Intel vs. AMD: General Faulse Assumptions for laptop CPUs

[IntelUser2000]

1. "Turion has worse battery life than Pentium M but Pentium M has advantage of low power platform".

False. Turion has memory controller on-die and memory controller part be taken out of the equation in power consumption.

2. "Pentium M systems have low power DDR2 memory while Turion uses power hungry DDR."

Not true. The power advantage that DDR2 has over DDR ONLY applies at same clock. DDR2-400 is lower power than DDR400. Its been said that DDR2 at same clock is 50% lower power consumption than DDR. With DDR2-533 however, DDR2-533 has only 33% reduction in power consumption against DDR400.

With DDR2-667 capable chipsets coming with Yonah, that advantage of power consumption goes down to 16.7%.

RAM doesn't take much of the power consumption part in laptops. They do contribute to power consumption, but its not a lot, maybe around 15%. Looking at the entire system, advantage of DDR2 over DDR decreases further in power consumption from 33% to approx. 5%.

855 chipsets support DDR, but it is HALF the power consumption of 915.

 

[dexvx]

Originally posted by: IntelUser2000
False. Turion has memory controller on-die and memory controller part be taken out of the equation in power consumption.

Umm... no. Memory controller is memory controller. Whether it is on-die or off-die, it still takes die space and thus power. It seems to me that the Pentium-M based laptops are better than Turion-ML laptops in terms of battery life on similarily configured laptops.

Originally posted by: IntelUser2000
Not true. The power advantage that DDR2 has over DDR ONLY applies at same clock. DDR2-400 is lower power than DDR400. Its been said that DDR2 at same clock is 50% lower power consumption than DDR. With DDR2-533 however, DDR2-533 has only 33% reduction in power consumption against DDR400.

With DDR2-667 capable chipsets coming with Yonah, that advantage of power consumption goes down to 16.7%.

Memory clock speed and power consumption are, to my knowledge, not a 1:1 correlation in percentage.

 

[Hacp]

Um the Pentium M does have the advantage of the lower power platform. The Pentium M has better battery life and battery saving techniques than the Turion. Plus, it was built as a laptop chip, so it was designed to use as less power as possible w/o sacraficing performance. The Pentium M has the better platform because it is usually combined with the Intel Centrino Platform. This platform emphasizes low power graphics/ low power chipset/ and low power wireless. These combined with the advanced power savings of the Pentium M give the Centrino platform an extremely long battery life. DDR2 just helps it even more.

The Turion on the other hand is an undervolted Athlon 64. It does not have the advanced power saving features of the Pentium M. I has cool and quiet, but thats all. THe chipset, graphics, and wirelss of most turion notebooks take up way more power than those of the centrino platform. Already, the Turion uses more power than the Pentium M in everyday applications. Combine that with a more power hungry platform and the Centrino Platform is a winner in terms of battery life.

Ow ya, both pentium M and turion have northbridges, so I don't know why your yelling about the memory controller. Traditionally, the Memory controller is placed on the northbridge.

 

[IntelUser2000]

You two may not be the ones I pointed out. But people pointed out that AMD has advantage of IMC and Intel has advantage of low power platform and DDR2. IMC DOES consume less power since CPUs are on smaller process tech than chipsets.

Memory clock speed and power consumption are, to my knowledge, not a 1:1 correlation in percentage.

That even reinforces my argument saying that DDR2 has less advantage over DDR in power consumption. Though using CMOS devices, the power consumption increase over increase in clock speed is very likely linear, just like the CPUs.

About people saying how Centrino has low power chipset and wireless and optimization, it really has no point, since you people are talking about the "potential" of what Turion can be.

I sometimes believe these discussions in forums are purely theoretical, cause I can say: "What if Prescott did reach 5GHz early this year, and Tejas came with 7GHz with better performance per clock?"

1. What if Turion had better chipset??
2. What if A64s clocked like P4s?
3. (old one) What if AXPs clocked like P4s?
4. What if PM had IMC??

In reality, none are true, and having looked at a review(which is hard to find) that says Intel's chipset has power consumption advantages over EVERY other chipset, I can say the low power chipset factor isn't gonna change.

Plus, even with PM, situation can improve further just like Turion, as 915M chipsets consume nearly DOUBLE the power of 855M, while with next generation, it will have lower power than 855M.

 

[IntelUser2000]

1. There are no real advantage of DDR2 over DDR in power consumption.
2. Indeed DDR2 is faster than DDR since DDR2 is clocked way higher, negating #1 somewhat.
3. Centrino is power optimized and Turion is not as optimized. Who cares?? The fact is that there is no Turion or equivalent that is as optimized. Saying that it COULD be optimized is PURE theory.

 

[fbrdphreak]

*sigh*

*shakes head*

Originally posted by: IntelUser2000
1. "Turion has worse battery life than Pentium M but Pentium M has advantage of low power platform".

False. Turion has memory controller on-die and memory controller part be taken out of the equation in power consumption. Is this supposed to be news? Turion MT chips have lower consumption when you DON'T take the memory controller out of the equation. Intel measured thermal design power as 75% of the maximum power consumption. Since TDP is technically used as a measurement to design heat dissipation systems, Intel figured why quote higher figures if most of the time you won't see 100% power dissipation. AMD on the other hand measures TDP as 100% maximum, worst case power consumption. If a Turion 64 consumes 25W @ 100%, multiply that *.75=18.75W. Or you can add 25% to Intel's 27W figure, *1.25=33.25W. Now, you say "Turion has worse/better battery life," this is very different than chip power consumption.

2. "Pentium M systems have low power DDR2 memory while Turion uses power hungry DDR."

Not true. The power advantage that DDR2 has over DDR ONLY applies at same clock. DDR2-400 is lower power than DDR400. Its been said that DDR2 at same clock is 50% lower power consumption than DDR. With DDR2-533 however, DDR2-533 has only 33% reduction in power consumption against DDR400. What you're assuming is that while DDR2 uses 1.8V & DDR 2.6V is that current is constant between the two. I = V/R Without knowing how much current the chips draw at a given state you can't make that assumption. Where is it "said that DDR2 at same clock is 50% lower?" I would like to see that.

With DDR2-667 capable chipsets coming with Yonah, that advantage of power consumption goes down to 16.7%.

RAM doesn't take much of the power consumption part in laptops. They do contribute to power consumption, but its not a lot, maybe around 15%. Looking at the entire system, advantage of DDR2 over DDR decreases further in power consumption from 33% to approx. 5%.

855 chipsets support DDR, but it is HALF the power consumption of 915.

Are you trying to start flamewars?

 

[fbrdphreak]

Originally posted by: IntelUser2000
1. There are no real advantage of DDR2 over DDR in power consumption.
2. Indeed DDR2 is faster than DDR since DDR2 is clocked way higher, negating #1 somewhat.
3. Centrino is power optimized and Turion is not as optimized. Who cares?? The fact is that there is no Turion or equivalent that is as optimized. Saying that it COULD be optimized is PURE theory.

All you're saying is pure theory. So you might have read a couple whitepapers.

What tests have you conducted? What degree do you have? You talk a lot of ****** without having anything to back it up there son.

 

[Accord99]

Originally posted by: fbrdphreak
False. Turion has memory controller on-die and memory controller part be taken out of the equation in power consumption. Is this supposed to be news? Turion MT chips have lower consumption when you DON'T take the memory controller out of the equation. Intel measured thermal design power as 75% of the maximum power consumption. Since TDP is technically used as a measurement to design heat dissipation systems, Intel figured why quote higher figures if most of the time you won't see 100% power dissipation. AMD on the other hand measures TDP as 100% maximum, worst case power consumption. If a Turion 64 consumes 25W @ 100%, multiply that *.75=18.75W. Or you can add 25% to Intel's 27W figure, *1.25=33.25W. Now, you say "Turion has worse/better battery life," this is very different than chip power consumption.

Except it's not true at all. 75% only refers to the Williamette P4. Not only that, all load measurements of Dothan have shown that it does not even come close to 27W even with a thermal virus or Prime 95 even at 2.26GHz.

34W is in fact enough to power an entire P-M 2.26 laptop with a 14" screen and integrated graphics under full CPU load.

 

[fbrdphreak]

Originally posted by: Accord99

Originally posted by: fbrdphreak
False. Turion has memory controller on-die and memory controller part be taken out of the equation in power consumption. Is this supposed to be news? Turion MT chips have lower consumption when you DON'T take the memory controller out of the equation. Intel measured thermal design power as 75% of the maximum power consumption. Since TDP is technically used as a measurement to design heat dissipation systems, Intel figured why quote higher figures if most of the time you won't see 100% power dissipation. AMD on the other hand measures TDP as 100% maximum, worst case power consumption. If a Turion 64 consumes 25W @ 100%, multiply that *.75=18.75W. Or you can add 25% to Intel's 27W figure, *1.25=33.25W. Now, you say "Turion has worse/better battery life," this is very different than chip power consumption.

