Sign in with
Sign up | Sign in
Your question

Which is the Best Mainstream CPU?

Last response: in CPUs
Share
May 4, 2007 2:06:18 PM

AMD and Intel are cutting prices faster than ever, which has moved high-end CPUs into the $200 range. We compare the Core 2 Duo E6400 against AMD's Athlon 64 X2 5600 , to see which might be the better choice for you.

More about : mainstream cpu

May 4, 2007 2:14:03 PM

Quote:
AMD and Intel are cutting prices faster than ever, which has moved high-end CPUs into the $200 range. We compare the Core 2 Duo E6400 against AMD's Athlon 64 X2 5600 , to see which might be the better choice for you.


x2 3600 and x2 5600 for AMD, e6320 and e6600 for intel. They are good deals, you just have to decide if you will overclock.
May 4, 2007 2:23:12 PM

The new 60/90 min power consumption test is very interesting, because it clearly shows how the X2s, even consuming more power at full load, overall have a lower mixed power consumption. This should be because Cool'n'Quiet is more effective than SpeedStep in the low power consumption modes; X2s can go as low as 800 MHz while Core2s are stuck at the double; 1600 MHz lowest.
Related resources
a b à CPUs
May 4, 2007 2:35:11 PM

The power consumption is the main reason I might build an AMD system this fall. I don't do anything high-end, so I'm not looking at CPU's over $150.
May 4, 2007 2:42:57 PM

in my country you can get an e6600 for 197€ anyway, so there really is no competition here :) 
May 4, 2007 2:49:41 PM

Quote:
in my country you can get an e6600 for 197€ anyway, so there really is no competition here :) 

Then you should also get ~197€ X2 6000+s :wink:
May 4, 2007 3:04:08 PM

In a thread not long ago, after several folks tried to paint the x2 6000 as a power-hungry toaster, etc., I was pointing out that most x2 6000s will end up spending a lot of time at idle with QuietN'Cool, and that the max power draw wasn't so much of an issue. I even calculated the electricity cost difference for heavy use over time, which was interesting, since although heavy use isn't that common, it's still interesting to folks who like to look at numbers.

I think it's finally a no-brainer at this point to choose an am2 5600 for a new build, with the idea of upgrading to a quad core in time.

The only last instance where I would recommend a C2duo is only for something like constant encoding work where the cpu is not at idle much during the day, but....someone doing this should actually look at getting quad anyway. So I'm not sure there is a situation to recommend c2duo for now, and that's a change. They are good processors, but their advantages don't add up for most situations. Only overclockers now might prefer them on the specs.
May 4, 2007 3:06:40 PM

with intel price cuts amd has no chance, even though i dislike the big monopolistic manufacturers like nVidia and Intel, if amd\ati shows better products, then we will have a interesting summer.
May 4, 2007 3:08:36 PM

It makes me wonder what the upcoming split power planes and such will do for power consumption (for both AMD and Intel platforms). Nice to see that a CPU will be able to idle with most cores clocked down to barely anything and one hardly running at all to keep background processes churning, which will cause power consumption to be...nil. Now if only graphics cards....
May 4, 2007 3:11:40 PM

Typo on page 1. An E6600 costs around $220 and clocks at 2.4GHz. I think the author meant to say E6700.

When you buy the AMD6000 X2 you are buying an end-of-life product with minimal upgrade potential. Meanwhile the E6600 overclocks past 3GHz and there are stock parts that clock at 2.93GHz (as well as a quad core part)

Not mentioned in this article, but Intel CPUs are far more sensitive to memory performance than AMD. If you want to build an Intel system and performance is your goal, you should factor another $100+ for top-end memory.

The power question is mostly spurious; both chips run cool, neither requires special heatsink considerations. Building a solution with a 420W supply is easy for both CPU solutions - your decision on power supply is guided more by your choice in graphic cards. If you are a super-greenie you should be using a notebook.

