NotebookForums.com › Forums › General Notebook Discussions › Notebook Forums - General › Extremely good Turion / P-M technical comparison
New Posts  All Forums:Forum Nav:

Extremely good Turion / P-M technical comparison - Page 2

post #21 of 49
Quote:
Originally Posted by Karma
Blink, I believe you did not even read it, as it answers your questions. As a matter of fact, I went over the TDP in my very first post. There is a very good technical reason for it, and if you are not aware of it, then you miss the entire point of not just the article, but of this thread. Read both articles, then please return.

I hate to be so lazy, but i skimmed both articles, and it (seemingly) did not

make any mention of these numbers

total power consumption of

Turion ML + ATI RX480M + ATI SB400
Turion MT + ATI RX480M + ATI SB400
PM Dothan C0 + Intel 915 PM + ICH6-M

if it is in the article, would you mind linking me, I just took a second look and did not find it.

Also the second point still stands.
post #22 of 49
Thread Starter 
Short story: chipsets make no difference in battery tests, especially since both the Intel and AMD chipsets in this case both have north-bridge and south-bridge chips (even the ATi/AMD chipset has a northbridge, regardless of the integrated memory controller).

Previous comparisons between the P-M and the Turion were between the first version of the Turion and the Dothan: 21W Dothan vs the 35W Turion ML (the one most readily available at the time). A few even had the Dothan against a 25W MT, but the RAM and/or CPU speeds were too different to make a valid comparison between them: i.e. 2.0 Ghz Dothan vs 1.6 Ghz MT Turion, or the 533-DDR2 vs the 333-DDR of the Turion, even though the Turion supports the DDR-400.

Harald Thon and Bert Töpelt both received extremely good access to AMD's Turion due to the fact that AMD has their new 90nm Fab 36 LLC manufacturing facility near where Toms Hardware is based: Dresden, Germany. This article even covers information that debunks what alot of people here previously thought: that higher rated Turions have lower TDP's at the same speed as lower rated Turions. This disproves one thing I was unaware of: that if a person buys a higher performing Turion notebook, it will run quieter and cooler than a lower performing notebook.

As to your question for them to use RMclock, you can always email them to ask them why they didn't include it in their test bed.
post #23 of 49
I notice an excess amount of fanboys on this site I mean why be brand conscious? It's to the point that some people are seemingly blinded by the facts laid right before them. I am not pointing to anyone in particular but what happened to sound judgement? I believe that the report favours neither of the products. If i read correctly it basically says you can't lose with either. Both have pros and cons...depending on personal usage then an educated choice can be made.
post #24 of 49
Thread Starter 
Quote:
Originally Posted by pygo
I notice an excess amount of fanboys on this site I mean why be brand conscious? It's to the point that some people are seemingly blinded by the facts laid right before them. I am not pointing to anyone in particular but what happened to sound judgement? I believe that the report favours neither of the products. If i read correctly it basically says you can't lose with either. Both have pros and cons...depending on personal usage then an educated choice can be made.
That was the conclusion of the multi-part article as well.
post #25 of 49
hold off.
post #26 of 49
Amd turion is rellay very similar to pentium m just about the same speed and its cheaper...ummmm..maybe i hsould have went with turion but the fans!!!!!!!!!!!!
post #27 of 49
Quote:
Originally Posted by Karma
Short story: chipsets make no difference in battery tests, especially since both the Intel and AMD chipsets in this case both have a north-bridge and a south-bridge (even the ATi/AMD chipset has a northbridge, regardless of the integrated memory controller).

Previous comparisons between the P-M and the Turion were between the first version of the Turion and the Dothan: 21W Dothan vs the 35W Turion ML (the one most readily available at the time). A few even had the Dothan against a 25W MT, but the RAM and/or CPU speeds were too different to make a valid comparison between them: i.e. 2.0 Ghz Dothan vs 1.6 Ghz MT Turion, or the 533-DDR2 vs the 333-DDR of the Turion, even though the Turion supports the DDR-400.

Harald Thon and Bert Töpelt both received extremely good access to AMD's Turion due to the fact that AMD has their new Fab 36 LLC manufacturing facility near where Toms Hardware is based: Dresden, Germany. This article even covers information that debunks what alot of people here previously thought: that higher rated Turions have lower TDP's at the same speed as lower rated Turions. This disproves one thing I was unaware of: that if a person buys a higher performing Turion notebook, it will run quieter and cooler than a lower performing notebook.

As to your question for them to use RMclock, you can always email them to ask them why they didn't include it in their test bed.

i just read both articles and nowhere do they make _any_ talk about how the chipset effect power, how do you know that they have no effect, i think they do, and unless you can link me to an article that says otherwise i will hold true to it.

and thanks whoever modded me down 315 points, i was right, the article did not talk about what i asked, and because of your fanboyness you ****tard i am in the red for the first time since i became part of this forum, thanks jackass.
post #28 of 49
Thread Starter 
Quote:
Originally Posted by BlinKSilver
i just read both articles and nowhere do they make _any_ talk about how the chipset effect power, how do you know that they have no effect, i think they do, and unless you can link me to an article that says otherwise i will hold true to it.

and thanks whoever modded me down 315 points, i was right, the article did not talk about what i asked, and because of your fanboyness you ****tard i am in the red for the first time since i became part of this forum, thanks jackass.
That was me, Blink. I did not mean to say the article mentioning chipsets effecting power, instead I was talking about the articles mentioned specifically reasons why they chose the low-power MT vs the Dothan P-M.

