[MacBook] The Mother of all Benchmarks

[Abstract]

Does one HD have more cache available or something, because those results are counterintuitive, especially since harddrives are harddrives, regardless of system, and in other harddrive tests done in laptops, a 5400 rpm laptop doesn't run as fast as a 7200 rpm laptop, and never does it go faster. 😉

Dual Channel Memory Performance
Something we’ve been told is that using two pairs of the same memory size will allow the Macs to use dual-channel mode, increasing performance. But by how much?

Of the configurations tested, only two – 512MB and 2GB – used two like-paired memory sticks and were running in dual channel mode. Of these two configurations the Xbench Memory Test – one of the few consistent tests – did show a memory performance improvement between 5 and 10%. Overall system performance, however, was minimal, except in rendering in Cinebench: the dual-channel 2GB configuration saw a 15% in OpenGL software rendering and a 35% improvement in OpenGL hardware rendering.

I don't really understand the result here. I can't see how you could test to see whether there's an improvement when taking advantage of the dual channel memory configurations. 😕

The only way I can think of to test this would be to, for example, use 2 x 512MB sticks of RAM (1GB in total) in one configuration, and a 1GB stick in the other in order to get the same amount of RAM in a different manner. Or if both memory slots MUST be filled, then how about comparing a 2x 512MB configuration with a 1GB + 256 MB configuration. The totals aren't the same, but they're similar enough. If the performance of both systems is the same, or if the 1GB + 256MB configuration is slower than the 2 x 512 MB configuration, then you can say that using 2 identical sticks of RAM matters.

 

[nighthawk]

That was what I was basically wondering at in my earlier post. Based on my experience comparing external 7200 and 5400 rpm firewire drives on my mini (which itself has a 4200 rpm drive) I have found a material (in terms of test results) and noticeable (in terms of performance) difference going from 4200 to 5400 and 5400 to 7200.

Intuitively these results make no sense unless there is something choking the performance inside the MB. Maybe the SATA interface. Can't say if Apple did anything on purpose here of course.

SATA has a throughput of 150 Mbits/s whereas Firewire has a throughput of 400 Mbits/s or 800 Mbits/s depending on 400 or 800 connector/device. But the technology is such that it is unlikely that any 2.5 HD completely floods the SATA 150 bandwidth unless there is a second drive/optical running on the same channel. The advantage with Firewire is that even reading/writing mutliple hard drives on the same channel will not significantly choke the bandwidth.

 

[Abulia]

I don't really understand the result here. I can't see how you could test to see whether there's an improvement when taking advantage of the dual channel memory configurations. 😕

Look at the Xbench memory results. The 512MB (dual-channel) config beats the 1GB and 1.25GB memory tests. The Xbench memory test tests the ability to read/write to the memory, nothing else. The speed to read/write to the memory is directly affected by dual-channel being enabled or not.

While I'll grant you that the ideal test would have been to mix and match memory configurations (one 1GB vs 2 512MB sticks), the reality is that I don't have those resource available to me. Again, if anyone would like to send some hardware my way, I'd be happy to re-test. 🙂

In any event, I think the results bear out the minimal performance increase that dual-channel provides over single-channel.

 

[daveaudio]

To resurrect a thread that is a week old. I do Audio work and 7200 rpm drives are pretty much a given. After getting my new macbook, I immediately maxed my ram and was prepped to get a 100GB 7200 rpm Sata drive to replace my stock 60GB. After doing extensive reading of Don's and others tests. I decided to give Seagate a call and ask them what they thought of this issue.
I got two answers of interest. When I presented Don's results the tech proceeded to be a bit perplexed because he felt there would be a speed increase seen across the board.
His response was it seemed there was a controller issue.
My second question was and I did not expect him to give me an answer, was is it possible Apple had crippled the controller for the sake of keeping the Macbook at a certain performance level in regards to the Macbook Pro.
Of course he would not comment saying it was Apple and he could not make a statement.
Seeing as Seagate is a primary source for Apple, for the drives in these machines. If indeed Apple had done, what I previously theorized, indeed he would not be able to comment.
I will add in closing, this kind of crippling of hardware to keep the performance down on the lower lines is nothing new to Apple. They have done this sort of thing in the past.
The other theory I have come to conclusion on is also that cinebench tests are flawed due to the fact that they were originally compiled on non intel based machines and on 10.3.
Let the debate continue.....

