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 Final Showdown

After a dozen or so benchmark results, it is getting more and more difficult to keep track of what happened where and when, and in the final analysis for this review, it mostly matters how the QX9770 compares to the QX9650. To make it a bit easier, we have taken the results of the QX9770 and normalized them to the results of the QX9650. In most benchmarks, as not expected otherwise, the 3.2 GHz version wins by more or less, prime examples are SiSoft Sandra (~ 15.6% gain) and 3DMark’06 (~12.5% gain) whereas most of the other benchmarks are in the same order as the increase in clock frequency, that is 6.7% or less. Needless to say that the scaling is more than adequate since there is hardly ever even a 1:1 translation of frequency to performance in real world benchmarks. Bear in mind also that in some of the encoding benchmarks rounding errors may skew the results ever so slightly (only full seconds are displayed in Dr.DivX) and Virtualdub).

We didn't expect the QX9770 to lose but the actual gain in performance is not earth-shattering. At identical settings (e.g. 8x 400 MHz), both CPUs performed identical

Let’s Play Devil’s Advocate

The biggest trend worldwide is green. Green cars, green fashion, green computing are the buzzwords and green is measured using energy efficiency as metric. If we look back at the overall power consumption and assume that during benchmarks we usually draw maximum power (barring some single threaded games), then we can simplify the equation to get an energy efficiency estimate for both the QX9650 and the QX9770. Even in applications that are not using maximum power, a case in point are the above mentioned single-threaded games, we can apply pretty much the same equation – remember, the Windows “Idle” power consumption is quite a bit higher on the QX9770 than on the QX9650 , even at matching settings.

Bear in mind that we are not trying to come up with precise numbers here, rather, we are trying to give an estimate of how efficient or inefficient the QX9770 is compared to the QX9650 by taking into account the actual benchmark results and the power draw of both CPUs under load. The fact that the numbers we have are power before the VRM doesn’t really matter either because everything was measured on the same identical motherboard, meaning that what applies to one CPU necessarily also will apply to the other one. There are, of course, different ways to show the results but we settled on the following equation:

and show the results as + [%] of energy efficiency for the different workloads used.

On average, the QX9770 loses out to the QX9650 by roughly 40% in energy efficiency. On a system level, the difference is going to be even more pronounced because of the higher power consumption of memory and chipset due to higher frequencies and voltages. On the other hand, if the QX9650 is substituted for the QX9770 and overclocked to the same operating conditions, at least on the CPU level, the efficiency delta becomes negligible with respect to running the same processor at 3 GHz (9x333 MHz) or 3.2 GHz (8x400 MHz).

(BX80557E6300)

next page: => Conclusions and Dis-Illusions =>