We've been looking at Apple's latest 15in. and 13in. MacBook Pros, which include Intel's second-generation Core ix (Sandy Bridge) processors, updated integrated and discrete graphics, and the brand-new high-speed Thunderbolt I/O technology. We don't have any Thunderbolt peripherals to evaluate yet (products are expected this summer from Promise and LaCie), but we've taken the opportunity to run some benchmarks to complement those already reported by our CNET US colleagues.
Processor performance To assess the impact of the new Sandy Bridge processors, we ran the Cinebench 11.5 CPU test, in both multi-core and single-core mode. The 15in. MacBook Pro runs a quad-core 2.2GHz Core i7-2720QM, while the 13in. model has a dual-core 2.7GHz Core i7-2620M. For comparison, we tested a 2008-vintage 24in. iMac based on a 2.8GHz dual-core Core 2 Duo with 2GB of RAM.
The Cinebench 11.5 CPU test scene contains some 2,000 objects, which in turn contain more than 300,000 polygons in total (longer bars are better)
The advantage of the 32nm Sandy Bridge architecture is evident when we look at the single-core results: the 2.7GHz and 2.2GHz Core i7 CPUs in the 13in. and 15in. MacBook Pro perform 53.6 and 48.8 percent better respectively than the 45nm 2.8GHz Core 2 Duo in the 2008 iMac. In multi-core mode, the dual-core (2.7GHz) Core i7 delivers a 98 percent improvement on the dual-core (2.8GHz) Core 2 Duo, while the quad-core (2.2GHz) Core i7 in the 15in. MacBook Pro is in a class of its own, delivering 3.5 times the performance of the 2008 iMac.
Graphics performance Turning to the graphics subsystem, Cinebench 11.5's OpenGL test exposes a big difference between the two 2011 MacBook Pros: the 13in. model uses Intel's integrated (on-die) HD Graphics 3000 with 384MB of shared video memory, whereas the 15in. model has the much more powerful AMD Radeon HD 6750M with 1GB of dedicated video memory.
Cinebench 11.5 uses a complex 3D scene depicting a car chase; the graphics card displays a huge amount of geometry and textures, plus a variety of effects including environments, bump maps, transparency and lighting (longer bars are better)
The 15in. MacBook Pro's 35.8fps result represents perfectly acceptable graphics performance, but the same can't be said for the 13in. model's 11.5fps — or indeed the 2008 iMac's 14.8fps (with a discrete GPU).
Battery life We estimated battery life for the 13in. and 15in. MacBook Pros by fully charging their batteries and then measuring the power consumption under a variety of conditions using a Voltcraft VC940 Plus multimeter. Dividing the average measured power consumption (in watts) into Apple's reported battery capacity (in watt-hours) gives a figure for the expected battery life.
We ran four tests for each notebook: idling at the Mac OS X desktop with the screen at 100 and 50 percent brightness, and fully loaded (running Cinebench 11.5's CPU test), again with 100% and 50% screen brightness. The results give a spread of battery life estimates, within which most real-world usage patterns should fit.
Battery life estimates for the 15in. and 13in. 2011 MacBook Pros with different workloads and screen brightness settings. Shown in red are the results reported by CNET US's video playback battery drain test
No-one is likely to keep their MacBook idling with the screen at 50 percent brightness, but if you did you'd get around 10 hours (15in. model) or 11 hours (13in. model) of battery life. Similarly, no MacBook is going to be running workloads as demanding as the Cinebench 11.5 CPU test all of the time. Again, if this was the case, and with screen brightness at 100 percent, both notebooks would only manage around an hour (15in. model) to 1h 17m (13in. model) of uptime. Apple's battery life estimate for both notebooks of 7 hours seems reasonable, although you can expect considerably less if you run demanding workloads.