Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm using Linux and have an old P4 with about 3 GHz clock speed. Will a newer chip that had slower clock speed run my legacy applications faster or slower? I only use one application at a time, an old drawing program (windows app used in wine) that is unable to take advantage of multicore, and have no desire to run anything in the background.

For example - I was looking at used computer that was CORE 2 DUO 1.86GHZ. This is only half the clock speed, but a much newer chip. Will this run my single app faster or slower or about the same?

share|improve this question
unscientifically, a pentium M was about twice as fast as a system with roughly the same clockspeed, and the core and core2 family are a fair bit more advanced. – Journeyman Geek Oct 20 '11 at 7:19
up vote 11 down vote accepted

The clock speed is only partially in charge of 'how much stuff gets done' (a.k.a. 'work'). To compare 2 CPUs you should check out and find both your old and your new CPU and compare the results.

For example, a P4 with 3 GHz clock speed achieves a benchmark result of 491 (see here). the Core 2 Duo with 1.8 GHz clock speed achieves a result of 1115 (see here).

So, the Core 2 Duo is able to do more work in the same amount of time. With that in mind your drawing app should do calculations faster, and it will wait for your input much faster :) (it idles more since it finishes the tasks faster).

share|improve this answer
does anyone know if this cpubenchmark's software is running a multi-core version or single core version of its algorithm? because if so, it would give an unfair disadvantage to a single core processor with regards to a single-core application. – user114558 Oct 20 '11 at 7:08
thats irrelevant. even if you personally don't want to use the nth-core of your new cpu, your system will happily use it for all kinds of things: keeping the system running, handling devices etc. – akira Oct 20 '11 at 7:10
cpubenchmark is pretty bad imo. you should take a look at cpu-world instead – Marcus Ekwall Oct 20 '11 at 8:11
Just above the Pentium 4 result of 491 is the Core 2 Solo 1.4GHz result of 501, so the OP should get at least a 30% speed improvement with a 1.86GHz Core 2 Duo, even if one core is completely inactive. – Mark Booth Oct 20 '11 at 11:23

As been pointed out already, clock rate has very little to do with how well a CPU performs. It all comes down to how many clock cycles that is necessary to execute a given instruction. This is called CPI (Cycles Per Instruction), and it's what measures a CPUs performance.

So even though the Pentium 4 has a much higher clock rate it will be greatly outperformed by the much newer and more advanced Core2Duo.

If you want to crunch some numbers you should check out CPU Worlds benchmarking database.

share|improve this answer

As others have said, clock rate is only marginally useful at best when comparing different CPUs. If you are comparing different versions of the same CPU, then yes faster is, well, faster. :)

As for the "I don't need no stinkin' multi-cores" issue, yes, you do. :) Even if YOU are not actively doing multiple tasks, your operating system has many other responsibilities to perform that will greatly benefit from multiple cores. Would you rather your computer be doing what you told it to do, or doing one of it's background tasks like memory management, file indexing, virus scanning or printing? Maybe you don't need quad cores, but I wouldn't suggest springing for anything less than a dual core processor.

share|improve this answer
If he's using an old kernel distribution of Linux, then it wouldn't benefit (or he wouldn't even be able to use) the multiple cores. – HaydnWVN Nov 1 '11 at 13:50

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .