A few months ago, we looked into the effectiveness of using different numbers of CPU cores with various types of software. We received a lot of good feedback from that article, and there were some interesting suggestions from the community that we've taken to heart in this follow-up.
Primarily, there was a concern that part one might have been flawed technically, as the Core 2 Quad Q6600 we used in our testing does not share all 8 MB of its L2 cache between its four CPU cores. Intel's Q6600 instead has two separate 4 MB cache repositories, each shared between one pair of CPU cores. This means the quad- and triple-core results would have demonstrated the CPUs utilizing 8 MB of total cache, while the dual- and single-core results show that they were likely benefiting from 4 MB. Indeed, the benchmarks may have been reflecting the difference in L2 cache availability more than performance attributable to enabled processing cores.
To remedy this, we are using a different CPU this time around: AMD's Phenom II X4 955 BE. There are a number of reasons why the Phenom II is ideal for these tests. First of all, its 6 MB of L3 cache is shared between all four CPU cores, so the cache's impact on results will be kept to a minimum. Secondly, since there are now X2, X3, and X4 versions of the Phenom II CPU based on the same die, we will have the opportunity to test the validity of the method we use to simulate fewer CPU cores. By comparing simulated results to an actual retail CPU with fewer CPU cores, we will know more definitively whether disabling CPU cores in the operating system is a truly legitimate test.
At the end of these tests, we will be able to compare the Phenom II X4 results with the ones achieved by Intel's Core 2 Quad Q6600 to see if the impact of shared CPU cache is dramatic or minimal.
A few readers were also interested in simulating a scenario where multiple applications are running at the same time, in order to gauge the benefit of additional CPU cores while multitasking. We therefore ran a new test to analyze this type of scenario, too.
- Time To Follow-Up
- Test Methodology: How Do You Make It A Fairer Fight?
- Test System And Benchmarks
- Synthetic Benchmarks: 3DMark And PCMark Vantage
- Synthetic Benchmarks: SiSoftware Sandra
- Application Benchmarks: Audio Encoding
- Application Benchmarks: Video Encoding And Image Rendering
- Application Benchmarks: General Usage
- Game Benchmarks
- Multitasking Benchmarks
- Simulated Dual-Core Versus Actual Dual-Core Comparison
- Performance Analysis
- Conclusion
Nice article..
it clearly shows an improvement from dual to triple or even quads, and if most people are like myself when it comes to computing and gaming, pulling the best out of the system from the budget available, then the triple and quad core cpu's sure look better than the dual cores.
also down to price and overclocking ability, then the quads from intel can reach virtually the same speeds as the duals and with only a small price difference, making a better price to performance gain overall.
i'm a bit behind the curve and even i'm playing at 1680x1050?
cheers,
bill
p.s. stuff and nonsense: eupeople.net/forum
They should do gaming benchmarks at high deatails so i know how much of a difference a quad cpu will really make on the games.
I expected far better performance from dual / quad cores than a single core - basically like running multiple processors.
So I was expecting twice the performance with dual cores, and 4-times the performance with 4 cores.
I guess this might be expected where the processors were using dedicated caches?
Perhaps it also reveals that Windows isn't correctly taking advantage of the power of 2-4 cores - i.e. the kernel isn't too intelligently dividing multi-threaded / multi-apps capability to multiple cores.
Would be interesting to see a benchmark with Supreme Commander since that's meant to be a well threaded game, and really needs CPU performance when a large number of units are in the game.
Another application which can really use multiple cores is software compilation (e.g. try compiling a kernel with the option: -j 4). But since the compiler itself is not (normally) threaded, but just run several times in parallel, clearly the performance increase is linear.
That's why the socket AM3-based Athlon II/Phenom II x2 seem so appealing, considering their cost. Not saying Core 2 is bad, either, just that I have some Socket AM2 stuff lying around and for a relatively cheap price these deliver killer blows for games, as well as provide a very good overclocking potential. Most of my clients have AMD's due to price restrictions but I have no complaints and at least I know that when I install these parts they're gonna run cool and fast.
We still need faster fingers and eyes to go with the faster computers.