As we already wrote, the new line of Intel processors "Core 2 Duo" includes two subgroups — processors with the "E" index (their power consumption is 55-75 W) and currently the only processor with the highest clock in the series (2.93 GHz) and the "X" index that denotes power consumption above 75 W. This index certainly associates with the previous XE-series processors — eXtreme Edition. But unfortunately, the new extreme processor offers no additional bonuses versus regular desktop CPUs — just the highest clock (good) and higher power consumption (not very good...) In view of the above said, we shall not pay much attention to its technical characteristics. Let's proceed right to our tests.

Hardware and Software

Testbed configurations

• Video card — GeForce 7800GTX 256 MB (Gigabyte)
• Memory — 2 GB (2 modules)
• HDD - Samsung SP1614C (SATA)
• Coolers - standard models that come shipped with processors
• PSU - Thermaltake PurePower 680 APD

* — "2x..." means "per each core"
** — in AMD processors it's frequency of the memory controller bus
*** — it's measured differently in Intel and AMD processors, so a direct comparison is not correct

Software

1. Windows XP Professional x64 Edition SP1.
2. 3ds max 7.0
3. Maya 6.5
4. Lightwave 8.5 x64 Edition
5. WinRAR 3.51
6. 7-Zip 4.32 x64 Edition
7. LAME 3.98
8. Monkey Audio 4.01
9. OGG Encoder 2.8 (Lancer)
10. Windows Media Encoder 9 x64 Edition
11. MATLAB 7.1
12. Pro/ENGINEER Wildfire 2.0
13. SolidWorks 2005
14. Microsoft Visual C++ Professional 6.0
15. CPU RightMark 2005 Lite x64 Edition
16. F.E.A.R. 1.3
17. Half-Life 2
18. Unreal Tournament 2004 build 3339
19. Quake 4 Point Release 1.1
20. FineReader Professional 8.0
21. Adobe Photoshop CS2 (9.0)
22. Canopus ProCoder 2.01.30
23. DivX 6.1.1
24. Windows Media Video 9 VCM
25. x264 v.438
26. XviD 1.1.0 Release
27. Apache 2.0.55 for Windows

Drivers

1. NVIDIA ForceWare 91.31
2. NVIDIA nForce SMBus Driver 4.50
3. Intel INF Update

Testing

Some interesting details

But at first, a little lyrical digression. We adhere to the position that any processor should be tested with its issued cooler, if possible. There exist (almost always) better coolers, of course. But 90% of users won't take the trouble, because a similar device is already included into the CPU bundle. This principle is as vital as ever for Intel processors, because they are mostly sold in boxed modifications. But enough of this lyrics.

We got our Intel Core 2 Duo X6800 processor without a cooler. But we assumed that the standard boxed cooler would remain the same, as format of the socket and the retention mechanism was unchanged. So we tested our processor with such a cooler.

A standard boxed cooler with a copper core and
(in our opinion) a not very good retention module for Intel processors.

We should make another lyrical digression here and specify that we ran our tests on a testbed in a usual apartment. In Moscow apartments, unconditioned for several days, the standard daytime temperature was within +25—+30°C. In this case the environment temperature was +28°C. But it did not confuse us - as is well known, the new Intel Core 2 is much less hot than the old one. Boxed coolers were designed for Pentium 4 Prescott processors that broke all heat emission records.

But the results put us on guard, to say the least. CAD/CAE results in the first place (when you see them, you'll understand why). But on the other diagrams the new processors demonstrate much lower gains than expected (to say the least) versus the previous processor. That was when we remembered about a wonderful invention made in Pentium 4 times - Thermal Throttling. We ran the tests together with RMClock (before we had a look at the results, it didn't even occur to us that new Intel Core 2 processors could spring such a surprise with throttling...) Yes! It was throttling! What could we do? Can the X6800 work under such conditions without a water cooling system? Of course, we replaced the cooler in the first place.

GlacialTech Igloo 5071 PWM (on the right). Inexpensive, but good cooler for LGA775, which is screwed to the socket to provide good contact.