If I had to buy a system today it would be Intel for longevity and overclocking. But if I could wait it would probably be Barcelona. The question is when? And is Barcelona really as good as AMD says it is? I certainly hope so for AMD's sake and our collective pockets!
May 4, 2007 3:19:02 PM

I am no fan of the X2 6000+, I believe there are better options from both AMD and Intel, but to say it has 'miminal' upgrade potential is absurd considering upcoming K10 CPUs will run on current AM2 boards.

C2D is nowhere near as sensitive to memory speed as AM2 is. You can use DDR2-533 on a C2D and get similar performance to DDR2-800. You'll lose 10 - 20% performance on AM2 in the same situation. I don't know where you got the idea from, but it's clearly wrong.
May 4, 2007 3:19:54 PM

Quote:
It makes me wonder what the upcoming split power planes and such will do for power consumption (for both AMD and Intel platforms). Nice to see that a CPU will be able to idle with most cores clocked down to barely anything and one hardly running at all to keep background processes churning, which will cause power consumption to be...nil. Now if only graphics cards....

Well, idle values will stay more or less the same but full load power consumption will have the upper hand on process refinement and die size shrink; we're (not that slowly) getting used to 100W+ CPUs.
May 4, 2007 3:22:23 PM

I am fascinated you think that AMD has any useful upgrade potential. Barcelona requires a new motherboard (as does Penryn) so I would evaluate the upgrade path based on which line of processors has the most headroom. This favors Intel as the Core 2 Duo have proven capable of remarkable overclock potential and the available speed stepping are already 20% above the E6600 that is compared in this article.

Like most people who have any interest in building their own system, I am interested in ripping media and playing games. I backup my DVDs to 4.7GByte DVD+R which requires transcoding. So, to be fair here, both the AMD top-of-range and Intel middle-of-range are good solutions. But I also upgrade my system from time-to-time so there is really no contest. Intel are today's winner.

If I want quad-core then I suggest waiting for Barcelona. But if you decide today then it's Core 2 Duo.
May 4, 2007 3:27:19 PM

Quote:
Typo on page 1. An E6600 costs around $220 and clocks at 2.4GHz. I think the author meant to say E6700.

When you buy the AMD6000 X2 you are buying an end-of-life product with minimal upgrade potential. Meanwhile the E6600 overclocks past 3GHz and there are stock parts that clock at 2.93GHz (as well as a quad core part)

Not mentioned in this article, but Intel CPUs are far more sensitive to memory performance than AMD. If you want to build an Intel system and performance is your goal, you should factor another $100+ for top-end memory.

The power question is mostly spurious; both chips run cool, neither requires special heatsink considerations. Building a solution with a 420W supply is easy for both CPU solutions - your decision on power supply is guided more by your choice in graphic cards. If you are a super-greenie you should be using a notebook.

If I had to buy a system today it would be Intel for longevity and overclocking. But if I could wait it would probably be Barcelona. The question is when? And is Barcelona really as good as AMD says it is? I certainly hope so for AMD's sake and our collective pockets!


It's a "no-brainer" to buy an AMD now because you drop-in a agena quad core upgrade into that same motherboard (and same window's licence $100 also) later on. Your "end of life" idea is "spurious" as you say. Also spurious to suggest (even indirectly) that an AMD system will fail sooner than an Intel system. In fact, overclocking shortens cpu life even with good cooling (not that this matters to enthusiasts!).
May 4, 2007 3:36:08 PM

Quote:

Not mentioned in this article, but Intel CPUs are far more sensitive to memory performance than AMD. If you want to build an Intel system and performance is your goal, you should factor another $100+ for top-end memory.


Ummm, my understanding is that the opposite is true.
May 4, 2007 3:45:38 PM

Quote:
It makes me wonder what the upcoming split power planes and such will do for power consumption (for both AMD and Intel platforms). Nice to see that a CPU will be able to idle with most cores clocked down to barely anything and one hardly running at all to keep background processes churning, which will cause power consumption to be...nil. Now if only graphics cards....

