windows is dumb OS ?? ):

linux is just a big store of bugs

Um you guys, can't you read? AMD uses less power for the same performance. This is true even for a desktop setup...

Yes, it is true that Intel is faster clock for clock than AMD. But an AMD system uses less power for the same performance... AnIntel 2400 Ghz system is more power hungry than an AMD 2600 Mhz system... although the performace is very similiar.

Sadly some people just don't get it.

If it takes 45nm Intel to beat 65nm AMD on power efficiency then what does that say about Intel ? ;-)

Does it perform 41% less as well ? Because it sure as heck isn't as fast.
TPM is not a relevant measure to me. I just want to know about FPS. And in FPS, I'm sorry but it's Intel that takes the cake right now.
AMD, I loved your Athlon and your XP, but I'm just not interested in power efficiency if it means your product gives me less FPS.
I'm a gamer. For two hours a day I want all the FPS I can get. It's not like the PC is on all day anyway, so don't bash it. Besides, your widescreen TV probably burns as much power as my PC does.

Even twice slower and twice the price, I will always buy AMD.

Intel is a predator

Not even a semiprofessionell would ever dare to present his results like that.
Ok...the AMD fanbois would not even think about it, because somebody says AMD is cool, but...
If somebody makes a benchmark the whole setup should be listed. This way you do not know, if he forgot the coffee machine swithed on or not.
There are many ways to proove somthing, but this is the dumbest I have ever seen (in 35 years he has not learned anything about copyright notices either)

I wonder why the benchmarks were strictly I/O task -based. Could it be that it might mask the DDR2 platform's tendency to fall behind in performance compared to FB memory when the thread counts climb?

This looks more to do with the RAM than anything else. Energy gobbling FB-DIMMs have their place. Smart system architects will know the difference and will choose accordingly.

yep, as the title, why ZDNet. I personally dont like anything from ZDNet because it's the advertising tool for M$ so it's usually tyrant biased. Let's see what does one of its blogger say about herself: "an unblinkin' eye on M$"; WTF anyone feels sick?

back to the test. the owner said that he did review 2 IDENTICAL servers. not 2 top of the line servers. Does it make any sense at all?

Neal Nelson is the only lab in the entire world that claims AMD machines to be more power efficient, they always have, they always will. Their tests are always paid for by AMD. Their hardware is always provided by AMD. So essentially, they are an indpendent mercenary--or a paid pumper. How come they never test Penryn?

I guess it wasn't pointed out that they were comparing a processor that wasn't released (from AMD), to older, 65nm Intel chips.... Instead of their newer processors.

It's a shame that it's come down to power consumption, rather than performance anyway.

"If your real world experience with one-year-old Xeons versus Barcelona that AREN'T EVEN BEING SOLD YET, I'll give you your money back."

Fish n' rat smells abound at this "independent tester". So what if he's not on the AMD payroll (even that's in doubt...)?

AMD needs to release the amazing Quad-collumn abacus. Based on 45-bead architecture. It uses an amazing 0.00 Watts!!! I think it will sell millions.

When it comes to clock to clock ticks on real server benches....AMD wins....
And consumes less power.
Linux shows the real numbers, not Windows, Windows is a dumb OS, anybody knows that.

It uses 45% less power and how much percent less performance? Regardless of the CPU, a turned off computer uses 100% less power, if you don't care about the performance!

Who cares how much energy the cpu draw from the grid? As if we use an efficient cpu, we can save the polar bears?

I personally won't change a cpu just because i want to keep my energy bill down. i probably spent more on cpu than able to save on energy.

And a watch is more power efficent than a laptop, a watch is still totally unable to do the workload that a laptop can do, you got the same issue here!

the link to the power consumption summary is here

http://www.worlds-fastest.com/wf3400.html

These days people pay to get the most expensive and best processor, without actually needing it. Personally, I think it's a great thing if these tests are accurate, I would go for a product that I spend less on each month, which more than likely also gives me more than enough performance for what I use it for. Games included. I play in 1248*1024 so paying top of the line for everything is a complete and utter waste of money and power.

FB-DIMMs consume A LOT of power. It's around 10 Watts per DIMM. There's nothing new, here.

I LOVE 'IntelApparatchNick'!!! Brilliant!

By the way - not all Intel based servers use FB-DIMM - some use DDR2. Intel's latest 45nm Xeon processors are even more power efficient and even faster that the 2006 CPU's used in this claim.

If you want industry standard power efficiency benchmarks, look here: http://www.spec.org/power_ssj2008/results/res2007q4/ 

Intel ApparatchNick

I work for Intel - for another opinion - look here: http://blogs.zdnet.com/Ou/?p=956

as long as you want to pay more money for less performance.

So it's the memory that consumes more power.... not the cpu?

All the power efficiency in the world won't make it faster than Intel.