 

[JAT]

SATA has a throughput of 150 Mbits/s whereas Firewire has a throughput of 400 Mbits/s or 800 Mbits/s depending on 400 or 800 connector/device.

Isn't that:

SATA: 150MBps
1394a: 400Mbps (50MBps)
1394b: 800Mbps (100MBps)

 

[BWhaler]

Thanks for the information. This was a great post.

The MacBook is a promising laptop. It's just a shame the quality sucks so badly.

Hopefully Apple will fix the heat, wireless, staining, etc, problems in the next revision.

 

[daveaudio]

Thanks for the information. This was a great post.

The MacBook is a promising laptop. It's just a shame the quality sucks so badly.

Hopefully Apple will fix the heat, wireless, staining, etc, problems in the next revision.

I would like to add short of the hardrive issue. I have do not have many if any complaints about my Macbook its very fast and blows my old G4 Powerbook right out of the water.I do watch my cpu temps and it does get a bit warm but then again my old G4 Aluminum Powerbook would get warm as well.

 

[Abulia]

To resurrect a thread that is a week old. I do Audio work and 7200 rpm drives are pretty much a given. After getting my new macbook, I immediately maxed my ram and was prepped to get a 100GB 7200 rpm Sata drive to replace my stock 60GB. After doing extensive reading of Don's and others tests. I decided to give Seagate a call and ask them what they thought of this issue.

Well THAT'S interesting! 🙂

I still like my MB, although it gets a little warm for my liking. I'm also getting some discoloration on my white keyboard. (Grrrr....)

But I really wish I saw better performance from the 7200rpm drive. 😡

 

[EricNau]

[Resuscitating Thread]

I found it interesting that Bare Feats found Photoshop to run much faster in XP than in OS X on a MacBook Pro in their Windows XP Pro versus Mac OS X shootout.

Why would it be the opposite for the MacBook? Am I missing something? 😕

 

[Muriac]

Ok, I just bought a MacBook with the 60GB drive. I've been looking around, but nobody other than Abulia seems to have any experience with a 7200rpm drive in a MacBook. Is it simply that after this set of benchmarks, nobody has bothered to upgrade their MacBook from 5400rpm?

 

[anthonymoody]

Well Muriac, fwiw, as soon as Apple drops a merom into the macbook I'll be picking one up, along with a 2 gig ram upgrade and a 7200 rpm drive...the original benchmarks be damned 🙂

TM

 

[Muriac]

Well Muriac, fwiw, as soon as Apple drops a merom into the macbook I'll be picking one up, along with a 2 gig ram upgrade and a 7200 rpm drive...the original benchmarks be damned 🙂

TM

It's good to hear that at least one person is getting a 7200rpm drive. I find it really odd that wherever people are saying that these results are bogus or make no sense, nobody responds when someone says, "I know it should be faster, but does anyone actually have one?"

Frustrating, given how old this thread is.

 

[Muriac]

Ok, this is my last bump and then I'll go. Just to be sure: nobody has installed a 7200rpm drive in their MacBook since Abulia posted these benchmarks?

 

[Mackilroy]

Ok, this is my last bump and then I'll go. Just to be sure: nobody has installed a 7200rpm drive in their MacBook since Abulia posted these benchmarks?

You know, that's an unanswerable question, really, because there are plenty of people who have bought MacBooks and don't come here. I haven't installed one, personally – I'm still using the stock drive.

 

[Muriac]

You know, that's an unanswerable question, really, because there are plenty of people who have bought MacBooks and don't come here. I haven't installed one, personally – I'm still using the stock drive.

Well, obviously I wasn't expecting some omniscient forum-goer to show up and tell me whether this is true for all MacBook owners. 🙂

I still find it weird that nobody (even just on this board) has done it. In any case, if these are the only numbers I have to go by, I'll just get a 5400rpm drive.