Except it's not true at all. 75% only refers to the Williamette P4. Not only that, all load measurements of Dothan have shown that it does not even come close to 27W even with a thermal virus or Prime 95 even at 2.26GHz.

34W is in fact enough to power an entire P-M 2.26 laptop with a 14" screen and integrated graphics under full CPU load.

That sounds plausible, but without information to back it up it is just as worthless as the OP's claims. Numerous tech sites (AT, THG, HH, etc) have cited that Intel still measures TDP at 75% max current draw. Where is your source?

 

[bearxor]

I thought bout jumping in on this fbrdphreak, but thought you would just drop a link to the LL comparison. That would end this conversation pretty quick.

 

[Accord99]

Originally posted by: fbrdphreak

That sounds plausible, but without information to back it up it is just as worthless as the OP's claims. Numerous tech sites (AT, THG, HH, etc) have cited that Intel still measures TDP at 75% max current draw. Where is your source?[/quote]
Intel provides maximum power draws for their Xeon line.

For example the 130nm Xeon 3.06 has a TDP of 85W and a maximum power of 101W. The 90nm Xeon 3.8 is 110W TDP and 120W maximum.

http://users.erols.com/chare/elec.htm

And the P-M is a completely different architecture so its relationship its completely different. Measurements would indicate that the TDP is far overrated compared to the actual maximum power draw of the P-M.

 

[Accord99]

Originally posted by: bearxor
I thought bout jumping in on this fbrdphreak, but thought you would just drop a link to the LL comparison. That would end this conversation pretty quick.

Except there are other reviews that show that the P-M uses half the power of a comparbly clocked Turion ML.

Like here:
http://babelfish.altavista.com/babelfis...on-64-Pentium-M-Asus-A6000-327%2f11%2f
where when playing Doom 3, the P-M uses 30W less power. The two laptops are more equal than the Laptoplogic review.

 

[fbrdphreak]

Originally posted by: Accord99

Originally posted by: fbrdphreak

That sounds plausible, but without information to back it up it is just as worthless as the OP's claims. Numerous tech sites (AT, THG, HH, etc) have cited that Intel still measures TDP at 75% max current draw. Where is your source?

Intel provides maximum power draws for their Xeon line.

For example the 130nm Xeon 3.06 has a TDP of 85W and a maximum power of 101W. The 90nm Xeon 3.8 is 110W TDP and 120W maximum.

http://users.erols.com/chare/elec.htm

And the P-M is a completely different architecture so its relationship its completely different. Measurements would indicate that the TDP is far overrated compared to the actual maximum power draw of the P-M.[/quote]Wow, so guess how pissed I was when I typed up a bunch of stuff and accidentally closed the window, LOL. Let me see if I can sum this up:

*You tend to talk a lot about stuff you can't prove or support well, if indeed what you've said isn't completely off base as it is. Use some logic, please.
*TDP, max power draw, and max power usage during normal operation are all very different things. Every expert source I've seen and talked to says that Intel measures TDP as 75% of max power consumption and AMD measures as 100% worst case. Prove me wrong and I'll listen to you.
*How is that review you linked any more matched than ours? Besides not being able to understand a damn thing in it, how is it better? The only way ours could have been better matched is if we had two same branded laptops with the ATI Radeon Xpress200M chipset; one Intel and one AMD of course.
*In regards to Doom 3 power consumption, assuming one can trust a foreign review that we have no reason to put any faith in, I am not surprised the Pentium M would come out a little on top. Dothan can disabled registers, buffers, parts of the cache, etc on the fly to save power; even when under heavy load like Doom 3. Turion does not do that to my knowledge; Turion is simply a very efficient processor made with low power transistors.

I think that's it but I'm sure I'll remember more to mention here in a bit.

 

[frostedflakes]

Directly from a Pentium-M 2MB/533FSB datasheet:

NOTES:
1. The TDP specification should be used to design the processor thermal solution. The TDP is not the maximum theoretical power the processor can dissipate.

