The last home computer I bought was a maxed out iMac in late 2010. I recently got a Retina Macbook Pro and thought it would be interesting to compare the machines’ performance for both work and play purposes.
The configuration policy I adopted for my personal purposes a while ago is “buy a maxed-out configuration, it will last a while”. This is true for both the iMac and MBP. They were both maxed out at the time of purchase in terms of CPU and RAM, and I was interested to see how it affects performance and how/if things have evolved in two years.
The hardware specs of the machines:
27” iMac, 2.93 GHz Intel Core i7, ATI Radeon HD 5750 1024 MB graphics, 8 GB 1333 MH DDR3 RAM, OS X on SSD, Windows on secondary HDD
15” Retina Macbook Pro, 2.7 GHz Intel Core i7, NVIDIA GeForce GT 650M 1024MB + Intel HD Graphics 4000 512MB graphics, 16 GB 1600 MHz DDR3 RAM, OS X on SSD, Windows also on SSD
OS X general performance
I didn’t want to spend money on benchmark tools so I just used the first free one I could find in the App Store, which was NovaBench. Here’s the results for both machines.
The MBP score is slightly higher, since it performs better in all aspects except graphics. I admit that it’s still confusing to me why recent Macs (all? or only laptops?) have multiple graphics chips, and how/why the system or user switches between them.
OS X disk speed
Both computers are running the Apple-provided SSD-s. Looks like there has been a huge speed leap in two years. MBP is more than 2x faster for both read and write speeds. I used Blackmagic Disk Speed Test for this benchmark.
This is interesting for games. I used the standard free 3DMark from Futuremark. The iMac was running 32bit Windows 7 from a HDD Bootcamp partition, and MBP was running 64bit Windows 7 from a SDD Bootcamp partition.
My hypothesis was that even though the iMac is 2 years older, it has a stronger graphics chip, but apparently, this is not true. MBP performed better than iMac, though not by a huge margin. I don’t know which of the two graphics subsystems the MBP used for 3D, or whether both of them are even active or usable under Windows.
PCMark (Windows general performance)
This test measures the general performance in Windows. This was clearly unequal, as iMac was running from HDD and MBP was running from SSD, so iMac was disadvantaged and it definitely shows in the score. Still, many of the tests were CPU- or network-bound (e.g browsing and encryption) where the disk should not matter so much. PCMark has many categories, I compared them individually, and MBP beat the iMac in every single category, which is also clearly reflected in the overall score.
More ideas to test?
Is there any more ideas how I could test the performance of these two machines? The Windows tests were clearly unequal, as MBP was Win7 64bit from SSD and iMac was Win7 32bit from HDD. In an ideal world, I would test with the same OS version. I haven’t really kept up with the testing scene recently—any other benchmark softwares in OS X or Windows?