Sign in with
Sign up | Sign in
Your question

Green Machine: AMD Dual-Core Platform at 54W

Last response: in CPUs
Share
September 25, 2006 11:01:38 AM

AMD has introduced low-voltage versions of several of its Athlon 64 models, promising decreased power consumption and heat output for smaller systems. We compare these to standard parts to put this power savings in perspective.
September 25, 2006 11:27:56 AM

useless for desktop. I have no problems with my overclocked singlecore Athlon64
September 25, 2006 12:02:53 PM

Quote:

It's good article though and shows AMD is now looking to decrease their already excellent 90nm power specs. I'm sure 65nm will set power efficiency records.

They have a long way to go to catch Yonah and Merom.
September 25, 2006 12:07:22 PM

Quote:
Were Athlon 64 X2 Energy Efficient processors to reduce average power use by 10 W, the average 10 hour per day office system would cost only few dollars less per year to operate.

This would certainly be a very viable option for alot of schools who are either looking to upgrade from the energy consuming Netburst P4's or from older model PC's or Macs that they seem to be plagued with.
One of the biggest factors for school is price so if this saves money in operating costs, then districts would probably think of changing over.
Then again, Intel or whoever the schools buy their boxes from could be giving them a bigger discount than AMD ever could.
September 25, 2006 12:29:33 PM

Lemme translate the article:

AMD: Since we can't beat Intel with performance per clock, lets change tactics and say we're better performance per watt. But lets not mention enthusiasts could care less about power consumption, most desktop users don't give a darn, mobile wise Intel still kicks our ass, and not say how Opterons do compared to Woodcrest.

Intel: Gaijin Smash!
September 25, 2006 12:30:05 PM

I have to agree with DaSickNinja, my highschool is always trying to save money from electricity I can see where these would be useful processors
September 25, 2006 12:59:10 PM

To be truthful, I don't care if my PC uses 58W or 1KW or even if it needs

or
.
Why? Cause the stove uses more power than the PC does.
But for schools that have boards that pinch pennies down to the erasers and the markers, if any of these schools have even heard of AMD, it might look good.
September 25, 2006 1:40:54 PM

Quote:
I have to agree with DaSickNinja, my highschool is always trying to save money from electricity I can see where these would be useful processors



Where's the savings? The price hike for this processor outstrips any gain in energy savings.

hball
September 25, 2006 1:48:41 PM

I think the article paints a picture of savings not only in the monetary sense, but in noise. If you are a HTPC freak, you might give these processors a serious look, especially when AMD transistions to 65nm.
September 25, 2006 1:56:11 PM

Am I the only one thinking that the guys who wrote this article don't know what efficiency means?

Don't get me wrong: EE Athlons are a great idea. They probably go a long way in enabling fanless systems. But if you want to measure efficiency, or performance-per-watt - which by the way are the same - you should also have compared performance.

Anyone who concludes that the energy efficient-athlons are more efficient because of this article has been mislead: power efficiency was never measured here. The only conclusion that can be drawn here is that EE Athlons consume less power, that's all.

It's easy, it's simple math.

Sorry for the rant.
September 25, 2006 2:33:07 PM

That was the entire point of the article. . . to show how energy efficent the processors were. Not how powerful they were at any specific application. School systems, or whomever chooses to buy this may not be looking for how well it can encode DVD's or how well it can keep up with todays latest games. They just need an energy efficent processor at a decent price. AMD seems to have thier number on this one.
a c 471 à CPUs
a c 118 À AMD
September 25, 2006 2:39:25 PM

Quote:
I have to agree with DaSickNinja, my highschool is always trying to save money from electricity I can see where these would be useful processors



Where's the savings? The price hike for this processor outstrips any gain in energy savings.

hball

Actually, it depends on what the cost of electricity is. In NYC the electricity is about $0.19 per KWH, which I believe is significantly higher than the national average.