Well, idle values will stay more or less the same but full load power consumption will have the upper hand on process refinement and die size shrink; we're (not that slowly) getting used to 100W+ CPUs.

I think it will really work with quad-cores more than dual-cores. Two cores can be nearly inoperative until a highly threaded or demanding multitasking situation is happening. If even only two cores are really running at any given, average time, a quad-core CPU would be consuming in the ballpark of ~65w.

Quote:
I am fascinated you think that AMD has any useful upgrade potential. Barcelona requires a new motherboard (as does Penryn) so I would evaluate the upgrade path based on which line of processors has the most headroom.


K10 works in AM2 and 1207 except without HT3, split power planes, memory burst transfers, and other little things. The gist is, it will work.
May 4, 2007 3:51:01 PM

Quote:
I am fascinated you think that AMD has any useful upgrade potential. Barcelona requires a new motherboard (as does Penryn) so I would evaluate the upgrade path based on which line of processors has the most headroom. This favors Intel as the Core 2 Duo have proven capable of remarkable overclock potential and the available speed stepping are already 20% above the E6600 that is compared in this article.

Except for some features, AM2+ CPUs WILL work on all exicting AM2 boards and if penryn needs a new socket, then AMD wins this upgrade game.
May 4, 2007 3:54:45 PM

AMD are talking about an AM2+ board which suggests the final cut of Barcelona for the desktop (Agenda) is, like Penryn, going to require a modded motherboard. I wouldn't hold your breath on using Agenda as an upgrade path on your AM2 mobo. In any case, AMD are talking about moving to DDR3 in 2008 which, for a quad core, is very reasonable and I applaud AMD for taking the leap.

I did not imply a "shortening of life" and I'm curious how you could possibly imply this from my words. Is English a second language for you? I have both Intel E6600 systems and an AMD 6000X2 solution. The E6600 overclocks easily to 3.2GHz (a 25%) whereas I had difficulty getting the AMD6000X2 past 3.2GHz (a mere 6.6% overclock). Both systems using a Tuniq tower heatsink. I believe this is typical for both processors.

I noticed almost a 10% performance jump on my Core2Duo when upgrading from cheap PC6400 to top-end PC8500 sticks. This was a big suprise to me. The identical memory upgrade on the AMD6000X2 yielded a 2.5% performance jump (on the applications I favor). These figures are pre-overclock. In my opinion the AMD memory controller is less sensitive to memory than my P965 and P975 mobos.

Could you also explain your point on window's license to me? Microsoft state their license on Windows is based on the number of processors; not the number or cores. So your comment is meaningless as far as solutions from both Intel and AMD are concerned.
May 4, 2007 3:59:46 PM

Quote:
AMD are talking about an AM2+ board which suggests the final cut of Barcelona for the desktop (Agenda) is, like Penryn, going to require a modded motherboard. I wouldn't hold your breath on using Agenda as an upgrade path on your AM2 mobo. In any case, AMD are talking about moving to DDR3 in 2008 which, for a quad core, is very reasonable and I applaud AMD for taking the leap.

I did not imply a "shortening of life" and I'm curious how you could possibly imply this from my words. Is English a second language for you? I have both Intel E6600 systems and an AMD 6000X2 solution. The E6600 overclocks easily to 3.2GHz (a 25%) whereas I had difficulty getting the AMD6000X2 past 3.2GHz (a mere 6.6% overclock). Both systems using a Tuniq tower heatsink. I believe this is typical for both processors.

I noticed almost a 10% performance jump on my Core2Duo when upgrading from cheap PC6400 to top-end PC8500 sticks. This was a big suprise to me. The identical memory upgrade on the AMD6000X2 yielded a 2.5% performance jump (on the applications I favor). These figures are pre-overclock. In my opinion the AMD memory controller is less sensitive to memory than my P965 and P975 mobos.