 

[fbrdphreak]

Originally posted by: frostedflakes
Directly from a Pentium-M 2MB/533FSB datasheet:

NOTES:
1. The TDP specification should be used to design the processor thermal solution. The TDP is not the maximum theoretical power the processor can dissipate.

Yep.

 

[dexvx]

Originally posted by: fbrdphreak
*TDP, max power draw, and max power usage during normal operation are all very different things. Every expert source I've seen and talked to says that Intel measures TDP as 75% of max power consumption and AMD measures as 100% worst case. Prove me wrong and I'll listen to you.

What expert source(s) are you talking about? The guy at LL who runs benchmark software on new toys does not count as an expert. In fact, his review has more than 1 inconsistency due to a lack of either research or HW knowledge.

Originally posted by: fbrdphreak

Originally posted by: frostedflakes
Directly from a Pentium-M 2MB/533FSB datasheet:

NOTES:
1. The TDP specification should be used to design the processor thermal solution. The TDP is not the maximum theoretical power the processor can dissipate.

Yep.

LoL... you forgot to include the previous sentence:

The processor?s power is specified as Thermal Design Power (TDP) for thermal solution design. TDP is defined as the worst-case power dissipated by the processor while executing publicly available software under normal operating conditions, at nominal voltages that meet the load line specifications. The TDP definition is synonymous with the Thermal Design Power (typical) specification referred to in previous Intel datasheets. The Intel TDP specification is a recommended design point and is not representative of the absolute maximum power the processor may dissipate under worst case conditions.

Of course its not the maximum theoretical power dissipated. A Pentium-M 1.6Ghz will dissipate LESS power at "maximum" than a 2.26Ghz Pentium-M, even though they have the same TDP. On the other end of the spectrum, software overvolting, software PLL clock modifications, and even thermal viruses under load will exceed the TDP. In layman's terms, Intel's TDP is the the worst case power dissipated at stock speeds with stock voltage with stock current with software that does not modify hardware settings in normal operating evironment (Which is... 5C - 40C?). Example to your contrary: if I were to run a Turion/Pentium-M in an oven at 100C, they would exceed their TDP very quickly (yes even AMD's "maximum TDP").

HardOCP.

 

[fbrdphreak]

Thanks for that info, I hadn't read that part of the Intel whitepapers in a while myself. As far as who said Intel's TDP is 75% max, I know AT & THG said that at one point. I can dig up the articles if you really care, but I should get back to studying soon 😉

From a whitepaper on A64 power dissipation, here is AMD's definition of TDP:

Thermal Design Power (TDP) is measured under the conditions of TCASE Max, IDD Max, and VDD=VID_VDD, and include all power dissipated on-die from VDD, VDDIO, VLDT, VTT, and VDDA.

It certainly seems to me from that statement that AMD measures TDP as a dead sure worst case scenario. It makes sense, as AMD stock heatsinks always work rather well; they're designed for much worse than average operating conditions!

Anyway, the conclusion I'm seeing here is that Turion MT's dissipate 25W dead maximum and a 533 Dothan will probably never dissipate more than 27W (assuming 2.2 & 2.26GHz models respectively). Chip to chip, my money is on Turion for lower power consumption. Is it enough to make a huge difference? Probly not.

But the fact still remains that when these chips are put into a notebook system, the Pentium M is overall more efficient under normal operating conditions. Chipset optimizations alone account for a decent power savings and with Dothan's ability to deactivate parts of the cache, buffers, etc to save power, it isn't much of a contest. However I still believe overall Turion is a better value.

Chip to chip, Turion has Dothan beat hands down:
MT-37 2GHz $223: http://labs.anandtech.com/alllinks.php?pfilter=3865
760 2GHz: $329: http://labs.anandtech.com/alllinks.php?pfilter=1450

On the other hand, Turion's are available in far fewer laptop configurations than Pentium M's are. When you look at two similar Dothan & Turion systems, you will generally see Turion winning in price by a very decent margin; just look at HP's machines.

The guy at LL who runs benchmark software on new toys does not count as an expert. In fact, his review has more than 1 inconsistency due to a lack of either research or HW knowledge.

And which inconsistency is that?

