I’m thinking of getting a new MacBook Pro as a replacement for my 2 year old Mini. I know how much RAM I want and the size & speed of the hard drive. The only question remains is what CPU speed should I get?The choices are: 2.4Ghz and 2.53 – a difference of 133mhz. The price difference between them, however, is $250 making each Mhz about $1.87. That’s the cost of one Grande coffee from Starbucks. But is the price difference really worth the nominal speed boost from the faster CPU?I’m not a heavy user but I don’t accept slowness from my computers. When people have come to me for PC advice, I’d always steer them away from the fastest CPUs because I never thought the premium was worth it. But the idea of spending $2300 dollars on a laptop vs $2500 – a 10% increase – is just that: 10%. I know I won’t get a 10% speed boost. But over the lifetime of the laptop, will get a good return on my investment?Without doing a ton of math, I’m inclined to say the $250 over the life of the laptop will save me more than $250 dollars of my time. It’s just too complicated to do. I’ve already spent more than $250 of my time trying to figure it out.
The fundamental flaw in the logic is that the machine will not run 10% faster, unless the CPU is where your major bottleneck is. In real life, your bottlenecks are disk speed, memory speed, and network speed.
On average you, in particular, use a computer for 3 things: Web, Documents, and Photos. Let’s take each in turn.
Web: Let’s assume typical cable modem, typical latency, typical page size. Odds are you’ll spend about 0.5 seconds downloading content (network speed) and 0.0001 seconds rendering it (cpu & memory speed). Net savings with a 10% faster CPU? 0.00001 seconds.
Documents: Again, assuming averages, time to open will average about 2 seconds (disk speed to load software and document), time spent scrolling around and refreshing screen while typing will again be about 0.0001 seconds (CPU & memory speed). Again we see no clear benefit
Photos: Here we have the most likely benefit from an increase in CPU speed, but again, your heavy uses are more likely to be spent on disk and in memory moving those large images around. Prime example, when I was working on my 2 Gigabyte image, it didn’t take 15 minutes to scale the image due to CPU speed. Besides, you’d probably use your desktop for this anyway.
We’re not talking about multiple cores here, which could spend up multitasking, just a straight MHz increase, so where are your real bottlenecks?
That $250 would be far better spent with more/faster memory. Preventing the system from needing to page data out to disk will help your experience way more than a paltry 10% CPU increase.
Comments on this entry are closed.