Overclocking II – Quad-Core Q6600 at 3.20 GHz
We continue to push for higher clock speeds. Next, the processor hit the 3.20 GHz mark, albeit at a much higher core voltage than its dual-core counterpart. To reach this frequency, we had to increase the voltage by a full 0.100 V to enable it to complete the Prime95 test without issue on a front-side bus of 356 MHz (1424 QDR).
Along with the clock speed, the memory frequency increases to DDR2-854 (427 MHz), while still running at a latency of CL 4.0-4-4-12. That such a dramatically increased core voltage is needed to achieve this speed is due to the Q6600’s design, which uses two dual-core chips. Here, the probability that one of the two cores has lower overclocking potential is twice as high, even though both of the dual-core parts inside use the G0 stepping. Also, two dual-cores dissipate twice as much heat, which in return results in a higher CPU temperature, leading to a higher risk of instability.
| CPU Frequency | 3.20 GHz (+33.3 %) |
| FSB | 356 MHz (1424 QDR) |
| Core Voltage | 1,4125 Volt |
| Memory Multiplier | 2.40x |
| Memory Frequency | DDR2-854 (427 MHz) |
| Memory Latency | CL 4.0-4-4-12 |
Latest Motherboards News
Latest Motherboards reviews
- 12/01 – Seven $260-$320 X79 Express Motherboards, Reviewed
- 07/12 – Ultimate X79? Five £230+ LGA 2011 Motherboards, Reviewed
- 07/11 – Five £130 To £195 990FX-Based Socket AM3+...
- 13/10 – Man Vs. Machine: Four Automatic Overclocking Techs, Compared
- 19/09 – Round-Up: Four Z68 Motherboards From £190 To £260
so what you just said, is the newer stuff is better
i dont know why they take 10 pages to show what can be said in 1 page + a few graphs
On top of that Crysis is meant to be very cpu dependant and prefers 4 cores - it's the way things are going
Hmm, if they'd have gone for a different motherboard they could have gotten the q6600 to 3.6 on air.
Dunno if you have a duff chip or mobo. I have a Q6600 and exactly the same cooler and I can do 3.41 GHz at stock voltage on a Gigabyte 965P-DS3P.
"Its stock clock speed is 2.4 GHz, which it operates at using a comparatively low core voltage of 1.3125 Volts – the lowest core voltage available for this chip"
Wrong my Q6600 is 1.26V although it runs slightly higher in practice.
http://i152.photobucket.com/albums [...] s/Quad.jpg
I guess this shows how OC'ing can vary depending on luck. Even hand picking the best S numbers is no guarantee. A bit of luck (unless you have deep pockets) can be key.
Fortunately for myself, my Q6600 is 1.28v core, and hits 3.6GHz with only slight bump.
Indeed electron migration is a significant issue at high Vcore but realistically most of us overclocking are probably running 6-12 month cycles on our hardware (at least from my experience) and the cost of killing a mid range part every year against taking a top end part is still more cost effective.
That's given only one CPU in the past 15 that I've OC'd have failed (possibly luck?) on me and that was due to a faulty voltage regulator on my motherboard
Very nice review/test. Highly informative. I was gonna buy a 6850 or a quad core but now im just gonna grab the low cost msi board and a 6750 and spent my cash elsewhere.
question, you end up recommending the MSI motherboard, but the test system states that you used the gigabyte for the test. Will i be able to get the same clocking abilities with the MSI?