Navigate: Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors WebNews Home	.	.
Prices: Mainboards ABIT ASUS Chaintech Shuttle Soyo Tyan CPU Intel P4 2.4C-800 P4 2.6C-800 P4 2.8C-800 P4 3.0-800 P4 3.2-800 AMD AthlonXP XP 1700+ XP 2000+ XP 2400+ XP 2500+ XP 2700+ XP 3000+ XP 3200+ Athlon64 Athlon64 3200+ Athlon64 FX-51 Opteron Opteron 240 Opteron 242 Opteron 244 Opteron 246 Memory Corsair Crucial Kingston Mushkin OCZ Search Prices:

LOSTCIRCUITS	SHORTCUTS: A Higher Bus Frequency Test Configurations Memory Subsystem Power Consumption 3D Rendering Energy Efficiency Cinebench AV Encoding: DVD-Shrink, DrDivX, MainConcept AV Encoding: Virtualdub / DivX and SSE4 Gaming: 3DMark '06 FarCry F.E.A.R. The Down and Dirty Dis-Illusions Give Us Some Feedback on this Review
Intel's Yorkfield QX9770 at 3.2 GHz I have the Power!!
(Review by Michael Schuette, November 25, 2007)

The Power Paradox

All around the web, the system power consumption of the Quad Extreme QX9770 CPU has given reasons for concern. Arguably, increasing the chipset frequency from 333 MHz to 400 MHz will increase the power consumption by 20% (on the chipset itself). In addition, the same 20% power increase has to occur by increasing the memory frequency from 1333 to 1600. That's not where it stops, though, because the memory voltage (VDD) has to be raised from the nominal 1.5V to 1.8V on average to achieve 1600 MHz, the realistic memory power consumption at 1600 MHz is 1.2/(1.5^2)*(1.8^2) or 1.728 that of running at 1.5V and 1333 MHz. At approximately 30W power draw going into the memory under load nowadays (at 2 GB system memory, 2 ranks), this alone can can add some 20W to the overall system power equation.

All of this makes perfectly sense in the context of the data shown in various reviews, with a nominal increase in clock speed from 3 GHz to 3.2 GHz adding a few extra percent in power consumption on the CPU level. A few extra percent, though, with the emphasis on "few" does not explain the somewhat inflated TDP of the QX9770. Reason enough to look at the isolated CPU power consumption using the same test platform that already served for the QX6850 and the QX9650.

Idle Power

Windows idle power consumption of the CPU as measured before the VRM. One thing to keep in mind is that with a lowest supported multiplier of 6x, the QX9770 cannot be throttled back below 2.4 GHz, whereas the QX9650 idles at 2.0 GHz and the 266MHz bus CPUs are sitting at 1600 MHz. Still, 20% increase in clock speed do not correlate well with an almost 60% increase in power-consumption.

Max Load

We were getting a little suspicious about what we saw in the "Idle" test but it got worse when we started hammering the CPU with full load.

Max Load power consumption in WinXP-32. What we measured here is outright scary, a 6.7% increase in clock speed causes an almost doubling in power consumption. Intriguingly, the QX9650 comes in at a higher nominal Vcore of 1.321V compared to the 1.246 of the QX9770 (both shown in the BIOS Hardware Monitor)

To get to the bottom of this, we performed a very simple sanity check by running the QX9650 at the same settings as the QX9770, namely 8 x 400 MHz or 3.2 GHz and the results are "nothing but shocking" as THG would put it.

Idle and max load power consumption of the QX9650 and the QX9770 at identical settings on the identical platform.

So... there we have it. The QX9770 draws about 50% more power than the QX9650 under identical operating conditions. Who would have thought....

(BX80557E6300)

next page: => TrueSpace and Render Power =>