EDIT Thought I would also add this: if Intel's TDP specs are so close to real world usage, why have a large number of consumers had their toasterbox P4 3.6-3.8's throttle on them because the Intel HSF can't keep it cool enough? I'm not trying to bring P4's into the discussion, I'm merely using this as a point that I see no reason that Intel's actual power consumption is as low as they claim. Intel can say "worst case power dissipated at stock speeds with stock voltage with stock current with software that does not modify hardware settings in normal operating evironment" all they want, several review sites around the web and own user reports I mentioned above all lend credence to the fact that Intel's power dissipation isn't as low as they'd like everyone to think.

 

[Accord99]

Originally posted by: fbrdphreak
Anyway, the conclusion I'm seeing here is that Turion MT's dissipate 25W dead maximum and a 533 Dothan will probably never dissipate more than 27W (assuming 2.2 & 2.26GHz models respectively). Chip to chip, my money is on Turion for lower power consumption. Is it enough to make a huge difference? Probly not.

http://babelfish.altavista.com/babelfis.../le-turion-dans-les-desktops/page4.php
The difference between idle and full load is 17w for a 2.13GHz P-M and 26W for a 2.2GHz Turion MT. And this is with the MT using 1.2v.

several review sites around the web and own user reports I mentioned above all lend credence to the fact that Intel's power dissipation isn't as low as they'd like everyone to think.

The P4 is a different matter. However, measurements of the P-M would indicate that it is significantly lower than the TDP. 17w for the above review with cpu-burn,

16w for the 2GHz P-M 755 with P95: http://www.x86-secret.com/pics/divers/conso/conso_result.png

27w for the 2.26GHz P-M which includes the power used at the MB voltage converter and a CT-479 adapter while running the most heat producing power virus available:
http://www.xbitlabs.com/articles/cpu/display/pentiumm-780_9.html

30w advantage over a Turion ML with the same laptop chassis, screen, video card and the exact same battery while playing Doom 3. The P-M laptop itself is using only a total 44.4W so it has a 67% advantage in battery life. The chipset and memory difference only accounts for a fraction of it, the rest comes from the CPU.
http://babelfish.altavista.com/babelfis...on-64-Pentium-M-Asus-A6000-327%2f11%2f

 

[dexvx]

Originally posted by: fbrdphreak
From a whitepaper on A64 power dissipation, here is AMD's definition of TDP:

Thermal Design Power (TDP) is measured under the conditions of TCASE Max, IDD Max, and VDD=VID_VDD, and include all power dissipated on-die from VDD, VDDIO, VLDT, VTT, and VDDA.

Basically what Intel's TDP definition is, except using words instead of symbols that would appear obscure to most people.

Originally posted by: fbrdphreak
However I still believe overall Turion is a better value.

Chip to chip, Turion has Dothan beat hands down:
MT-37 2GHz $223: http://labs.anandtech.com/alllinks.php?pfilter=3865
760 2GHz: $329: http://labs.anandtech.com/alllinks.php?pfilter=1450

It'd be a better value if you purchased retail. However, in the real world, most people buy notebooks from big OEMs, ala Dell, HP, IBM. So far, Dell has been king in the hot deals notebook on a regular basis (once in a year deals like BF hardly count).

Originally posted by: fbrdphreak

Originally posted by: dexvx
The guy at LL who runs benchmark software on new toys does not count as an expert. In fact, his review has more than 1 inconsistency due to a lack of either research or HW knowledge.

And which inconsistency is that?

1. He confuses TDP with power consumption. Those are totally different measurements.
2. He claims that raising the FSB to 533 and adding XD increased TDP by 30% in Dothans (B to C stepping). He seems to suggest that at clock B0 Dothan offers 30% savings in power compared to C0 Dothans, which is not true. Intel's docs for the Celeron-M is at 21W TDP, regardless of stepping, so his idea that the transition from B0 to C0 stepping causing a 30% jump in power dissipation is just stupid.
3. His erraneous "75% law" for TDP extention to Dothans.
4. His erraneous idea that AMD's TDP is at "maximum" current draw, which is also inaccurate. Current at load and current at max are different.

I'm going to stop there.

 

[fbrdphreak]

Once again, I lend no credence to a review in a foreign language that I do not know to be a reliable site.

Now the X-Bit Labs & X-86 articles provide some interesting numbers, numbers that we would expect. Thank you for those.

But those articles still do not compare Turion to Dothan.

 

[dexvx]

Originally posted by: Accord99
27w for the 2.26GHz P-M which includes the power used at the MB voltage converter and a CT-479 adapter while running the most heat producing power virus available:
http://www.xbitlabs.com/articles/cpu/display/pentiumm-780_9.html

Oh wow... thats actually the first power consumption measurement of a Dothan CPU alone that I've seen. Is there one for the Turion?