Could you also explain your point on window's license to me? Microsoft state their license on Windows is based on the number of processors; not the number or cores. So your comment is meaningless as far as solutions from both Intel and AMD are concerned.


Any motherboard capable of supporting an X2 6000+ has the power capabilities to run a quad-core at stock. Overclocking it may be a different matter.

I don't know why you said it was impressive "for a quad core." It's an entire platform shift. Changing the memory controller isn't a big deal for AMD.

You did imply that AM2 had a short lifespan as of now. While that's true in one respect, as it will be replaced in the mid- to high-end by AM2+ shortly, AM2+ processors will still be compatible with AM2.

As to memory performance changes, I think you should brush up on the information surrounding AMD's memory division and speed reduction. A 6000+ at stock runs DDR2-800 at something like DDR2-750.
May 4, 2007 4:04:51 PM

To be honest I never considered DDR2 533Mhz memory in my assertion regarding memory performance. I only compared PC6400 against PC8500 so perhaps you are correct. I am astonished that anyone would dream of running DDR2 533 on either dual core processor. You might as well tie your shoe laces together then try running a marathon.
May 4, 2007 4:07:59 PM

Interesting article, as far as it goes. :roll: Why not use the Intel E6420? 8O

Why stop there? :oops:  Even mainstream users, I dare say many of those that read these articles anyway, will overclock. Take the equipment you tested, maximize your overclock, and run a few (or all) benchmarks again. Satisfy both user camps, "mainstream" non-overclockers AND "mainstream" overclockers! :D 
May 4, 2007 4:14:49 PM

Quote:
......
Could you also explain your point on window's license to me? Microsoft state their license on Windows is based on the number of processors; not the number or cores. So your comment is meaningless as far as solutions from both Intel and AMD are concerned.



It's not hard to find this kinda info. Just do a google search. I'm lazy at the moment, so here's the wiki:

http://en.wikinews.org/wiki/Microsoft_changes_OEM_licen...

You shouldn't be lazy though, in this forum. You'll find it has rather knowledgeable folks.
May 4, 2007 4:58:29 PM

Quote:
Even after the latest price cuts, the Core 2 Duo E6600 (2.66 GHz) is tagged at $316


Umm... the $316 price is the OLD price, the new one is $224 which invalidates this statement:

Quote:
This is 30% more expensive than what AMD asks for the Athlon 64 X2 6000+ top model (3.0 GHz), which is priced at $241.
May 4, 2007 5:06:23 PM

Can I guess that you did not intend to reply to me? You're post has you responding to me with quotes from someone else.
May 4, 2007 5:12:48 PM

Quote:
Even after the latest price cuts, the Core 2 Duo E6600 (2.66 GHz) is tagged at $316


Umm... the $316 price is the OLD price, the new one is $224 which invalidates this statement:

Quote:
This is 30% more expensive than what AMD asks for the Athlon 64 X2 6000+ top model (3.0 GHz), which is priced at $241.


You got that correct in spades! :) 

Intel Core 2 Duo E6600 Conroe 2.4GHz 4M shared L2 Cache LGA 775 Processor - Retail $227.00

http://www.newegg.com/Product/Product.aspx?Item=N82E168...
May 4, 2007 5:16:23 PM

Quote:
The new 60/90 min power consumption test is very interesting, because it clearly shows how the X2s, even consuming more power at full load, overall have a lower mixed power consumption.


What's even more interesting is how all the sites downplay this! Intel fanbois! Agh! It's a conspiracy by the power companies!!!

I took the liberty to do some simple calculations and I figure the "average" pc user will save about $25 a year just by running the AMD equivalent. Assuming electricity costs $0.20 per KWh and the machine is running 50% of the time. On a 24/7 system, you'd save $50 a year. Since the performance is equal, it makes no sense to NOT tack those extra costs onto the Intel CPU! I guess the electricity costs have to be > $100 a year (like back in the XP vs P4 days) before ppl begin to take it seriously.

