|
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 
|
LOSTCIRCUITS
|
|
| AMD Athlon64 "Venice" May Low Power be with you! | |
|
(Review by MS May 2, 2005) |
| AMD Athlon 64 4000+ |
CMOS Setup Power Consumption
Contrary to common belief, entering the CMOS Setup does not leave the CPU at idle, on the contrary, since it is a pre-OS operating state, none of the OS-based power management features are active and the CPU incurs considerable load. Since the load appears constant and it is further possible to monitor the core temperature in the Hardware Monitor, the CMOS Setup provides an excellent condition for looking at the impact of temperature on power draw of the CPU under defined conditions.
With both Winchester and Venice cores, we reached a temperature plateau at approximately 28 degree Celsius (38 degrees for ClawHammer and Newcastle), therefore, we disconnected the CPU fan for all measurements to reach temperatures relevant for conditions under full load.
As the CPU die warms up, it draws more current, which could in theory result in a vicious cycle. In practice, an equilibrium is reached at a point that largely depends on the cooling solution used.
It is interesting to compare the different power numbers for the cores we looked at. As expected, the highest power consumtion, along with the highest temperature-dependent increase was found for the Clawhammer, starting at roughtly 53 W and increasing by some 10% over a temperature increase of 18 degrees Celsius to reach 60 W at 47 degrees. As further expected, a delta in core frequency by 9% (2400 => 2200 MHz yields a rather constant 12% delta in power over the entire temperature range measured which conforms to a non-linearity of the power vs. frequency relation.
What is also interesing is the impact of the reduced L2 cache size. The Clawhammer core is 34% larger than the Newcastle, yet, the power consumption is "only" 16% higher. Keep in mind, though, that the cache does not necessarily qualify as prime power hog, even though high-speed SRAMs as used in the on-die cache use quite a bit of power.
At the same speed, the dual stress liner technology used in the Venice core is supposed to yield approximately 24% power reduction. What we see here is an 11.1 % delta but keep in mind that we are looking at the Veince core running at 2.4 GHz and the Winchester core at 2.2GHz. If we factor in the additional 12% power delta for the clock difference between 2200 and 2400 MHz as taken from the Clawhammer results, we end up with a 24.3 % power savings for the Venice core compared to the Winchester core (using the 29 degree Celsius reference point) which is consistent with the IEDM white paper.
|
Athlon64-3500+ (Venice Core) |
next page: => Windows Idle vs. Sandra Burn In =>
All advice and educational articles on LostCircuits are free, but if you feel you can, please make a small donation to us!