At idle, the difference between the Core 2 Duo E6400 and the Athlon X2 3800+ EE is about 26w. Assuming the PC is on 24 hours a day for 365 days that works out to a total of 227.8KW extra per year (26 x 24 x 365 /1000).

Under full load, the difference between the Core 2 Duo E6400 and the Athlon X2 3800+ EE is about 35w. Assuming the PC is on 24 hours a day for 365 days that works out to a total of 306.6KW extra per year (35 x 24 x 365 /1000).

Since electricity is $0.19/KWH in NYC that works out to an addtional cost of between $43.27 (227.8KW x $0.19) at idle 100% of the time, and $58.25 (306.6KW x $0.19) at full load. Therefore, the additional cost of the Athlon X2 3800+ EE can be fully recouped in 1 year.

For people living in areas where the cost of electricity is only $0.07 per KWH that would be a little different. In this case having the "more power hungry" Core 2 Duo E6400 will only cost between $15.94 (100% idle) and $21.46 (100% load) of additional electricity per year. In this case it can take up to 3 years to recoup the extra cost of the Athlon X2 3800+ EE.

From these two examples one can argue that paying more money for an "EE" version of the Athlon X2 3800+ (say $50 more) only makes sense for those people or businesses in areas where the cost of electricty is high. In areas where the cost of electricity is relatively low it only makes sense if the PC will be kept for at least a few years.
September 25, 2006 2:51:48 PM

Thomas,

Thanks for this useful article. I enjoyed it immenesely. Just a couple of questions, though. Where in the hell did you get the ADD3800IAT5CU? You must be a lot closer to Henri than the rest of us. And, buy the way, if you're so close to him, why didn't you get the retail one. The ADD3800CUBOX. After all, what end user would ever buy an OEM CPU, with its one year warranty. Oh, I see, it was all that was available. Dell must be eating'em like crazy? No - wait, they're all going to Apple for the new MAC mini X2? That must be it.

This is the second time I've seen one of these AMD 35 Watt CPUs pop up. The other time was over in Russia, I think. Maybe I'm a little hazy on that.

I noticed that you didn't show any BIOS settings for the ADD3800IAT5CU. In case you took down the data, but just didn't report it, I'd be interested to know if Vcore was 1.025 Volts. That's the Vcore I've read about before. Makes this Athlon X2 look a lot like a Turion X2 - incognito, doesn't it? But, who'd give a damn about that. Perfect for the MAC mini X2. Something, though. Why don't we see any of these in the US retail market? It's been over three months now. The only thing I've seen is advertisments for $400 figment units. I wish AMD would shove some of the retail ones out the door. I'd sure as hell buy one for $150. Maybe a lot of other people would to?

By the way, I noticed that you went with DDR2-533 memory in your tests. Why not DDR2-800?

Orville
September 25, 2006 2:52:19 PM

Quote:
I have to agree with DaSickNinja, my highschool is always trying to save money from electricity I can see where these would be useful processors



Where's the savings? The price hike for this processor outstrips any gain in energy savings.

hball

Actually, it depends on what the cost of electricity is. In NYC the electricity is about $0.19 per KWH, which I believe is significantly higher than the national average.

At idle, the difference between the Core 2 Duo E6400 and the Athlon X2 3800+ EE is about 26w. Assuming the PC is on 24 hours a day for 365 days that works out to a total of 227.8KW extra per year (26 x 24 x 365 /1000).

Under full load, the difference between the Core 2 Duo E6400 and the Athlon X2 3800+ EE is about 35w. Assuming the PC is on 24 hours a day for 365 days that works out to a total of 306.6KW extra per year (35 x 24 x 365 /1000).

Since electricity is $0.19/KWH in NYC that works out to an addtional cost of between $43.27 (227.8KW x $0.19) at idle 100% of the time, and $58.25 (306.6KW x $0.19) at full load. Therefore, the additional cost of the Athlon X2 3800+ EE can be fully recouped in 1 year.

