How Many CPU Cores Do You Need?

In the early years of the new millennium, with CPU clock speeds finally accelerating past the 1 GHz mark, some folks (Ed.: including Intel itself) predicted that the company's new NetBurst architecture would reach speeds of 10 GHz in the future. PC enthusiasts looked forward to a new world where CPU clocks kept increasing at an accelerating pace. Need more power? Just add clock speed.

Newton’s apple inevitably fell soundly on the heads of those starry-eyed dreamers who looked to MHz as the easiest way to continue scaling PC performance. Physics doesn’t allow for exponential increases in clock rate without exponential increases in heat, and there were a number of other challenges to consider, such as manufacturing technology. Indeed, the fastest commercial CPUs have been hovering between 3 GHz and 4 GHz for a number of years now.

Of course, progress can’t be stopped when money is involved, and with folks willing to shell out cash for more powerful computers, engineers set out to find ways to increase performance by improving efficiency rather than relying solely on clock speed. Parallelism presented itself as a solution--if you can’t make the CPU faster, well, why not add additional compute resources?

The pentium EE 840, the first commercially available dual-core CPUThe pentium EE 840, the first commercially available dual-core CPU

The trouble with parallelism is that software has to be specifically written to run in multiple threads--it doesn't offer an immediate return on investment, like clock speed. Back in 2005, when the first dual-core CPUs were seeing the light of day, they didn’t offer much in the way of tangible performance increases because there was so little desktop software available properly supporting them. In fact, most dual-core CPUs were slower than single-core CPUs in a great majority of tasks because single-core CPUs were available at higher clock speeds.

However, that was four years ago and a lot has changed. Many software developers have since been hard at work optimizing their applications to take advantage of multiple cores. Single-core CPUs are actually hard to find and two-, three-, and four-core CPUS are now the norm.

Which begs the question: how many CPU cores are right for me? Is a triple-core processor good enough for gaming, or should you splurge on a quad-core chip? Is a dual-core CPU good enough for the average user, or do more cores really make a difference? Which applications are optimized for multiple cores and which ones react only to specifications like frequency or cache size?

We thought it would be a good time to run some tests with apps from our updated benchmark suite (there are still more to come, too), running the gamut of one, two, three, and quad-core configurations to illustrate what multi-core CPUs really offer in 2009.

This thread is closed for comments
188 comments
    Your comment
  • Wayoffbase
    Very neat article, this is the kind of thing i read tom's for.
  • SpadeM
    Good piece, this will probably clear up some misunderstandings about dual vs. quad core processors. With just a clarification, that this article is based on the same clock speed (2.7Ghz) for ALL processor cores. And so, in some cases where the software isn't optimized for a multi core experience then going up on the Mhz scale is a valid option. If it wouldn't be, why do we overclock :).
  • You have to consider that dual-core processors will often be clocked faster than a quad-core processor at the same price range (Intel Core 2 Duo E8400 @ 3.0GHz vs. Quad Q8200 @ 2.33GHz at $165 on Newegg). Inversely, to get the same clock speed, to get the same clock speed for a quad-core as a dual-core, you will have to pay 2x or more of the dual-core price (E8400 vs. Quad Q9650 - $165 v. $325). You have to ask yourself whether a 30% increase in speed justifies a 100% increase in cost. In certain scenarios, yes, it is worth it. For the average user on a budget, however, dual-core will often be the most cost-effective. Not to mention the over clocking power of mature dual-core processors (but then, an overclocker is not an average user).
  • giovanni86
    Great article, i enjoyed seeing some of the gaming conclusions. I guess my CPU having 4 cores is a bit useless even though on some of your graphs it still shows 2 to 3 frames more in some of them compared to having only 3 cores. None the less though great article i enjoyed it very much.
  • martindamp
    What about the average user running multiple programs at the same time. I often run both virus scanner, office applications and multimedia encoding at the same time. With four cores this runs smoothly, but with only one core it would be a pain.
  • WheelsOfConfusion
    Very informative! I keep seeing comments regarding which applications or games are good for multi-core versus single etc, it's good to have some hard data.
    And that's a neat trick for creating a standardized platform for the tests, eliminating the architectural differences between single and various multi-core processors.