Of course one could save even more money by running one of the EE versions like the 4600_65W that's selling for $120.

And we wonder how it is that we consume 22 million barrels of oil every day!
May 4, 2007 5:22:32 PM

Quote:
Can I guess that you did not intend to reply to me? You're post has you responding to me with quotes from someone else.


I just randomly hit the reply button that was closest to me.
May 4, 2007 5:28:51 PM

No big deal. When you are responding to a topic there is reply button at the bottom left corner of the page. When you want to reply to a specific comment there is a reply button at the top right corner of that post.
Hitting the right one avoids misunderstandings.
May 4, 2007 5:31:21 PM

Boy is that the pot turned black... :wink:
May 4, 2007 5:37:20 PM

Quote:

Not mentioned in this article, but Intel CPUs are far more sensitive to memory performance than AMD. If you want to build an Intel system and performance is your goal, you should factor another $100+ for top-end memory.


Ummm, my understanding is that the opposite is true.
yep... because of the on-chip (NB) memory controller and big cache.

for Intel, you can get RAM from DDR2-533 all the way up to DDR2-800, without seeing much performance gain.

on the other hand though, in order for AMD processor to function optimally, you have to get DDR2-800. you'll see a substantial hit in performance if you use lower grade DDR 2 like 667 or even 533.
May 4, 2007 5:39:52 PM

Quote:
Typo on page 1. An E6600 costs around $220 and clocks at 2.4GHz. I think the author meant to say E6700.

When you buy the AMD6000 X2 you are buying an end-of-life product with minimal upgrade potential. Meanwhile the E6600 overclocks past 3GHz and there are stock parts that clock at 2.93GHz (as well as a quad core part)

Not mentioned in this article, but Intel CPUs are far more sensitive to memory performance than AMD. If you want to build an Intel system and performance is your goal, you should factor another $100+ for top-end memory.

The power question is mostly spurious; both chips run cool, neither requires special heatsink considerations. Building a solution with a 420W supply is easy for both CPU solutions - your decision on power supply is guided more by your choice in graphic cards. If you are a super-greenie you should be using a notebook.

If I had to buy a system today it would be Intel for longevity and overclocking. But if I could wait it would probably be Barcelona. The question is when? And is Barcelona really as good as AMD says it is? I certainly hope so for AMD's sake and our collective pockets!


It's a "no-brainer" to buy an AMD now because you drop-in a agena quad core upgrade into that same motherboard (and same window's licence $100 also) later on. Your "end of life" idea is "spurious" as you say. Also spurious to suggest (even indirectly) that an AMD system will fail sooner than an Intel system. In fact, overclocking shortens cpu life even with good cooling (not that this matters to enthusiasts!).
for Intel, you can also buy quad core upgrades (not penryn though) without purchasing a whole new motherboard. :) 

i guess both Intel and AMD are going in the right direction :p 
May 4, 2007 5:41:47 PM

Quote:
No big deal. When you are responding to a topic there is reply button at the bottom left corner of the page. When you want to reply to a specific comment there is a reply button at the top right corner of that post.
Hitting the right one avoids misunderstandings.


I quote those that I reply to.
May 4, 2007 5:51:31 PM

Most likely because it was a lazy article........

1) He failed to account for the 6420. Its just dumb to buy the 6400 with the 6420 selling for the same price which is the norm.

2) He was "Speculating" about how the 65nm parts would OC. Unfortunately, if you read review sites, they appear to OC just fine.

3) OC tests "MUST" be included. This is an enthusiasts site. The overwhelming majority of folks who are reading this post and are going to build their own system either will OC or should OC. The few posters who start their post by claiming they will not OC wind up doing so after being explained to them how safe and easy it is to perform reasonable OCs that yield huge performance bonuses. Gone are the days when you might OC a chip 10-20%. Ushered in are the days some chips can do well over 50% of an OC and still not require anything but an OEM cooler and inexpensive DDR2-667 ram. (I.E, Take an 4300 from 1.8 to 2.7Ghz).

