery last longer by not playing music and otherwise minimizing background tasks.

At IETF meetings, there are always power strips strewn throughout the meeting rooms. A decade ago, I'd always connect my laptop to power whenever I could, and run a few hours on the battery otherwise. More recently, I wouldn't bother to connect the power adapter during the morning session, but I'd have to do so during the afternoon sessions at some point. I think with this computer I could easily leave the power adapter in my hotel room and run on the battery the entire day. I typically don't stay in the conference hotel, so this would be taking a slight risk. But with more than a thousand Internet engineers around you can probably bum fifteen minutes of charging in a pinch...

Speaking of charging: the included 60 Watt power supply charges the battery at about 1% per minute, but this slows down after reaching 80%, so charging from dead to full takes two hours. In other words, much the same as charging an iPhone.

Conclusion: the battery life is everything Apple promises and then some. You really don't need to forego the high resolution screen and get a MacBook Air if you want extremely good battery life.

² The SSD holds 250,140,434,432 bytes.

• Battery life
• CPU performance
• Ports and networking
• Audio
• General impression and the screen(s)

Permalink - posted 2013-10-28

Archiving data: DVDs, flash drives, hard drives?

▼ Originally, I intended to write a story under the title "saying goodbye to the optical drive". But after a few minutes of research, it turns out that, in my particular case, the optical drive may be on its way out, but it hasn't quite left the premises yet, and not just because I may need to rip a DVD or CD once or twice a year. Let me tell you a tale of backups and storage management.

I think my first USB flash drive was 128 MB. (Yes, that's MB, not GB.) And my first CompactFlash card 16 MB. So when I got my PowerBook G4 back in 2003, being able to burn a whopping 4.5 GB to a DVD was huge.¹ Back in those days, there was no Time Machine, so I'd often burn important data to DVD to back it up. I also got in the habit of burning video files of stuff I may want to rewatch at some point to DVD, as well as other types of files that I want to keep but don't necessarily need to have on my internal drive. So over the last decade I amassed seven containers with about 20 burned DVDs (or CDs) each. So that's over half a terabyte worth of data.

Then, in 2007 we got Time Machine, which backs up your internal drives to an external one. So with backups out of the way, the question is: where do I archive data that I can't or don't want to keep on my computer's internal drive?

Burning DVDs

In theory, you can burn (write) data to a DVD a bit at a time. In practice, you burn an entire DVD at a time. So you first need to collect all the files that you want to burn, and then it takes about 30 minute for the actual writing and another 15 or so for the system to check whether all the files were written correctly. DVD-Rs and DVD+Rs are slightly different kinds of recordable DVDs, but I think all DVD burners support both types now. They hold about 4.5 GB¹ and can only be written once. There's also dual layer DVD+Rs that hold about 8.5 GB, but they are a good deal more expensive and the burning is a lot slower. There are also DVD-RWs, which you can erase and use again for something else, but they're notoriously unreliable.

So burning DVDs is pretty inconvenient—even before optical drives started to disappear from Apple's computers. These days, you have to buy a separate USB Superdrive (or third party external DVD drive) to read and write DVDs and CDs. However, there is a big upside: they're pretty cheap. If you buy them in bulk, you can get them for as little as 20 cents a piece. Speaking of bulk: a DVD weighs about 17 grams. That's not a lot, but I keep them in containers that hold about 20. Those end up weighing about half a kilo for about 100 GB of data.

If you want to stick with optical disks (discs, whatever) for your own use, then you may want to look at blu-ray. But especially in the Apple world, few people can read them.

External hard disks

Hard disks are everything DVDs aren't. You can just write data without thinking about it, erase, write something else, no problem. And they're fast. HDDs have been limited by the speed of the USB 2 bus for a long time now. I'm very interested to see how fast my latest 2.5" USB 3 drive performs once I get a computer with USB 3. External HDDs have also gotten pretty cheap: that 1 TB one cost me 60 euros.

External HDDs have two downsides. Because they're so big, that's a whole lot of eggs in one basket. And even though they're fast, it still takes hours to transfer all that data, making it hard to make extra copies. The fact that there's an extra box tethered to your laptop all that time doesn't help.

But at least the 2.5" models that get their power over USB are infinitely better than the 3.5" models that come with their own power brick. I had three of those, but in each case either the power brick or the enclosure failed after some years. The drives themselves are fine though, but now I need one of these to use them. Also, the 3.5" drives are more than half a kilo, while a bare 2.5" drive is only 100 grams, 170 grams in the USB enclosure.

Mac - Windows compatibility is also an issue. I formatted my first USB HDD with the FAT file system, which both MacOS and Windows can read and write. But then I accidentally switched the power off, which corrupted the FAT so that new files started to overwrite old ones, so I lost a lot of data without knowing it. So you really need the journaled HFS+ file system, but then you lose Windows compatibility. (And FAT can't handle files larger than 4 GB.)

USB flash drives and SD cards

A year ago, I was shopping for USB flash drives and SD cards, and I thought that it wouldn't be long before those would be so cheap that they're a good alternative for DVDs. But at about 50 cents a gigabyte, we're not there yet.

