Core 2 Duo vs Core 2 Quad - Overclocking the Quad-Core

Article Index
Core 2 Duo vs Core 2 Quad
Overclocking the Quad-Core
Synthetic Quad vs Duo Performance
Multi-Core In The Real World
Multi-Core Gaming


Both of these processors had quite a bit of headroom when it came to overclocking.  The X6800 comes clocked in at 2.93GHz with an 11x Multiplier.  With modest core voltage increases we actually managed to post at 4GHz but we couldn't boot into Windows at this speed.  We had fair success at 3.8GHz, but settled at 3.641GHz at 1.37v.  At this speed the system was rock solid for a week of 100% CPU load with a Zalman CNPS8000 cooler.

Reference X6800 O/C
Reference X6800 O/C

I didn't know what to expect with the QX6700 processor.  This little beauty comes clocked in at 2.66GHz with a 10x Multiplier.  Our original goal was to make 3.0GHz with the smaller Zalman CNPS8000 cooler, but we reached that at stock voltages and pushed higher.  We ran this processor at 1.37 as well, and managed to reach 3.2GHz rock solid.  We hit 3.25GHz and completed quite a few tests at this speed, but then started to experience system instability.  Once we dropped it down to 3.2GHz, we remained stable for a solid week at 100% CPU load.  At 3.25GHz with a quad core processor, we had 13GHz of screaming Core technology under a single HSF.  When we ran full load tests, we topped out the temperature around 60C with high fan speed.  I believe this is partly why we experience instability.  Better cooling would have confirmed this.  Unfortunately, a better cooler didn't arrive in time.

Reference QX6700 O/C
Reference QX6700 O/C

Test System:

Our test system consisted of the following components:

  • MSI 975X Platinum v.2 "PowerUp Edition" Motherboard
  • Intel Core 2 Duo X6800 CPU - Memory Express
  • Intel Core 2 Quad QX6700 CPU - Memory Express
  • 2GB Kit Crucial PC2-8000 - Crucial
  • eVGA 7950GX2 1GB Graphics Card - Memory Express
  • Seagate 7200.10 SATA II HDD
  • Plextor PX-750A DVDRW
  • Windows XP Professional SP2

On the next page we'll jump into some synthetic benchmarks.