    Since I see a lot of Tom's articles considering power efficiency and read a lot of comments asking for underclock results, it would have been nice to throw some data about power usage with each configuration. Does disabling a core (or three) significantly reduce power consumption? What about temps?
  • Proximon
    I too expected to see some even priced CPU comparison, but this worked out well. Since it's so easy with any quad core system, all we need now is some gamer with a ton of titles to put up a list.

    Oh, such things already exist, whaddya know :)
  • I second martindamp's question... what happens when you run iTunes, lame, antivirus, and winrar on 1-4 cores?

    But what I'm most interested in is what would happen when you move this to a Corei7. It seems to me that some of the apps that see a slowdown while moving to four cores are likely bumping into bandwidth and bus arbitration overheads, as the Q6600 is essentially two C2D's packaged on the same chip, sharing the FSB. The Corei7 eliminates this bottleneck, and I'd be willing to bet the performance decrease from 3->4 cores goes away as well. And when you play around with the i7, you can toy with Turbo and HyperThreading as well, but it'd be most interesting to directly compare the two architectures based on real cores.
  • swyn01
    Nice article. A few comments for thought though. On the first page it is mentioned how single core CPUs often had a higher clock of newer dual-core CPUs. This fact still exists today between dual-core and quad-core. For example, on newegg, a 2.33 GHz quad-core is about $165. A 3.0 GHz dual-core is $168. That's almost a 29% clock increase for the same amount of money. In the gaming benchmarks, if you multiply the dual-core results by 1.29, you will find that this shows an increased performance over the quad-core benchmark. Just like it did years ago (a faster single core cpu was better for gaming than a slower dual-core), this shows that a faster dual-core is still better for gaming than a slower quad core. At least for now. This will change in the future just as it did for single and dual cores, but I'm sure it will take still a few more years before a higher-clocked dual-core is dethroned to a slower quad-core.
  • Summer Leigh Castle
    Is there a comparison between the i7 and a standard 4-core CPU? I was wondering if the real-world gains are there to justify spending more money for a 4-core hyperthreading CPU?
  • enterco
    Well, I don't know how much it would matter, but there is a thing to consider regarding this benchmark. All the tests were made on a C2Q, which has two L2 cache arrays of 4 MB each, and each array is used by up to two cores. I think that is possible that this benchmark would yelded different results in the following scenarios:
    1. the dual-core test would be made on cores 0 and 2, each of the cores using 4 MB l2 cache;
    2. the triple-core test would use an equal L2 cache size for all cores.
  • mitch074
    Xvid 1.1.3 isn't multithreaded! At the same time, Xvid 1.2.1 (which IS smp-friendly) is now out. While it's multithread implementation is far from perfect, it does get a significant boost over 1.1.3 on SMP machines.
  • thexder1
    I just wanted to say that I have seen this same effect when going from single core to dual core and from dual core to triple core. I believe that something that needs to be looked at for the games going from dual core to triple core and with the applications that showed improvements when going from single to dual core I believe that in both cases the improvement could be from having a spare core for the OS and background processes to run on versus having to run them on the same core that the application is running on. In either case I believe that the best for gamers is a triple core and for video conversions you want as many cores as possible at least until the major codecs support GPGPU technology then the CPU will not matter very much for that and when games start using the GPU for the physics and AI then the CPU in games will become less important. At that point the CPU will not really matter much at all but we still have a while before that happened.
  • Sharft6
    i dunno if its possible with the hyperthreading thing but it would be cool to see the difference between 1,2,3,4,5,6,7,8 cores on an i7 :)
  • jtbm
    Multiple cores benefit multiple programs. For example Vuze uses upto 50% of 1 core if it is loaded. Windows + other background programs run all the time on most PCs. This can use up 1 full core. Most games use 2 cores.

    This means that today 3 cores could be utilised on a common PC during work or gaming and 4 cores are advisable to have some spare capacity. 4 cores are fully utilised during video processing or other CPU heavy tasks.
  • apache_lives
    yet again - no one uses a fresh install platform (windows) with nothing else running - we all have MSN, Torrents, Antivirus, windows update, web browsers running, steam, ea games manager running etc - all chew up performance!

    Real life the quads (and better etc - Core i7) would stand out further

    I remember THG had a video - 3.6ghz vs 3.06ghz with HT - unreal tournament 2003 with some other heavy app running showed that with HT is was doable vs non ht - thats the stuff im talking about!

