AMD Phenom II X4 965 BE: Same Speed, Less Power

Page 3 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
I wonder what the idle power would have been with a reasonably rated PSU? The efficiency of the typical PSU drops off at very low power, so they aren't even certified below 20% of rated power. The highest load in this test is below the 20% cutoff for the PSU used, and the idle power is only ~6.5% of rated PSU power. You can tell that the difference between the two chips is small, but the reported idle power is likely to be way higher than it would be with a good 250 W to 380 W PSU.
 
I guess someone should go ahead and clarify things in more straight-forward fashion:

Just because a psu is rated to output up to 1200 watts doesn't mean it's drawing 1200 watts the instant you power on your PC. It will draw power in response to system demands. So if your cpu needs 15 watts less power under load than another given cpu, the psu will draw a proportionally lower amount of power from the power outlet.

So if you hook up a 1200 watt psu to a system that can only consume 250 watts, the psu will never output more than 250 watts. This will be true for a 400 watt, 800 watt, or a 1200 watt psu.
 
I think everyone realizes this. However, a PSU is most efficient at a certain percentage of its max load, with efficiency dropping off on either side of this peak.

Here's the curve for the VX450:
http://www.corsair.com/_images/charts/vx450w_efficiency.jpg

And here for the HX1000:
http://www.corsair.com/_images/charts/hx1000w_efficiency.jpg

So, say you have a system that draws 80 watts at idle. That's 18% load of a 450W PSU. A 450W PSU like the 450VX will be at over 80% efficiency with this load. However, it's only 8% load with a 1000W PSU, which likely results in significantly less than 80% efficiency. The 1000W PSU needs a load of 50%, or 500W to be at its most efficient. If your system only uses e.g. 250W at full load (not uncommon for any mid-range system), then a 450W PSU would be more efficient.

So even though your components use 250W, power usage at the wall might be higher with an over (or under) powered PSU that's not operating at peak efficiency.
 
For now,AMD has one of the best processor line-up when it comes to price/performance (excluding the 965).I've used the PII 955 for my new system(built for games) & up till now,there isn't a game that I can't run smoothly at full graphics settings @ 1280x1024 so,I don't know why anyone would buy intel & Sacrifice longevity(I mean,2sockets less than a year apart,they've got to be kidding)
 
I have just received a news from this article about Six Core Desktop version of AMD Processor... Hmmmmm.... When it will come to the market??? Eagerly waiting for that...

The article only tells us a new stepping of AMD Phenom II X4 965 arrival. Also this will encourage the users who were thinking to buy AMD's highest performed CPU but because of motherboard limitation they could not upgrade...
 
[citation][nom]Cleeve[/nom]I disagree that the 1200w PSU makes a mockery of power efficiency, since it is an 80+ efficiency CPU. What data do you base your assumption on?You can be as annoying as you like, but unless you have data proving that there's a notable amount of difference between PSUs of different wattage but similar efficiency - and I don't believe there is - I will probably ignore your concern.[/citation]

i believe i shuld not be the one to tell you that a bigger psu is much less effcient than a smaller one at a very small load. i mean it is a big difference in power efficiency between a 1000w psu and a 250w when you have a load of 50w for example. for 1000w psu that is 5% load but for 250 it's 20%. the rated efficency is measured between 20% and 80% as far as i know. at 50w the small psu is just entering the ideal load interval but for the larger psu... it may not even reach 20% in full system load (200w). you may ignore me but that does not make you a smart person or a good professional, doesn't? 😉
 
thanks for this nice review... i use AMD Athlon XP 2600+ ... AMD's performance are always great... cost is less than other vendors, but performance is great..!... well..i have found a good online store... which is specialized in selling hard drives... powersupply, processors..... etc... http://www.micropartsusa.com ... i felt that i should let you know that, that's why sharing with you all.
 
well... lemme give you another one... but i also have found it good, though seems it's gonna be launched very soon .... but i have found it!!!.. and wanna share with you all... http://www.buyergen.com well... i liked it's options... best of luck ...
 
[citation][nom]HalfHuman[/nom]You may ignore me but that does not make you a smart person or a good professional, doesn't?[/citation]

That assumes what you're saying matters.
If what you say doesn't matter -- and it is my belief that in this case your concern is insignificant -- then it wouldn't be smart or professional if I paid attention to you. :)

 
Well I now own this chip after owning the C2 140W version for a month. With the 140W version I was able to overclock to 3.8gh @ 1.55 volts ROCK SOLID (prime95, Intel Burn In Tool, 3D mark 2006, etc). With the C3 125W version I am able to overclock to 3.8gh @ 1.4375 volts ROCK SOLID. I idle at 31 celcius and load at 43 celcius using the Thermalright Ultra 120 with DUAL Noctua NB12 fans. I run Windows 7 Professional 64bit. I can not go any higher than 3.8gh. Both version hit a wall at 3.8gh. I do not mess with FSB when overclocking. I only overclock using the multiplier. One has to remember that these reviewers receive hand picked samples directly from AMD, so this is probably why they are able to hit much higher speeds.
 
One more thing to note is that I recall reading an article somewhere that Intel has a similar 6 core plan. So if that's true this is gonna get really interesting.
 
[citation][nom]Cleeve[/nom]That assumes what you're saying matters.If what you say doesn't matter -- and it is my belief that in this case your concern is insignificant -- then it wouldn't be smart or professional if I paid attention to you.[/citation]

your occupation says "writer for thg". that is a bit (actually a lot) misleading in the sense that you do not contribute to anything except being a smartass. you seem to think like a 15 years old and believe that bigger is better. if you think that what others think is insignificant and you may be also inclined to believe that you may be the smartest person in the room. chances are that you are not. you can actually read a lot of books about people that think they are the smartest in the room.

if you were a good professional you would bring arguments to the table. instead you and your rubber duckling choose to splash around in ignorance. your attitude need a bit of adjusting to your age like that 1200w psu must be adjusted to the darn configuration and the purpouse of the test.
 
TDP is not directly related to power usage. It is a parallel stat, but not a direct correlation. Professionals should know that.

TDP = Thermal Design Power (Amount of heat generated during normal operation that needs to be dissipated to keep CPU in normal operational range.

Fail.
 
half human... it doesnt matter if its a 5000000000000000000000000W 80 plus PSU. the power supply will generally only draw what it needs. in fact, if it is 80 plus, thats better. so if it draws 235W it uses... guess what. 235 W. thats it. 235 watts. two hundred and thirty five watts. The size of the wattage (two hundred and thirty five watts) with the powersupply does not (two hundred and thirty five watts) in any way shape or form (t-w-o h-u-n-d-r-e-d t-h-i-r-t-y f-i-v-e w-a-t-t-s) consume more power than the aforementioned two hundred, and thirty five... bananas? no WATTS. Do yourself a favour, go to the hardware store, by a reciprocating saw, and use it. on your Rick Astley Ogglers. AKA your eyeballs.
 
[citation][nom]HalfHuman[/nom]want a rubber duckie ignorant boy? guess it would be simpler to try and do some tests first.[/citation]

Sure!

Turns out it actually makes more sense accept your rubber ducky than to perform meaningless tests suggested by someone who doesn't impact my reality. 😀
 
Oh boy. More efficient but also more costly... granted not by much. I like this!
 
[citation][nom]Anonymous[/nom]test that reciprocating saw. i hear they are pretty good[/citation]

This is Tom's Hardware, but we're into PC hardware... not so much the Home Depot stuff.

But thanks for coming out, sport! 😀
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom