Looks like no one’s replied in a while. To start the conversation again, simply ask a new question.

Level 1

16 points

i5 / i7 MBP / Logic Pro benchmark test

So I have been thinking about consolidating both my Mac Pro and MacBook Pro into one newer MBP i5 or i7. I stumbled on a benchmark test and ran it on both my Mac Pro and MBP.

http://www.evan.se/logicprobenchmark/EvanLogicBenchmark.zip

My MacBook Pro is a 2.2 Core 2 Duo (3.1, not unibody) and I was able to play around 18 tracks by dragging the loop point right before the tracks, un-muting all tracks, starting playback and then dragging each track one by one until it can't played anymore.

I did the same thing with my Mac Pro 2 x 2.66 Xeon (first gen). I was able to get about 28 tracks going.

I got to stop in the Apple store for a short while tonight and was really disappointed by the new i5/i7's with this benchmark test. I actually got better results with the i5 than the i7 (really strange). I tried the i5 2.53 and was able to get about 22 tracks going and on the i7 2.66 I was only able to get about 16 tracks going.

Something is definitely wrong because the i7 should definitely be able to do more than the i5. I for sure thought the i7 would at least be able to match my current Mac Pro.

One more thing is that Logic Pro is not loaded on the Macs in the Apple store, only Logic Express is, so Space Designer was not present in the test at the Apple store. This was even more of a let down because when I was running the test on my Macs, Space Designer was obviously running as well.

Other people have gotten similar results in this thread on Gearslutz:

http://www.gearslutz.com/board/music-computers/371545-logic-pro-multicore-benchm arktest.html

Please run this benchmark test and give some feedback on your results, there might be something I'm missing here...

Mac Pro, MacBook Pro, Logic Studio, iPhone, Mac OS X (10.5.8)

Posted on May 1, 2010 12:21 AM

169 replies

thedomus

Level 1

20 points

May 7, 2010 7:22 AM in response to Jazzmaniac

I'm not sure 'complaining' is the vibe, more 'concerned' that a hefty 2k investment is actually an upgrade for some!

Also Jazz, some reports compare C2D 2.53 with same RAM & HD spec as an i7 uMBP and get very similar performance with that test. Also I would question whether the RAM & HD access times are the culprit for the perceived lack of improvement when an i7 iMac with similar RAM & HD spec (i.e. 4GB ram @1066 / 7200rpm HD) achieves approx 65 tracks (as opposed to around 22 on the i7 MBP) which implies sheer CPU grunt is being measured, not just RAM / HD handling!

So please, don't patronise others and consider what you are actually saying.

Cheers

Dom

May 7, 2010 9:00 AM in response to Jazzmaniac

Jazzmaniac wrote:
I'm not so sure that there is something to fix. EXS24 and ES1 use very different abilities of the system. EXS needs a lot of memory bandwidth and HD bandwidth for streaming. ES1 uses raw floating point performance and can probably rely on cached memory access almost entirely.
So depending on which plugin you use you can determine limits of certain bottlenecks. The test therefore implies that floating point and cache performance have gone up, but that either HD streaming performance or main memory interfacing bandwidth hasn't improved that much. And that's consistent with the technical specification changes between the models.

So please, stop thinking and complaining that something is wrong. Rather consider what you are actually measuring.

There is definitely something to fix, it just may or may not be with the EXS24. You are right that ES1 is a totally different instrument and can't be directly compared with the EXS. But that doesn't mean that there's no problem.

With all other machines with HT, Logic uses those extra cores. The new laptops are the only ones that neither show nor use the HT cores. While there is some streaming involved, it's fairly minimal (look at the EXS instrument, it's tiny) and the test is much more CPU intensive.

That definitely is a problem between the MPB and Logic and definitely should be fixed by either a Logic update or an update for the MPB.

May 7, 2010 9:04 AM in response to djanthonyw

Here's a question - with the ES1 version of the test, how many tracks are C2D MBP owners getting?

The EXS may have something to do with it, but since the machines aren't even showing the right number of cores, I think it probably goes beyond that.

May 7, 2010 10:19 AM in response to thedomus

thedomus wrote:

Also Jazz, some reports compare C2D 2.53 with same RAM & HD spec as an i7 uMBP and get very similar performance with that test.

That's actually exactly along the lines of what I'm saying.

thedomus wrote:
Also I would question whether the RAM & HD access times are the culprit for the perceived lack of improvement when an i7 iMac with similar RAM & HD spec (i.e. 4GB ram @1066 / 7200rpm HD) achieves approx 65 tracks (as opposed to around 22 on the i7 MBP) which implies sheer CPU grunt is being measured, not just RAM / HD handling!

Your assumption that the specs of the iMac and the Macbook Pro are nearly identical is not really valid. The iMac processor does not only have two additional complete cores as you obviously argue, it also has a much larger 2nd and 3rd level Cache, a different memory bridge and a much faster hard drive. Neither memory clock nor HD rpm are good clues for actual performance.

thedomus wrote:
So please, don't patronise others and consider what you are actually saying.

