Introduction

Just over two years ago Intel started a comeback with the Core processors. A few months later, summer 2006 the Core 2 architecture was presented to the audience. A real hit as the Core 2 Duo chips were very fast. Although Intel didn't release a processor faster then 2.93 GHz, enthusiasts were able to push these processors to over 4 Ghz. And some lucky ones topped the 5 Ghz barrier.

The months after Intel continued improving their 65 nm architecture. More models (often budget minded) were released. And the Core 2 Quad series followed not that much later. But no big changes were made....

Some interesting Madshrimps reviews and overclocking articles of 65 nm processors:

Intel Core 2 Extreme X6800 Overclocking with Air Cooling

Overclocking the Intel X6800 with Liquid Nitrogen (LN2)

Intel Core 2 Duo E4500 M0 Stepping CPU Review

Intel Core 2 Duo E2160 vs E6300: Budget CPU Comparison

Intel Core 2 Performance, Overclocking, Power Usage Review

Battle of the Mobile CPUs: Core Duo vs Pentium M

Until winter 2007. The 45 nm manufacturing process was finally ready to be commercialized. The biggest advantages were more chips from a wafer (and thus cheaper processors) and also more energy efficient and thus less heat output. But we only saw one processor being released back then. The Core 2 Quad Extreme QX9650. A very expensive chip (+/- 1000 euro), but the performance was better, and more interesting for us, overclockability also. Where a year before 4 GHz was a reasonable overclock with sub zero cooling, we now reached 5 Ghz. And the 6 GHz barrier was also broken.




Some interesting Madshrimps reviews and overclocking articles of 45 nm processors:

Mainstream Quad Core Processors Compared from AMD and Intel

Intel SkullTrail Super High End PC Platform Review

Intel Core 2 Duo E8500 45nm Wolfdale CPU Review

Intel Core 2 on 45nm: Performance, Overclocking, Power Usage

Intel Xeon X3320 45nm Quad Core Overclocking Test

Intel Core 2 Duo E8500 OC: AIR vs Phase vs LN2 Cooling

Intel Quad Core Q9300 Goes Subzero - OC Fun [OCTB]


At the same time Intel also announced a cheaper dual core series manufactured the 45 nm, the E8xxx series and Q9xxx for the quad series. They were very hard to find the first months but then finally, end Q1 2008, beginning Q2 2008 more and more chips became available. One record after another was broken...

And just as before Intel continued optimising their 45 nm architecture. We see now chips reaching 6.3 GHz and higher and also more budget friendly processors, such as the E7xxx series. Recently I bough the cheapest 45 nm processor I could find, the Intel Core 2 Duo E7200.

As there are already plenty of reviews on the internet, I decided to take this processor for a spin, overclocking wise. This is how it looked like before we pushed it beyond reference speeds....



Test platform

Test Setup
CPU	Intel Core 2 Duo E7200 2.53 GHz
Mainboard	Asus Commando
Memory	2 * 1024 Mb DDR2 PC7200 OCZ
Other	Seagate 7200.10 500 Gb SATA Nvidia Geforce 7100 Gs Zalman 1000 watt Power Supply

Let's put it all together. >>>

Cooling



More about Dry Ice on Wikipedia and if you’re interested in cooling your CPU with Dry Ice yourself, our in-house guide will surely be helpful.

Dry Ice is quit easily available and not very expensive. For about 12 Kg Dry Ice I paid 10 euro. You can bench for over an entire day, play with some ice and use a bit the day after. It can be kept in a cool box for about 2 to 3 days.

Installation

Dry Ice itself doesn't turn into water. But as it's so cool you'll get condensation. To prevent that you need to insulate (keep away from air that can contain humidity).
I used some foam I cut to the correct size and squeezed around the socket. This is far from perfect, but sufficient for a couple of runs.



Once ready I installed the copper tube on the cpu and poured a bit of acetone in the tube. As Dry Ice pellets themselves don't make good contact with the copper they can't extract the heat very well. Just for information, the Dry Ice doesn't put cold to the tube, but to sublimate (= turn to gas) it needs energy (= heat) from its direct environment. That's how the tube and CPU will get cold. But adding a fluid that's doesn't freeze easily you can make much better contact. Here you see the temperature measured with a k-probe and Fluke thermometer on the bottom of the tube.



The tube installed and cooling down....



Just after adding new Dry Ice there is more energy (heat) in the tube and you see for a few seconds a lot of gas sublimating.



This is what we see in Bios with the processor at default speed and voltage. Once we'll raise the voltage CPU temperature will raise to about -40° C.




On to the results. >>>

Overclocking Results

To get the maximum out of this processor we used the following settings in BIOS,



Results on air: 4200Mhz

First I installed the setup with the boxed cooling. After some playing with it I could hit a frequency of 4.2 GHz. This was reached using a mere voltage of 1.475 Vcore. The little boxed cooler could even keep the processor around 65° C. This was stable enough to survive some basic benchmarks as SuperPi 1M calculation. But not stable for daily use. Lowering the voltage with 0.025 volt and the frequency with 200 Mhz to 4 GHz I could obtain a perfectly stable system.

Results on Dry Ice: 5225Mhz

Dry Ice cooling made it possible to user higher voltages as heat wasn't really an issue any more. The bios could handle a maximum of 1.85 Vcore. Added a minor voltmod and vdroop mod for voltage stabilisation the Vcore was set to 1.845 Vcore. The setup booted at 4275 Mhz (9,5 * 450 Mhz) and the frequency was raised using setfsb in Windows. The maximum frequency reached was 9.5 * 550 Mhz, which gave us 5225 Mhz. Already a very impressive overclock. As I didn't have more test equipment at the time of testing, there also is a chance that this was the limit of the board rather than of the processor.

To be able to run a few benchmarks I had to lower clockspeeds by some Mhz, to about 5170 Mhz. At this speed I could run Superpi 1M:




But to consider it really stable I also now lowered the CPU frequency. Finally I set it to 5.05 GHz, which equals a FSB of +/- 531 Mhz.

In the end this is a 100 % overclock. Surely not bad. I ran some quick benches for those who would like it. This is what you can expect from this chip with default clock speeds, with the maximum air overclock (+/- 4 GHz) and with Dry Ice (+/- 5 Ghz)


Time in seconds, lower is better...


Max frequency reached with saved print screen ;-) :




Conclusion

The E7200 processor is a more than a decent overclocker. Even with the boxed cooler and a minimal voltage raise I was able to add more then 50% clock speed. 4 Ghz everyday stable with the boxed cooler is surely possible. You just need memory and a board that can handle just over 400 Mhz FSB. But most boards with some overclocking options can handle this. I also tested this CPU with the Asus P5K (about 85 euro) and that worked flawlessly.

But if you want to overclock the chip higher the motherboard could be the limiting factor. To reach 5 Ghz you already need more then 525 Mhz front side bus. This nowadays isn't that difficult anymore, but it's more difficult to find a board that can do that. Our Asus Commando priced at +/- 125 euro had no problems with that.

With a price tag of 115 euro the Intel Core 2 Duo E7200 surely is worth considering for those who are on a budget, but still want to do some overclocking.