For people living in areas where the cost of electricity is only $0.07 per KWH that would be a little different. In this case having the "more power hungry" Core 2 Duo E6400 will only cost between $15.94 (100% idle) and $21.46 (100% load) of additional electricity per year. In this case it can take up to 3 years to recoup the extra cost of the Athlon X2 3800+ EE.

From these two examples one can argue that paying more money for an "EE" version of the Athlon X2 3800+ (say $50 more) only makes sense for those people or businesses in areas where the cost of electricty is high. In areas where the cost of electricity is relatively low it only makes sense if the PC will be kept for at least a few years.

Very good point there, although schools may not be looking for the best cpu around i think the longevity the better cpu (the e6400) would out due the price one would have to pay to upgrade more often combined with the premium you have to pay for the EE version
September 25, 2006 2:55:42 PM

Quote:
Am I the only one thinking that the guys who wrote this article don't know what efficiency means?

Don't get me wrong: EE Athlons are a great idea. They probably go a long way in enabling fanless systems. But if you want to measure efficiency, or performance-per-watt - which by the way are the same - you should also have compared performance.

Anyone who concludes that the energy efficient-athlons are more efficient because of this article has been mislead: power efficiency was never measured here. The only conclusion that can be drawn here is that EE Athlons consume less power, that's all.

It's easy, it's simple math.

Sorry for the rant.


know what a word means before you criticize its use. heres a link for dictionary.com's definitions, since use it next time: efficiency

ef‧fi‧cien‧cy  /ɪˈfɪʃənsi/ Pronunciation Key - Show Spelled Pronunciation[i-fish-uhn-see] Pronunciation Key - Show IPA Pronunciation
–noun, plural -cies.
3. the ratio of the work done or energy developed by a machine, engine, etc., to the energy supplied to it, usually expressed as a percentage.
September 25, 2006 2:57:45 PM

Its nice to see that you bring nothing to the discussion except negativity. For all your two posts thats all you've done, troll the author. Thank you for showing yourself.
September 25, 2006 3:11:55 PM

who gives a damn about how much power it uses up. How fast is it compared to the 6800? Only 1 benchmark chart for power only? come on, talk about lame! :( 
September 25, 2006 3:14:31 PM

Doughbuy: Maybe you should have read a little closer... Athlon isn't saying that, as the author mentioned, AMD isn't even marketing this feature much at all.

Jaguarskx: Yes, and no... In the summer, ever spend a week with your computer on, then turn it off for a week? What about 2-3 computers? The difference in HEAT output is also quite stunning. I am not a PC gamer any more, and down here in Texas, anything that lets me keep things in my house cooler is a VERY welcome addition to my house. So, if these systems take less power, and give off less heat, then I will not only save on the lower power consumption, but I will save from my AC using less electricity too.

So, even with $0.06, it is still a deal.
September 25, 2006 3:26:53 PM

Since I pay my electric bills I am concerned about the cost of using the PCs at home. I bought a electric meter device called "Kill-A-Watt" which measures the amps and watts per device. I was surprised to see how much power my PC uses daily so I swapped out the power supply with a higher efficency rating such as the Antec H.E. 550 watt power supply. Turns out this new power supply actually use HALF the electricity over the old unit of the same rating! To explain this, my old power supply burns about 300 watts which translate into heat. The new one now uses 150 watts and runs much cooler. Granted, the unit is capable of producing 550 watts but the PC only needs 150 watts at idle and about 245 watts at full load.

So I have to wonder how accurate are the power requirements in PCs these days? And do we really need 1000 watt power supplies? Probably not.

I was very impressed with my friend brought over his XPC Shuttle running AMD 4400+ dual core to see how much power it uses. At idle it uses total of 65 watts and about 110 watts full load. This will make a good media center with MythTV.