Perhaps, you have already got it: the results got noticeably better. We ran the tests one more time to be on the safe side, having added a huge 60 cm fan blowing right to the open PC case. That was unnecessary — the difference from the previous test session was within a measurement error. RMClock was "silent" both times.

That's why diagrams in this article contain two Intel Core 2 Duo X6800 columns. Fiery red, "overheated", with a standard boxed cooler from Intel processors; and blue, "cooled", with GlacialTech Igloo 5071 PWM. Don't think that we advertise this cooler — it was just one of good models at hand. Our tests demonstrate that there are even better models.

Necessary preface to the diagrams

Our test procedure features two peculiarities of data representation: (1) all data types are reduced to one — integer relative score (performance of a given processor relative to Pentium D 805, if its performance is taken for 100 points), and (2) detailed results are published in a table in Microsoft Excel format, while the article contains only summary diagrams for benchmark classes.

3D Modeling and Rendering

Strange as it may seem, overheating has little effect on performance in 3D modeling packages. That may be because of the chaotic load, as interaction takes up most of the time. You can also note that adding 266 MHz produces a lesser effect than it used to: performance difference between E6600 and E6700 models is approximately 10%, between E6700 and X6800 — 7%.

CAD (computer-aided design)

The only group of tests, where throttling "kills" performance. The diagram shows a total score. The results of the overheated processor are very low in two applications out of three: SolidWorks 2005 and Pro/ENGINEER Wildfire 2.0. We can also note that the advantage of X6800 over E6700 is minimal even in case of intensive cooling. But RMClock indicated no throttling. We don't know any other utility that can dynamically indicate throttling for Intel Core 2 Duo processors. We can assume that performance is limited by the video card — CAD programs actively use OpenGL for rendering. We shall analyze this issue in the nearest future.

Compile

The columns are quite even (E6600—6700—X6800). But the difference between the last couple is still slightly lower, so the clock scalability is not ideal anymore. Overheating has almost no effect on performance.

RightMark Project

This situation with performance and throttling is similar.

Processing Bitmap Images (Photos)

Everything is the same again. Pay your attention: we have seen only one case so far, when performance slumps real earnest.

WEB server

Scalability is record-breakingly low. Perhaps, the system is limited by memory? It's a pity we cannot get higher results on the AMD platform — we might have been able to check this assumption then :).

File Packing

It's another case of a serious performance drop due to overheating. Ironically, throttling just levels down the difference between E6700 and X6800 :).

Audio Encoding

We have been suspecting that audio encoding is quite an easy task for modern processors. At least the processor didn't get overheated even in our extreme conditions.

Video Encoding

However, video encoding turned out not a much more complex task. Perhaps it happens because it heavily loads mostly the same CPU units and their dissipated power is just not enough?

OCR

Quite natural results again. Throttling is hardly noticeable again.

3D Shooter Games

Here is a much more serious situation: games prove themselves good CPU heaters. Of course, they are not as good as CADs, but still...

Total score

Efficiency per GHz

Conclusion

So, that's what the extreme Core 2 Duo processor is like: rather fast (no one has ever doubted that) and... rather hot. This hot summer has positively spoilt the triumph of the new Intel processor. In return, we have learnt lots of interesting things about throttling in a CPU with the new core — a simple fact of its presence there and a whole palette of results demonstrated by a slowed-down overheated processor in comparison with a more or less cooled processor. We have also found out that throttling "the beast" is not that bad — most real tasks do not heat a processor that much to put it to "idling speed".

So: don't forget about thermal compound, throw out the issued cooler (if Intel bundles the same trash), buy a good one (the price is nothing compared to that for such processors as X6800) — and Intel Core 2 Duo X6800 will yield as much performance as it is capable of. And its capacities are impressive...

P.S. When it got colder outside, we couldn't help running the heaviest tests (CAD) one more time with the "standard" cooler. We were somewhat consoled that it coped with its task, when the environment temperature was about 21-22°x