What is a good way of testing the speed of a quad core MBP (with SSD) so that I can finally feel how powerful this machine is supposed to be? I didn't get too much time to play with it last night, but installing applications still took a long time and transferring large files from an external hard drive wasn't amazingly fast, so I'm starting to worry that I paid a premium just to have my apps open in 1 second rather than 3.
PLEASE tell me this thing is more than that. Should I do some photoshop work? Garageband? Let me know so I can test it out tonight when I get home. Hopefully I'll be singing it's praises by tomorrow.
Most people have no need or use for 4 cores (nor 8 virtual cores). Cores increase throughput, not necessarily speed. Speed only increases if the work being performed can:
1) Be broken into smaller chunks that can be performed in parallel and combined at the end
2) Programmed by the software you're using to take advantage of this characteristic.
3) Isn't bottle necked by some other system resource such as disk IO, network IO, or RAM.
This is not true for the vast majority of office productivity tasks. Things like browsing websites, email. office documents, copying files, and even most games benefit very little from additional cores. (In some cases they will actually run ever so slightly slower do to the lower total thermal envelope available for turbo boost, but that's another story.)
Encoding video, running multiple virtual machines simultaneously, compiling code, certain scientific simulations and other more intensive tasks benefit immensely from the multiple cores. If you're not doing those sorts of things, then you likely have way more horsepower than you will ever be able to use in a practical way.
A rough way to think of this is as follows. If each core were a car that could go 100 MPH, having 4 of the same cars wouldn't make it any faster to deliver an item from point A to point B. However, if that item were so big that it couldn't fit in a single car, AND that item could be disassembled into 4 pieces with 1 piece in each car being simultaneously delivered and reassembled at point B, it would be 4 times faster than making 4 trips with a single car. (I'm ignoring the 3 return trips for the single car here - it's a rough analogy like I said.) So if you're particular workloads and applications don't break down like this, you'll only see marginal improvements over your prior machine. That's the case for most end users currently using dual core CPUs.
This is actually one of the biggest challenges in computing today. Current manufacturing processes are bumping up against the physical limits of making a single core in a chip go faster. Now most work is being put into adding cores to chips and figuring out ways to divide problems into parallelizable chunks, while performing various incremental optimizations of the single core case. 10 years ago if you wanted a chip to go faster you largely just increased it's clock speed. It's not realistic to do that today without some crazy cooling techniques that are neither affordable or pragmatic.