Darkk
September 25, 2006 3:28:00 PM

Quote:
who gives a damn about how much power it uses up. How fast is it compared to the 6800? Only 1 benchmark chart for power only? come on, talk about lame! :( 

The point of the article was not to measure performance, if you look closer and actually read you'll see that the entire theme was power usage and efficiency.
People who will be concerned about performance will not look at this processor. Its meant for people who are concerned about how many watts they use.
How fast is it compared to the X6800? It doesn't even match up with the FX-62 with is the competing flagship processor from AMD.
Read the article again and don't just look at the performance benches.
September 25, 2006 3:33:38 PM

Quote:
But if you want to measure efficiency, or performance-per-watt - which by the way are the same - you should also have compared performance.


Yes [Edit]efficiency[/Edit] (and then, performance:watt) matters, but no, they are not the same.

The typical system, even a high-end gaming system, does not run at 100% CPU load (even in the extreme of full throttle gaming, the CPU is not at it's 100% max TDP). The typical desktop probably averages under 10% utilization if not considering a solely-gaming build or otherwise atypical, mission-specific system like video encoding or rendering, etc.

Therefore, efficiency is as always, what the entire box uses. Some things are static for a reasonable comparison- memory, HDD, video, etc, but the CPU and chipset may not be.

Quote:
Anyone who concludes that the energy efficient-athlons are more efficient because of this article has been mislead: power efficiency was never measured here. The only conclusion that can be drawn here is that EE Athlons consume less power, that's all.


The article does overlook a lot, like that AMD had Geode, and low power CPUs for years- it seems to suggest a low power CPU has to have same equivalent current-gen full power part, when it is not often true, rather low power CPU is targeted for a use where it matters more than typical desktop use.

On the other hand, the article was steering towards total system power, and it is nice to see under 60W @ idle though most of us aren't going to be interested in the integrated video. Even so, 45% difference at idle IS substantial and WILL result in lower power usage on the vast majority of systems, even if the performance:watt is lower.

IF you have an exception you will know it is (an exception) because of your extended full load use, even then only if the load is such that it can actually reach 100% TDP, which almost nothing can.

One big problem still remains though, they still produce far too much heat to use a passive heatsink unless you install a Dell & Co. ducted fan alternative cooling strategy. IMO, this misses the market for low power, people are wanting either a desktop that is full on high performance and accept higher heat, or a special purpose system that would ideally have a CPU consuming less than 20W peak. I haven't even mentioned notebooks though, because that's not what was tested.
September 25, 2006 3:34:14 PM

People dont seem to be getting the point of this article. Either that, or they are ignoring it and becoming complete blubbering idiots.
Lets establish right now what the entire point of this article is:
TO SEE POWER EFFICIENCY IN AMD'S NEW EE (energy efficent) PROCESSORS.
That being said, dasickninja this wasnt aimed towards you but i just clicked reply to your post because you seem to be trying to establish the same thing i am.
September 25, 2006 3:38:43 PM

Its all good. Nice to have some backup.
September 25, 2006 3:40:52 PM

Speaking of backup, im suprised B-Matrix hasnt spammed anything. He has limited himself to one post. . . How unorthadox of him.
September 25, 2006 3:42:02 PM

Its quite scary but he may be at work. Wait for about 5:00 EST to see his posts.
September 25, 2006 3:45:40 PM

Quote:
Its quite scary but he may be at work. Wait for about 5:00 EST to see his posts.


I can hardly wait. Back on topic though, im actually glad AMD came out with this, i've been meaning to piece myself together a low power pc that i can run without it sounding like a jet taking off and manage to keep the temps relativley cool. The pc i have now (in my sig) is pretty nice but while its running children could be on fire screaming for help and i would never know.
September 25, 2006 3:46:51 PM

Please, please, please tell me you're joking!

Efficiency in general is:

Efficiency: A / B, where
A: how much of what you want you get
B: how much you have to put in

Clearly, efficiency in this case is A: performance and B: energy consumption.