Originally posted by: fbrdphreak
Once again, I lend no credence to a review in a foreign language that I do not know to be a reliable site.

It may come as a shock to you, but people in France, Germany, Japan, China... you know, the "other half" of the world, also use computers.

 

[fbrdphreak]

Originally posted by: dexvx
1. He confuses TDP with power consumption. Those are totally different measurements.
2. He claims that raising the FSB to 533 and adding XD increased TDP by 30% in Dothans (B to C stepping). He seems to suggest that at clock B0 Dothan offers 30% savings in power compared to C0 Dothans, which is not true. Intel's docs for the Celeron-M is at 21W TDP, regardless of stepping, so his idea that the transition from B0 to C0 stepping causing a 30% jump in power dissipation is just stupid.
3. His erraneous "75% law" for TDP extention to Dothans.
4. His erraneous idea that AMD's TDP is at "maximum" current draw, which is also inaccurate. Current at load and current at max are different.

I'm going to stop there.

1. Maybe you didn't read the article closely enough or he didn't specify clearly enough, but he does know TDP and power consumption are not the same thing.
2. The only "claim" he makes is that TDP went up to 27 from 21W. That's a 28.57% increase; I'm sorry, he rounded.

With the release of the Sonoma platform, Dothan got yet another upgrade. The memory bandwidth bottleneck was alleviated with the support of a new quad-pumped 133MHz FSB, as well as the adoption of DDR2 PC4200 (533MHz) SO-DIMM modules. Intel also followed AMD's footsteps by including support for the XD (Execute Disable) bit, identical to AMD's NX bit, which prevents buffer overflows thus providing a form of virus protection via hardware. Currently, 533MHz Dothan CPUs run at up to 2.26GHz. Unfortunately, these features came at a price: Dothan suffered almost a 30% increase in TDP, going from 21W to 27W TDP. However, due to its high-tech design Dothan still offers incredible battery life while increasing performance.

3. 75% is what every other reputable review site uses, as I previously mentioned. We'll be sure to discuss that with Intel soon.
4. Once again, AMD's TDP is at max current draw. Link Once again:

Thermal Design Power (TDP) is measured under the conditions of TCASE Max, IDD Max, and VDD=VID_VDD, and include all power dissipated on-die from VDD, VDDIO, VLDT, VTT, and VDDA.

IDD MAX, otherwise known as the maximum drain current. I'd like to hear a real reason why you think AMD's specs aren't for worst case scenario, 'cuz you don't seem to have one.

 

[fbrdphreak]

Originally posted by: dexvx

Originally posted by: fbrdphreak
Once again, I lend no credence to a review in a foreign language that I do not know to be a reliable site.

It may come as a shock to you, but people in France, Germany, Japan, China... you know, the "other half" of the world, also use computers.

No ****** sherlock, but how do I or you know how trustworthy their testing methods are? They could be another THG for all you know, where half the articles are total bullshit.

 

[Accord99]

Originally posted by: dexvx

Originally posted by: Accord99
27w for the 2.26GHz P-M which includes the power used at the MB voltage converter and a CT-479 adapter while running the most heat producing power virus available:
http://www.xbitlabs.com/articles/cpu/display/pentiumm-780_9.html

Oh wow... thats actually the first power consumption measurement of a Dothan CPU alone that I've seen. Is there one for the Turion?

Unfortunately no. The closest ones are the two French sites, where since they are using cpuburn, if you take the difference between the idle and full load power usage, you should be able to get a good approximation because cpuburn doesn't use much more memory or the video card or HD, etc.

 

[dexvx]

Originally posted by: fbrdphreak

Originally posted by: dexvx

Originally posted by: fbrdphreak
Once again, I lend no credence to a review in a foreign language that I do not know to be a reliable site.

It may come as a shock to you, but people in France, Germany, Japan, China... you know, the "other half" of the world, also use computers.

No ****** sherlock, but how do I or you know how trustworthy their testing methods are? They could be another THG for all you know, where half the articles are total bullshit.

No, I just find it interesting that you take LL to be the spoken word of the hardware god (you've quoted them a few times), and take precedence to using any foreign language review site. BTW, Matbe and x86-secret are very reputable sites in France.

Edit: XbitLab's power measurements ALL skew on the high side.