Intel Core 2 Duo: affordable and powerful

Situation with Core 2 Duo processors makes no difference from what was before. The reviews and test configurations mention primarily the higher-end and extreme models, whereas the lower-end either get obnubilated or are lost in the whole mass of all novelties emerging onto the market and thus become forgotten undeservedly. But it is just they which are in facts the winners and which will gain most of the money, however trivial it may sound.

In our materials, we have given a quite detailed account of the architectural traits of Intel Core and conducted a large number of tests - from running new processors with math suites up to Core 2 Duo performance in the most favorite gaming tests. Today we using these tests to benchmark the most affordable Core 2 Duo processors based on the Allendale core. As may know, the major and only distinction (of course, apart from the clock speed) of the lower-end series from the Conroe processors is just in the twofold reduction of the L2 cache memory size (to 2 MB). In fact, running a bit ahead, we'd like to note that it hasn't much influenced the performance, which we'll prove today.

Note that we are not taking into account another representative of the series – Core 2 Duo E4300 which, apart from the remaining aspects, has undergone reduction in the system bus speed to 800 MHz. We'll be able to talk about the performance of this processor only after its announcement which, according to the latest information, is to be held in January 2007.

Intel Core 2 Duo E6300 and E6400 – "a sweet couple"

Currently, the prices for Intel Core 2 Duo E6300 and E6400 processors fluctuate within 200$ to 230$, respectively. That means their direct competitors are dual-core AMD Athlon X2 3800+ and 4000+ and, to some extent, their brother Intel Pentium D 945 of 3.4 GHz clock speed. The scattering of prices at distribitors' is wide enough, so let's at least schematically list the above models of the same pricing category.

Much has already been spoken regarding the success of Intel's new architecture. If you are still unaware, you can go deep into examining the articles «Core 2 Duo processors: shock and awe» and «Dual-core Intel Conroe processors». Our goal today is different – to explore the behavior of processors in practice, reveal their overclocking capabilities in conditions demanding for overclocking (to be told later), as well as compare the performance of Core 2 Duo Conroe versus the immediate contenders. That's just what we'll be doing today, but for now let's look at the summary table of specifications.

This table is a proof of our words – the differences among the processors are minimum, and the thrifty user can make up for the clock speed difference through overclocking. Now let's take a closer look at the processors.

By sight, the boxed versions of processors make no difference from one another. Here are two Allendale and one Conroe (Core 2 Duo E6600) processors.

As compared to Pentium D, the form factor of the box has not changed - only the color spectrum and positioning of some components has been changed.

The package bundle is also standard enough. It is the customary Intel's boxed cooler with factory-applied "chewing gum" and a list of operating instructions describing the main rules of using the processor: «Do not chew, do not throw, never heat up in a microwave, and so on

Benchmarking

Test configuration for Intel LGA775:

As usual, we start with synthetic benchmarks.

We see quite a natural result. The indicators for the Core 2 Duo line are rising progressively due to the rising clock speed of each processor. At the same time, we should note that the performance rise from E6300 to E6400 (according to the readings by SiSoftware Sandra) amounts to ~ 13%, and, in view of the fact that the rise of clock speed from E6400 to E6600 is directly proportional to the two previous processors, then the resultant 12% boost points to the minimum effect of the cache memory in this case. A similar picture is seen in PCMark‘05 as well. Overclocking of Core 2 Duo 6300 to 3030 MHz once again proves that the "redundant" 2 MB are not so much needed to the processor. The other tested specimens demonstrate a natural lag, at least in the test for the memory bandwidth AMD processor shows the best result.

We now move on to practical tests of data compression (archiving a folder containing programs and documents of 1 GB total size), as well as audio encoding (540 MB *.wav into *.mp3) and vide encoding (a 74 MB *.mpeg movie into *.avi). Here the faster is the better.

But at archiving the situation is somehow different. While a move from E6300 to E6400 would give only 8% of boost, then the boost gained through moving to E6600 from E6400 is already as much as 16%. Other processors are left well behind.

The influence of cache memory in this test once again shrinks. The processors demonstrate a proportional growth in performance.

The same picture is seen at video encoding either. In fact, this test has always been "biased" towards the high clock speed. It's just due to that Pentium D looks not bad at all as compared to its elder "congeners".

At SPECviewperf 9.0, the alignment of forces does not radically change. Now AMD no longer yields to Core 2 Duo so easily, and because of the poor optimization of the benchmarking suite for multicore processors, the single-core Athlon 64 3200+ Socket AM2 was even able to catch up with Core 2 Duo E6300. The same turned out to be possible for Pentium D 945, due to higher operating clock speeds.

Futuremark's benchmarking suites demonstrate a traditional dependence on video cards, so we don't see deep "drops" in performance. In fact, installation of a more powerful graphic accelerator changes the results somehow. We already were able to observe that in previous materials devoted to Core 2 Duo.

Let's now move on to the favorite stuff - the games.

Quake 3 is very sensitive not only to the performance of the CPU, but to the memory subsystem speed. The result for Core 2 Duo is quite expectable - all is fine with it: at performance and at the system bus which now has turned faster; and with the P965 chipset which does not introduce latency when addressing memory.

Again we see a natural and quite expectable result. The difference between Core 2 Duo processors is weakly felt. The approximate gap between this series and Pentium D and competitor solutions amounts to 25%, at Half-Life 2 the results are almost the same (again due to the relatively "weak" video subsystem – Core 2 Duo would perform much better with a more expensive 3D accelerator).

Final Words

We believe the readers have already made the right conclusions and no more comments are needed regarding all the above. The victory of Core 2 Duo is evident. As regards the purpose of our research, all is also clear. The reduction of cache memory size in E6300 and E6400 to 2 MB has not proved critical. At most tests, the difference is almost not seen, if any. In other words, Core 2 Duo E6300 and E6400 are perfect candidates for the "Must Have" title. They offer affordable price as well as high overclocking capability provided there is a good motherboard and well-arranged cooling system (the latter factor is not as critical as it used to be). Besides, which is the most important, in the pricing category in question there is no competitor which would be able standing on par with the new offspring of the semiconductor giant.

We appreciate LLC PF "Servis" (Dnepropetrovsk, Ukraine) for the processors and other equipment presented for tests.