Think of it this way: a 130W Kentsfield is MUCH more efficient than a 130W Pentium 4 extreme edition. An efficient worker is one who does more in the same amount of time! That already implies that you're looking at some sort of ratio! Oh yes, efficiency should be a ratio! Look at your definition:

Quote:

ef‧fi‧cien‧cy  /ɪˈfɪʃənsi/ Pronunciation Key - Show Spelled Pronunciation[i-fish-uhn-see] Pronunciation Key - Show IPA Pronunciation
–noun, plural -cies.
3. the ratio of the work done or energy developed by a machine, engine, etc., to the energy supplied to it, usually expressed as a percentage.


Is there any ratio whatsoever in THG's article? Go ahead, please, and tell me where.

Meanwhile, I'll tell you exactly what the problem is: THG keeps saying "efficiency, efficiency, efficiency" and concludes the article with, among other things, this quote:

Quote:
Intel's release of Core 2 may have stolen AMD's performance-per-clock thunder, but AMD retains its advantage in performance-per-watt when including the platform.

What? From where did they get this? Certainly not from the data they collected, that's for sure! There's no performance-per-watt ratio (= efficiency) comparison here!

I'm a physicist, so please don't tell me that I don't know what efficiency is. I know, and pretty well too. And I don't have to check my definition in a dictionary: I've seen and used the word many, many times.
September 25, 2006 3:51:25 PM

Uh ... if anyone actually bothered to check things they would realise that Conroe STILL has a higher performance / watt ratio than even these "green" AMDs :lol: 
September 25, 2006 3:54:27 PM

If you take note of the following paragraph from the article you'll find that it should be quite easy to work out your own performance per Watt or whatever else you want:

"Energy Efficient Athlon 64 X2 models provide exactly the same performance as standard voltage versions, allowing us to refer to our Interactive CPU Charts for performance comparisons. This also lets us to limit the scope of this article to measuring power consumption at maximum and minimum CPU load, using our Prime95 torture test."

Just look up the CPU chart results for the relevent result and divide by the Watts for e.g. CPU Points per Watt
September 25, 2006 3:58:46 PM

Quote:
I can hardly wait. Back on topic though, im actually glad AMD came out with this, i've been meaning to piece myself together a low power pc that i can run without it sounding like a jet taking off and manage to keep the temps relativley cool. The pc i have now (in my sig) is pretty nice but while its running children could be on fire screaming for help and i would never know.

Same here, I have a P4 Heatburst 3.78 that doubles as a stove if I don't feel like walking to the kitchen. Some of my friends complain that its to loud but my advice is always the same, turn the speakers up. Thank god for the X6800
September 25, 2006 4:00:39 PM

No doubt, x6800 would look very good in my collection. Unfortunatley i'm cheap as hell. Nbgannet, good job actually paying attention to the article. I'm glad someone actually picked up on that.
September 25, 2006 4:07:31 PM

You know... there is another way to look at performance/watt. I mean, when people say that they are generally talking about full load, but for people who never turn off their computers, and leave them on 24/7, the greatest impact of performance/watt IS when they aren't using it, since when I am asleep and at work, my computer is just checking my email and managing the occasional pidly download.

At which point, it really is pretty much at idle, and the AMD will blow away the Intel processors at a performance/watt scenario. Far far more efficient when I sleep, and I know that without even bothering to check their CPU charts.

<smiles>

:lol: 
September 25, 2006 4:08:57 PM

Quote:
(big snip) From these two examples one can argue that paying more money for an "EE" version of the Athlon X2 3800+ (say $50 more) only makes sense for those people or businesses in areas where the cost of electricty is high. In areas where the cost of electricity is relatively low it only makes sense if the PC will be kept for at least a few years.