I am trying to be very considerate. I only feel slightly annoyed about threads that use questionable measurements to underline how disappointing Apple's new products are. That's what I call inconsiderate. If those people actually used the new Macbook Pros to do real work they'd stop being concerned. If there really is an issue with the hyper threading, then Apple will surely fix it soon. If it's not an issue but a design choice, then Apple will have a good reason for it. The new machines are excellent, and they work extremely well. So start making music.

Cheers,

Jazz

May 7, 2010 10:31 AM in response to Jazzmaniac

Jazzmaniac wrote:
I only feel slightly annoyed about threads that use questionable measurements to underline how disappointing Apple's new products are. That's what I call inconsiderate.

With Logic, there seems to be little if any improvement from the previous generation of machines, even with two additional (partial) cores. Nothing "questionable" or "inconsiderate" about pointing that fact out.

If those people actually used the new Macbook Pros to do real work they'd stop being concerned.

That's exactly what people are doing. And they are concerned because they aren't seeing an improvement from the previous generation of machines.

If it's not an issue but a design choice, then Apple will have a good reason for it.

You're saying a new machine with no improvement in performance could be by "design choice"?

The new machines are excellent, and they work extremely well.

And you say this based on what? Do you have one of these machines? Have you compared performance with Logic to the previous generation of hardware?

Obviously, they're not terrible machines. But if they are unable to show any improvement with Logic over the previous generation, even with more cores, I can't imagine how you could argue that they are "excellent" compared to previous MPBs.

thedomus

Level 1

20 points

May 7, 2010 11:32 AM in response to Jazzmaniac

Jazz, I appreciate what your saying.
Clearly there are people out there who have been patiently waiting for this uMBP upgrade and probably feel disappointed that the 'power jump' is arguably not what they were hoping for!
Its also frustrating that when planning a computer upgrade there is scant reference on the web to audio dedicated benchmarks hence the Evansbenchmark on Gearslutz, a noble effort to remedy this!
If this is annoying then maybe suggest an improved benchmark using Logic that people can run.
I use Macs everyday for commercial music production, but when I upgrade I like to know I'm getting a worthy bump in power before spending the cash, and thats why these threads appear!
I'm sure your right that Apple will have a look at the HT issue, if there is one, and a fix if warranted will appear... nothing wrong with a nudge in the Logic feedback page!

Cheers
Dom

djanthonyw Author

Level 1

16 points

May 7, 2010 11:48 AM in response to djanthonyw

So what do you think, upgrade and wait for a fix or hold off on upgrading?

May 7, 2010 11:53 AM in response to djanthonyw

Apple can sometimes take a really long time to fix things like this. If it were me, I'd wait unless you absolutely needed a new machine right away.

May 7, 2010 1:00 PM in response to djanthonyw

I had some older benchmarks I created a while back so I updated them and put them up in a new thread. Go crazy.

May 10, 2010 4:17 AM in response to Mike Connelly

Hi all,

after being seriously disappointed with Logic's performance on my new i7 MBP, I decided to 'stress test' Cubase. I knocked out a project on my 3GHz Quad Mac Pro (12 parts Omnisphere, 12 parts Stylus and some hitting plug-ins across all parts) that just stayed out of the red in the performance meter, then loaded the same project onto the i7 MBP. Low and behold, the i7 was able to play it back with only the occasional overload.

Not the most scientific test, I know, but this at least confirms to me that Logic isn't using all the power of the i7 MBP. Not even close. Hopefully there will be fix soon.

S

Pancenter

Level 6

14,674 points

May 10, 2010 4:41 AM in response to Steve Helstrip

Steve Helstrip wrote:
Hi all,

after being seriously disappointed with Logic's performance on my new i7 MBP, I decided to 'stress test' Cubase. I knocked out a project on my 3GHz Quad Mac Pro (12 parts Omnisphere, 12 parts Stylus and some hitting plug-ins across all parts)

Steve, which version of Cubase, 4 or the most recent, 5?

pancenter-

nagi

Level 1

0 points

May 10, 2010 9:38 AM in response to Steve Helstrip

Thank you for your test! Steve
It look like logic's problem

without HT..
MBP C2D / i7(i5) no difference
Logic doesn't support MBP's ix HT

djanthonyw Author

Level 1

16 points

May 10, 2010 10:58 AM in response to djanthonyw

So I guess the question is how long before Logic is updated / fixed? I find it really strange that this is happening because as far as I know, no other program had to be updated to take full advantage of mobile i5/7. Logic works fine with the desktop processors.

May 10, 2010 11:05 AM in response to djanthonyw

Don't hold your breath. Logic still hasn't even been updated to use all the cores on the latest octo mac pros. And those have been out for over a year.

nagi

Level 1

0 points

May 10, 2010 12:06 PM in response to Mike Connelly

It's a huge disappointing.. apple..
Logic producted by steinberg? adobe?

Oh my god!
My first mac(i7 MBP) and Logic 9 is shipping now

i5 / i7 MBP / Logic Pro benchmark test