Its much like buy a sportscar and say we are not going to take it out of 1st gear but report how fast it can go.

I can't recall the last time EITHER of these chips was seriously recommend by anyone on the forums.

The AMD guys grab the x2 3600 or 3800 and crank it past the x2 5600.
The E6420 is not as bad of a choice but most folks would likely go with either the E4300 to save quite a bit of cash or just a few dollars more and grab the E6600.
a c 99 à CPUs
May 4, 2007 6:28:09 PM

Intel needs to allow for the lowest Speed Step multiplier to not simply be 6x regardless of FSB speed. Instead, Intel should aim for a bottom speed of 800-1000 MHz regardless of FSB speed. The current method uses 6x only as it's a simple rule that all chipsets can follow. Intel would have to make the bottom multiplier for a certain chip in the CPU's ROM somewhere and make the chipset read it. The VID-FID values change by what processor is running anyway, so this should be easy to accomplish.
May 4, 2007 6:32:44 PM

*sigh* Why are we still reviewing K8 versus Core 2 weeks away from a major release? A processor over 3 years old versus something thats been out since late July. We've seen all these benchmarks a hundred times over already.


Now this is news worthy...


Probably could have made an entire article about these two benchmarks alone.

Is it just me or does AMD actually look better on power consumption in these graphs. hmm.
May 4, 2007 6:42:56 PM

Quote:
The new 60/90 min power consumption test is very interesting, because it clearly shows how the X2s, even consuming more power at full load, overall have a lower mixed power consumption. This should be because Cool'n'Quiet is more effective than SpeedStep in the low power consumption modes; X2s can go as low as 800 MHz while Core2s are stuck at the double; 1600 MHz lowest.


I'd love to see tests in Vista to compare to...my cpu uses almost 10W more (according to Sandra anyway, which might just be lying to me) running Vista on average than in XP, and my higher average cpu temperature after the switch as well as improved cpu benchmark scores seem to agree as well. Anybody running old Socket 939 dual core cpu's noticed the huge difference that I have between XP and Vista??
May 4, 2007 6:58:08 PM

I find the lack of any info concerning exactly what the watts/hour test methodology (apart from looping sysmark) a bit confusing. How can those figures represent watts per hour?

watt= joule per second, lightbulb = 100W = 100 joules per second and in the bottom of the two graphs its like 144W per hour, which is obviously totally wrong. I presume they're averaging over the hour, but without a stated methodology the results seem meaningless, at least to me.

Also; is this total system draw, or just CPU?

An interesting test is here: http://www.pcstats.com/ArtVNL.cfm?articleid=2097&page=3
although with slightly different CPU's.

Comments?
May 4, 2007 7:13:44 PM

Why would it be totally wrong for the power consumption to go down after an hour? Sounds more likely to me that things like caching would compensate for the repetitive nature of the data and relieve the burden on the CPU and thus lower the power consumption. Especially in the Athlon case as the onboard memory controller would consume less power by anything the cache system can give the cpu instead of the onboard controller activating to fetch something from memory.

Perhaps the caching system is more tuned on the AMD system for repeated data? Kinda hard to tell for sure until Intel brings its own onboard controller to the table. Nothing in your link shows anything other than maximum and minimum consumption, there is no timed data in the link you posted so it isnt really helpful to prove or disprove your statement.
May 4, 2007 7:24:12 PM

Seems like your data follows along with the THG analysis.

Intel has the better short term consumption for spikes in performance except for the odd occurance with 3dmark06 but the AMD system wins out in Idle and in the long term tests.

Now the only thing to answer is what causes this?
May 4, 2007 7:31:36 PM

Quote:

An interesting test is here: http://www.pcstats.com/ArtVNL.cfm?articleid=2097&page=3
although with slightly different CPU's.

Comments?