Very good point there, although schools may not be looking for the best cpu around i think the longevity the better cpu (the e6400) would out due the price one would have to pay to upgrade more often combined with the premium you have to pay for the EE version

A large fraction of PC users spend nearly all of their time running email and other office apps. The EE 3800+ is more than enough processor for this kind of user, in fact it's overkill. So the 6400, using more power than the EE 3800+ is not really "better". And realize that the typical office PC user does not upgrade their box much. The thing is networked and possibly the occasional USB peripheral is plugged in - most often an MP3 player - and life goes on. The box is used day in and day out till it dies or until IT leadership justifies a mass upgrade. The typical office or student user does not upgrade the CPU - these boxes are simply maintained until declared dead, then the next commodity box is purchased.

Jag is right on about the economics but only HTPC DIY builders will probably pay attention and they will for noise reduction. In order for EE processors to make a real-world difference, OEM builders like Dell, HP, etc., need to be pushing them. Look at it this way: if a low-end CPU will get the job done, save energy and pay back its price premium in energy savings, then philosophically, isn't it the right thing to do?
September 25, 2006 4:09:59 PM

Quote:

At which point, it really is pretty much at idle, and the AMD will blow away the Intel processors at a performance/watt scenario. Far far more efficient when I sleep, and I know that without even bothering to check their CPU charts.

<smiles>

:lol: 


lol. . . Well i dont know about "blow away" but this article shows an apparent power drop with amd when they are on idle. And you do just as i do, i leave my PC on 24/7 or damn near close to it. So when it is on idle i would expect some significant power drops. And its apparent from this article that AMD is giving them.
September 25, 2006 4:13:10 PM

Quote:
Quote:
Quote:

Jag is right on about the economics but only HTPC DIY builders will probably pay attention and they will for noise reduction. In order for EE processors to make a real-world difference, OEM builders like Dell, HP, etc., need to be pushing them. Look at it this way: if a low-end CPU will get the job done, save energy and pay back its price premium in energy savings, then philosophically, isn't it the right thing to do?


Well said; I do belive that is exactly what the article was saying as well. If we can get excellent performance out of inexpensive, power effiecient chips, then why not get them instead? Especially for HTPC builders.
September 25, 2006 4:15:32 PM

Quote:

Yes performance (and then, performance:watt) matters, but no, they are not the same.


I didn't say those are the same, I said efficiency = performance per watt. Read again.

Quote:

Therefore, efficiency is as always, what the entire box uses.


No, that's total power consumption. Efficiency is a ratio and in order to measure it, you'd have to measure both performance and power consumption. If you want to compare efficiency, you need to measure performance. That's what I meant.

Hell, why aren't we discussing VIA then? They have wonderfully underperforming 5W processors that are certainly much more "efficient" by your definition of efficiency = total power consumption. Think about it.
September 25, 2006 4:16:34 PM

Have you seen the price on the EE SFF CPUs! AMD list the price as *** on their website which equates to Froogle prices of over $400. Are they available cheaper than this?
A Merom T5500 has a fairly similar performance to an X2 3800 and costs ~$215 according to Froogle. It will also beat it hands down in the power consumption stakes. I put together a desktop system with a T2600, Gigabyte GA-8i945GM, 2 GB DDR2-667, Samsung 250GB SATA drive and achieved these results:

Idle - 39W
Load @ 1.66 GHz – 50W
Load @ 2.16 GHz – 60W

As a comparison the 35W EE SFF system used 54.5W and 79.5W at idle/load.

Merom does consume a little bit more power than Yonah but not much. The load figure that I achieved was using dual instances of Prime 95 at the most power hungry setting.
I did ‘cheat’ in the test as I undervolted the CPU using RMClock. In some ways this is a fair comparison as the EE SFF parts are in effect factory under-volted and this is the major reason that they achieve lower power consumption.
The system was tested rigorously and was stable at the levels indicated above, but it obviously would not be practical for businesses to do this, so the power saving with Merom would then be reduced.
The advantage of Merom versus AMD EE SFF is that you can get higher performance at a low wattage if you are prepared to pay the price and buy a T7xxx series chip, whereas with AMD you are stuck with the one model for now.

