http://browser.primatelabs.com/mac-benchmarks
http://browser.primatelabs.com/geekbench3/2165465
2010 mbp 15" i7 @ 32-bit multicore = 4092
Core M @ 32-bit multicore = 3437
So, 2010 mbp wins in the CPU department. GPU definitely is faster too with the dGPU and dedicated GDDR3.
http://browser.primatelabs.com/mac-benchmarks
http://browser.primatelabs.com/geekbench3/2165465
2010 mbp 15" i7 @ 32-bit multicore = 4092
Core M @ 32-bit multicore = 3437
So, 2010 mbp wins in the CPU department. GPU definitely is faster too with the dGPU and dedicated GDDR3.
http://browser.primatelabs.com/mac-benchmarks
http://browser.primatelabs.com/geekbench3/2165465
2010 mbp 15" i7 @ 32-bit multicore = 4092
Core M @ 32-bit multicore = 3437
So, 2010 mbp wins in the CPU department. GPU definitely is faster too with the dGPU and dedicated GDDR3.
How can the single core scores be worse than multi core?
The GPU is also not too far apart. Depending on settings and CPU limitation the 330M can pull ahead but on playable settings (25-40 fps range) they are quit similar.
Here are some games tested on both. You just have to select archived old GPUs for the 330M to appear.
330M is quite bad by todays standards.
Yikes my i7 2011 MBA has a much better score than the Lenovo...
As stated by previous posters, Core-M CPU performance is quite comparable to the i7-620M if we look at micro benchmarks. In real life everyday applications, Core-M is likely to perform better because of its very quick power state switches and improved branch predictor.
As to the GPU performance, I'd like to wait for more game benchmarks. The Core M has around 2 times more computation power over the 330M, but I doubt that it will be able to make use of this advantage in games because of its limited TDP. Overall, I'd guess that 5300 will fare better in shader-intensive games while 330M should do better in games with simpler shaders and more overdraw.
Scores for geekbench are close:
http://browser.primatelabs.com/geekbench3/compare/608827?baseline=2060685
But you have to keep in mind that the arrandale can maintain this speed without throttling.
So for example, doing a long encode or any cpu intensive task that lasts a long time, will likely be faster on the the arrandale.