    Next article - how many apps can run while you game! Phenom 1/II vs Core 2 vs Core i7 etc - find the real benefits of a megatasker etc not just one app at a time like we all DONT do. Oh and also Vista and XP and maybe Win7 somewhere in there
  • apache_lives
    jigajigajooYou have to consider that dual-core processors will often be clocked faster than a quad-core processor at the same price range (Intel Core 2 Duo E8400 @ 3.0GHz vs. Quad Q8200 @ 2.33GHz at $165 on Newegg). Inversely, to get the same clock speed, to get the same clock speed for a quad-core as a dual-core, you will have to pay 2x or more of the dual-core price (E8400 vs. Quad Q9650 - $165 v. $325). You have to ask yourself whether a 30% increase in speed justifies a 100% increase in cost. In certain scenarios, yes, it is worth it. For the average user on a budget, however, dual-core will often be the most cost-effective. Not to mention the over clocking power of mature dual-core processors (but then, an overclocker is not an average user).


    Tis why i clock my Q6600 to 3.5ghz - best of both worlds :D
  • yonef
    Have anyone tried the method described by TH to disable the cores? I've tried and it didn't work.
    I made some search online and found that actually this setting in msconfig.exe is not disable cores, this setting is for 'how much cores OS will use while boot' (default is always set to 1 when OS installs) This does NOT mean that your system will run on 1 core!
  • apache_lives
    giovanni86Great article, i enjoyed seeing some of the gaming conclusions. I guess my CPU having 4 cores is a bit useless even though on some of your graphs it still shows 2 to 3 frames more in some of them compared to having only 3 cores. None the less though great article i enjoyed it very much.


    Ever seen an i7 destroy even a Core 2 Quad in GTAIV?
  • apache_lives
    apache_livesyet again - no one uses a fresh install platform (windows) with nothing else running - we all have MSN, Torrents, Antivirus, windows update, web browsers running, steam, ea games manager running etc - all chew up performance!Real life the quads (and better etc - Core i7) would stand out furtherI remember THG had a video - 3.6ghz vs 3.06ghz with HT - unreal tournament 2003 with some other heavy app running showed that with HT is was doable vs non ht - thats the stuff im talking about!Next article - how many apps can run while you game! Phenom 1/II vs Core 2 vs Core i7 etc - find the real benefits of a megatasker etc not just one app at a time like we all DONT do. Oh and also Vista and XP and maybe Win7 somewhere in there


    Yet again why do i get marked down by idiots on these forums when i put foward a valid point/question?
  • arkadi
    Wrong way to get to the right conclusion. No one run one program at a time any more. Hell look at your desktop. How many staff you have running one it simultaneously? Multi thread software or not, with the way I and many others work on a computer, I bet multi core any day. Try the last SSD article as analogy. That was a article that reflect something.
  • RazberyBandit
    Any chance to see the same test done on a Phenom/Phenom II X4? They may or may not scale similarly, seeing as they aren't designed the same as a Q6600, and therefore don't "communicate" quite the same way internally.
  • There are a couple of things to consider here that haven't been mentioned yet.

    Tthe slight speed-increase on single-thread applications, when moving from one to two cores, is easy to explain. On two cores, all the processes running in the background are handled by a different core, so you only measure the application itself.

    The AVG-test is also interesting. Most times, this will be a process runnning silently in the _background_. Now ask yourself this: do you _want_ it to occupy all four cores, potentially slowing your entire system down? Or should it stick to a single core, and leave the other three free for whatever it is that you're doing in the foreground?
    In some cases, multi-threading is a good choice. But in other cases, it seems to undermine the concept of multitasking, especially when the application fails to set a low priority for itself.

    Some benchmarks, like archivers, will probably always scale badly, simply because compression algorythms are hard to split up without sacrificing compression rate. Ofcourse, if you compress multiple archives at the same time, they will scale just as well as any other application.
  • ravenware
    For general desktop computing some of the performances differences aren't noticeable but once games came around to support dual core chips, there was a noticeable impact.

    The biggest increase in performance that was noticeable right away (even on single threaded apps) was the multi tasking.
    I used to run a 3.4GHz P4 2GB machine for my workstation; once our IT department finally sent us new c2d e6550 machines the performance difference was devastating. Everything ran faster; burning DVDs, querying databases, copying thousands files, remote managing a server and listening to mp3s (or sirius) all at the same time was almost seamless.