The other thing is that you might as well buy a standard X2 and under-volt it yourself using RMClock or CrystalCPUID rather than buying the EE 65W parts. They just don’t make sense for tweakers, but for businesses users they may be of interest if the pricing is right.
When AMD bring out affordable 65nm X2s, that’s when I’d be interested in buying one and under-volting it myself. I think the current EEs smack a bit too much of marketing BS for my liking.
September 25, 2006 4:20:10 PM

Like Orville, I'm frustrated at my inability to find ANY of the low-TDP Athlons or Semprons in the U.S. It's been months since their announcement, and using froogle.google.com I still can't find any U.S. vendors I trust that say they have them in stock. It would be nice if this story would include a note about the difficulty (to say the least) finding these processors, and the reason for this.

I have a feeling that I'm in the minority here, but it seems to me that one of these CPU's with an integrated-graphics motherboard would make a great low-cost server. The Asus M2NPV-VM used in the article seems the best choice at the moment. ECC memory support + integrated graphics + low-power CPU, attractively priced. Try finding that combo with an Intel cpu/chipset. 975X seems to be the only non-server Intel chipset to support ECC memory, I don't know of a single 975X motherboard with integrated graphics. Once you put a video card (not CrossFire for a server) in a 975X, how's the power draw? And 975X + video card is expensive.
September 25, 2006 4:21:15 PM

Quote:
Uh ... if anyone actually bothered to check things they would realise that Conroe STILL has a higher performance / watt ratio than even these "green" AMDs


Exactly. Isn't that efficiency, by wikipedia's definitions? Look at any one of them and you'll see:

efficiency: useful work per quantity of energy!


Guys, "efficiency" is being misused here. :( 
September 25, 2006 4:26:59 PM

Darkk: the reason power supplies are getting so high is video cards. With ATi x1900's requiring 30amps meaning you have to get a 500watt or better, because smaller power supplies can't or struggle to make 30amps, not to mention the video cards extra wattage needed. If you run a low end vid card with your normal setup then you won't need anything more than 400/450 for a while. Can even get away with a 300/350 but its all component depending.

Mephistopheles: The ratio in the article is Power input/Power usage, or Power usage/Heat production. No where do they mention a comparison in the form of (Power Input/Power Usage)/Perfomance. Your looking beyond all the article is trying to display, its just trying to stat the minimalism of power requirements needed to make an X2 3800+ EE SFF operate. We all know that it will be nowhere near the perfomance levels Core 2, but for the sake of large corporations or anywhere that needs a lot of general use PCs its an ideal part as it has more perfomance than one would need.
September 25, 2006 4:42:45 PM

Yes, well, I still think there is no ratio there whatsoever. If there were, it wouldn't be measured in Watts: watts by watts is still unity.

Anyway, my problem with this article is that it states clearly that AMD is efficiency leader, but efficiency was not compared, only total power consumption. So I'm not actually looking beyond the article, I'm thinking that there's something skewed with what's written in it.

Economical? By all means. Interesting for small systems? Oh yes, my brother will probably want one of those for his audio recording studio. But... "efficient" is another story.

The article even goes so far as to state that AMD wins in "performance-per-watt" without having compared that. I guess they're pulling comments out of their proverbial hats now. That's not nice to the readers.
September 25, 2006 4:48:28 PM

As mentioned in the article, a typical PC might be on for 10 hr/day. I think the article should cover the "sleep" and "off" power draws that cover the other 14 hrs of the day.

My two main machines are an old Barton 2500 multimedia setup and a A64 3400. I measured my energy draw and ran some scenarios a while ago and came up with the following.

Watts in various modes.
Machine Max Idle Sleep(S3) Off Switched Off
A64 3400 122.4 70.8 12.0 9.6 0
Athlon 133.2 108.0 12.0 9.6 N/A

With my use patterns and $.108/kWhr these machines cost $35/$41 to power for a year. I figure 14hr/day in the sleep and off modes.