Yes,
Very onesided test there which appears to be quite in favor for some reason of the core2....
the whole article prolly got them some Core2 hardware sponsored so I can understand the enthusiasm but why say first:
Quote:
he tone of the desktop computing world has palpably shifting towards power conservation, and even individual computer users are starting to wonder "just how much power is this going to take to get me there?"

Perhaps it's only when we consider the 24x7x365 electrical operating cost of a computer that any small power reduction musters any significance, but there it is.


and then in small letters:
Quote:
Please note that EIST is disabled for these tests, with EIST enabled, total system power draw will be even less that what is reported.

^^^ so EIST (and cool&quiet) which are just there for the 24/7/365 powerbills in the first place were not tested "for some reason"...

That and the weird choice of cpu's makes for a crappy objective/indepth powerdraw section IMHO, the Silentpcreview might not have lots of models of cpu's thrown in but the setups are identical and it goes very much indepth of what toms just found out.
May 4, 2007 7:35:29 PM

I'm not saying that it would be wrong for the average power consumption to drop over an hour usage, but that the graph makes it look like the total power draw over an hour is 144W which is a misnomer as a watt = joule per second and is a measure of power and not energy consumption.

They must be averaging over an hour, but a more useful figure would be watt hours or kWhours ie the total ENERGY draw for one hours operation, do you see what I mean?

this explains it pretty well:
http://www.gcse.com/energy/kWh.htm
a b à CPUs
May 4, 2007 7:38:49 PM

Quote:
*sigh* Why are we still reviewing K8 versus Core 2 weeks away from a major release? A processor over 3 years old versus something thats been out since late July. We've seen all these benchmarks a hundred times over already.


Now this is news worthy...


Probably could have made an entire article about these two benchmarks alone.

Is it just me or does AMD actually look better on power consumption in these graphs. hmm.


Yeah.. if you don't use your computer then the Athlon64 X2 is better. :p 

So X2 users are those who don't use their PC's, while C2D users use them.

That's what I gather from those power consumption tests. So X2 users, stop playing video games, encoding or doing all that other stuff and sit idle on THG arguing why the X2 is better while us C2D users enjoy playing games and paying a lower monthly electrical bill.
May 4, 2007 7:48:32 PM

is it just me or is THG still ignoring the existence of anything below an e6400? thats really annoying.
May 4, 2007 7:54:32 PM

The "use" vs "don't use" is not really valid since most of the time the PC is running it will likely be running far from full load.

The AMD's cycle far lower than Intel chips do and hence that is why they save more power on Idle.

Chalk one up for AMD.

As other's have stated, Intel should build in lower steppings.
Perhaps ones that kick in after longer periods of low use.

This is something Intel should still work on.
I don't think they have taken Idle use that seriously yet.
They have focused more on top power and staying cool under load.
May 4, 2007 8:01:10 PM

I believe the x2 5600+ is better than the E6400 for the guy who will not OC. No question here.

But I'm tired of reading reviews in THG with major flaws; in these case are prices and/or cpu models (1st page), measuring power requirements with EIST disabled, lack of the E6420 model (nitpicking here) and not including OC tests which are important for at least 80% of the readers.
May 4, 2007 8:01:37 PM

Quote:
Even after the latest price cuts, the Core 2 Duo E6600 (2.66 GHz) is tagged at $316


Umm... the $316 price is the OLD price, the new one is $224 which invalidates this statement:

Quote:
This is 30% more expensive than what AMD asks for the Athlon 64 X2 6000+ top model (3.0 GHz), which is priced at $241.


Yeah I wonder where the writer was looking at.
May 4, 2007 8:04:23 PM

Well, not all of us are overclocking fanboish nerds like you :p  j/k

I for example just use my pc for downloading *cough*legally owned content*cough* 75% of the time with it being on 24/7... with US$ 0.30/KWh rates for power here where 20% savings or more makes a BIG friggin difference as I pay the bills myself and not mommy or daddy.

Hence why I undervolt and go for silence like other people do.
!