USB flash drives are even more convenient than external HDDs: they have no moving parts and they're tiny, so they're easy to carry and when connected, don't get in the way of using a laptop. They can be fast, but especially older ones may be slow. You can get 128 or even 256 GB flash drives now, but to avoid the eggs/basket issue, you can also get several smaller ones. Avoiding the biggest and the smallest available gives you the best price per gigabyte, but on the other hand different sizes can be convenient for different things. For instance, I may want to make a copy of my photo library that's 90 GB on a 128 GB drive. Four 32 GB drives wouldn't work so well for that.

USB flash drives are especially useful for exchanging data with other people, or even with devices such as TVs and audio systems. However, for most other uses I prefer SD cards, because they provide a more consistent experience than USB flash drives that come in all kinds of weird form factors. It's also easier to find out whether an SD card is fast or not so fast before you buy it. With USB flash drives I'm reluctant to reformat them as HFS+, because I'm not sure I can format them in a way that allows other systems to read them. With SD cards this is not an issue, I can simply reformat those as FAT in my camera. SD cards weigh almost nothing at less than 3 grams. Another nice little feature on SD cards is the write lock switch.

Conclusion

Let's compare the different options:

	$/€/TB	Weight/TB
DVD±R	45	5 kg
2.5" HDD	60	170 g
3.5" HDD	35	310 g
USB flash	370	150 g
SD card	520	90 g

So burning data to DVD is still one of the cheapest options—but only if you buy DVD-Rs or DVD+Rs in spindles of at least 50. And then you have to figure out how to store them. I've already copied a lot of that data on DVDs to a USB HDD, but I was thinking about having the second copy² on SD cards and get rid of all those DVDs. But considering the price of half a terabyte worth of SD cards, that can wait.

¹ It's really hard to find out how much data fits on a recordable DVD, due to the multitude of definitions of the gigabyte, rounding at various stages, general sloppiness and the need to inflate numbers for marketing purposes. And then there's the file system overhead and a little extra overhead that Apple imposes. But it's not all that meaningful anyway, as you'll rarely be able to fill up a DVD exactly. So I'm using 4.5 GB as a nice, round number.

² Data doesn't exist until it exists in two places.

Permalink - posted 2013-10-14

2013, a good MacBook Pro year?

▼ I bought Apple laptops in 2002, 2003, 2007, 2008 and 2011. So that's one every two years on average. In 2008, I learned the hard way that I really need to have a decent backup computer in case my primary one fails. So a while back I decided that I should probably buy a new computer every two years, and keep the previous one as a backup for another two years. So I'm due for a new laptop this year.

Which brings up the question: is 2013 a good year for buying a new Mac laptop?

One important consideration is CPU speed. The good people over at Primate Labs have compiled a long list of results for different Macs running their Geekbench benchmark. What's interesting here is that single threaded performance for the highest end MacBook Pro increased by 70% between mid-2009 and late 2012. In addition to that, in 2011 the 15" MacBook Pros went from dual core to quad core, which obviously improves multithreaded performance quite a lot. But since then, the speed improvements have been rather modest.

But there's more to life than just a fast CPU. Such as fast I/O. So I went to support.apple.com and looked up the tech specs for the MacBook Pros released since 2006. By 2007, the MBPs were fully 64-bit capable, supported 4 GB RAM, had Firewire 800 and 300 Mbps Wi-Fi. In 2009, the Expresscard/34 slot was replaced with an SD card slot, and the RAM ceiling went to 8 GB. In 2011, the Wi-Fi speed reached 450 Mbps and Thunderbolt was added. Last year Apple added USB 3, and we saw the new high resolution, slimmed down MacBook Pro with Retina Display.

Once we got to see the new MBPwRD (or rMBP or high resolution MBP) last year, it was obvious that this was the future of laptop computing. Or all computing, really: it won't be all that long until all computer screens reach "retina" resolutions. However, the MBPwRD has a few important weaknesses. One is networking. The 300 Mbps limit on my mid-2011 MacBook Air is a frequent source of frustration: transferring files over my home network at 22 MB/sec (on a good day!) is no fun. The MBPwRD should be about 50% faster, and you can plug in the Thunderbolt-to-GE adapter without losing your external screen. But 1.3 Gbps 802.11ac would be really, really sweet.

Speaking of external displays: apparently the MBPwRD can handle two (three?) of those... but only at a resolution of 2560x1600. So either you get to use the built-in high resolution screen, or you can use a nice, big external low resolution screen. Thunderbolt 2 should allow for 4k resolutions on external screens.

In addition to faster Wi-Fi and Thunderbolt 2, based on this year's MacBook Airs and iMacs, we can expect updated MacBook Pros to have better battery life and very fast PCIe-based SSDs.

So for those of us who are still on 300 or 450 Mbps Wi-Fi, USB 2, and low-to-average resolution screens, the new high resolution MacBook Pros that we all think are coming soon will be a very solid upgrade. Apart from the high resolution display and faster Wi-Fi, I'm also looking forward to much better external HDD performance over USB 3. Sure, a Thunderbolt-connected SSD array is going to blow a 2.5" USB 3 drive out of the water, but then again, that USB 3 drive only costs 60 euros and it should still be a good deal faster than its USB 2 predecessor.

On the other hand, Intel's broadwell architecture is coming (although likely not until 2015) and maybe next year will be the year of the MacBook Air with Retina Display...

Permalink - posted 2013-10-09