With the new platforms, as the Max/Idle watts come down, what happens in the Sleep/Off modes? I would be surprised if they still draw 10-12W. If they draw less, the savings will add up 14 hrs of the day and will probably be significant relative to the 10 hrs the system is "on".
September 25, 2006 4:54:46 PM

Wow what is with all of the green stuff? So turn off the TV and run whatever processer you want? To much about about wattage and voltage. I don't care just let my games and data bases fly.. I'll cut down on airconditioning heat and light...just give the me the processing power when I want it!! If this is all AMD has left for marketing they will fail and the market share is going to plumit.. AMD give us a faster processor not and energy saving device!! Net thing you know the will have appliance ratings..

ENERGY STAR Qualified Products: AMD saves energy but don't play FEAR!!
September 25, 2006 4:57:45 PM

I'll give you that one Mephestopheles, i think its an implied relationship, although its not clearly stated, and Watts surprisngly is used in both definitons as a unit measure for power and for heat dissipation, its been a gripe of mine for a while that they need to start to clearly state which is being used in disscusion. But oh well, its just a good choice for the energy conservationists thats all.
September 25, 2006 5:18:51 PM

The point is power consumption. A high end gaming computer can end up using a lot of power. And not everyone uses thier pc for gaming, and not everyone needs thier pc to be the best. Some people like to have thier computers for simple things. Such as email, playing dvd's, perhaps browsing the internet. Those people wouldnt want a power consuming computer that is going to not only run hot, but potentially be loud as hell and end up costing them extra money for power usage when they dont need it. So naturally, the best choice for those people would be a processor that consumes low amounts of power, and yet is powerful enough to run the processes they need it to run. Wether its AMD or Intel. And, if you would of read the article you would of noticed that they are not losing any performance while yet they are consuming less power. So really, its a win win situation.
September 25, 2006 5:20:49 PM

You guys are all being silly. I hate to remind you of this, since you've all seemed to forgotten, but enthusiasts account for a VERY SMALL portion of the computer industry. We get lots of fun stuff, because we spend MUCH MORE than the average Joe. EE processors are not a big deal, but they are a good step in the right direction. MOST people don't need even a dual core CPU. So, most people DO have an overkill situation. And furthermore, a 3800x2 can play Fear just fine, and the GPU will be burning much more power than the CPU, anyway. I find it very cool that I can get any low-power or ultra-low-power CPU these days. You couldn't get one that long ago. Even VIA's chips haven't been available to the HTPC crowd for very long. I'm sorry everyone wants to complain about EE CPU's being overpriced right now. It is again, simple economics. Just like conroe was inflated at launch, there is not enough supply to meet demand right now. AMD's MSRP is worth it, the retailers just haven't got down there yet do to circumstances. Intel is supposed to be releasing a ultra-low-power CPU to the masses soon as well. This is all good for us as a whole. We use too much gas and we use too much electricity. Sure, maybe we can afford it now, but that doesn't mean we'll be able to next year or next decade....
September 25, 2006 5:38:06 PM

For the average workstation, cornrow is just overkill.

I think a point may have been missed in the article. Efficiency per clock cycle is being argued but I think the article was pointing out how EFFICIENTLY the processor meets user needs, not clock cycle efficiency. In that respect, AMD's EE line most definately is the most efficient choice for companies that need savings, not cutting edge performance.

Money being the bottom line for ANY company, not how well they can brag about their computers.
September 25, 2006 5:53:23 PM

Well i dont know about conroe being an overkill. Overkill is the cost/performance ratio being higher than what you need. Something like me buying an FX 60 when all i would be doing is browsing email. Of course, its your money and your perrogative. However, the cheaper conroes offer superior performance at a relativley good price. If you buy anything higher than a 3800 x2 right now with amd your just giving your money away. The point of this being that conroes arent an overkill, especially at the price they are offered at.
September 25, 2006 6:06:04 PM

i would just like to point out that in the uk and much of the rest of the world energy is much more expensive so the saving are larger!

also anyone had a go at ocing these chips yet?
!