As to why I would know about chipsets effecting power consumption, well, for that, read what's under my name.

Honestly I tried to give a small rap on the wrist, but I couldn't avoid the full hit. If you would like me to bump you back up again, please ask and I will give it back to you.
post #29 of 49
Quote:
Originally Posted by Karma
That was me, Blink. I did not mean to say the article mentioned chipsets effecting power, I was talking about the articles mentioned specifically reasons why they chose the low-power MT vs the Dothan P-M.

As to why I would know about chipsets effecting power consumption, well, for that, read what's under my name.

Honestly I tried to give a small rap on the wrist, but I couldn't avoid the full hit. If you would like me to bump you back up again, please ask and it will be as if nothing happened.

sure, first off, i am no amd fan, you can tell my latest rig.

all I am saying and have been saying is this

22 27 35

by nature you should choose to compare 22 & 27, but that the 22 and 35 have integrated memory controllers, and in case no one here knew, one of the most power consuming parts of a computer (second to the CPU and GPU) is the memory control, just look at most desktop, they have cooling systems (hence why laptops have special chipsets) for the northbridge, i think that is something that should be accounted for, as well as the whole chipset itself. intel bumped up the power consumption of the new sonoma chipset, i just think it would be fair to know, that is all a ever asked for. because when we are talking about numbers this close, every little thing counts.

and yes, reset the rep,
post #30 of 49
Thread Starter 
Quote:
Originally Posted by BlinKSilver
sure, first off, i am no amd fan, you can tell my latest rig.

all I am saying and have been saying is this

22 27 35

by nature you should choose to compare 22 & 27, but that the 22 and 35 have integrated memory controllers, and in case no one here knew, one of the most power consuming parts of a computer (second to the CPU and GPU) is the memory control, just look at most desktop, they have cooling systems (hence why laptops have special chipsets) for the northbridge, i think that is something that should be accounted for, as well as the whole chipset itself. intel bumped up the power consumption of the new sonoma chipset, i just think it would be fair to know, that is all a ever asked for. because when we are talking about numbers this close, every little thing counts.

and yes, reset the rep,
I'm not sure what you mean by saying the 22W has an integrated memory controller. All AMD chips now have integrated memory controllers and no Intel chips have it. The Turion CPU used was a MT-34, with a 25W TDP, while the P-M used was a 750, which has a 22W TDP. Honestly, almost all desktop motherboards have cooling solutions on their north and south bridge chips. It doesnt mean much, if only to add stability to those who want to overclock/loadup their computers. The notebook motherboards only have active cooling (due to the thermal releases) on the CPU, GPU, and sometimes hard drive.

The Intel 915 PM can cause a notebook to use more power than the 855PM, but not because the northbridge is a bad design, but that the PCIe bus on the 915 draws more power than the AGP on the 855PM chipset. In this case, both the 915PM and the ATI RX480M are running north and south bridge chips, with the same hard drive spec (5400 rpm IDE), and PCIe power draw as well. Both notebooks have the same number of So-Dimms, as close to the same CPU speed as possible, and the same RAM speed as well. The display is also exactly the same between them, and both were manufactured by the same plant (MSI).

Compared to the CPU, RAM, display, and hard drive, the chipset is at the very bottom of the list of power draw. Why do you think the most powerful desktop SLI motherboard costs $200.00 while the most powerful CPU costs more than $1,000? Because most of it is PCB, and routing, thats all. The north and south bridges are minor, and are differentiated through their performance and not their power draw (ATi's SB400, 450, and upcoming 600 with the USB/CPU issues, for example).

In either case, you got your rep back, plus another 325 for good measure. However, in the future, read first then ask questions later, and avoid obscenities and personal attacks on the forums, or you will be in the red again.
post #31 of 49
Quote:
Originally Posted by Karma
I'm not sure what you mean by saying the 22W has an integrated memory controller. All AMD chips now have integrated memory controllers and no Intel chips have it. Honestly, almost all desktop motherboards have cooling solutions on their north and south bridge chips. It doesnt mean much, if only to add stability to those who want to overclock/loadup their computers. The notebook motherboards only have active cooling (due to the thermal releases) on the CPU, GPU, and sometimes hard drive.

The Intel 915 PM can cause a notebook to use more power than the 855PM, but not because the northbridge is a bad design, but that the PCIe bus on the 915 draws more power than the AGP on the 855PM chipset. In this case, both the 915PM and the ATI RX480M are running north and south bridge chips, with the same hard drive spec (5400 rpm IDE), and PCIe power draw as well. Both notebooks have the same number of So-Dimms, as close to the same CPU speed as possible, and the same RAM speed as well. The display is also exactly the same between them, and both were manufactured by the same plant (MSI).

In either case, you got your rep back, plus another 325 for good measure. However, please read first, then ask questions later.
thanks, I don't know if you should have, seeing i probably don't deserve it, but thanks

The problem is I am just to damn incapable of saying what i mean. When i had made my first post, I had skimmed enough to understand everything you had posted and and why THG had decide to go with the 22 watt, but i just was not convinced. that is where i made my first post. and all I have been trying to do it express (and failing at is) this one thing, hopefully i will make sense.

27watts-22watts=5watts
35watts-27watts=8watts

now assuming that the all the chips had memory controls that where inside the northbridge and chipset with same power requirements then you just compare the two with the lowest difference. In this case the you would use the MT chips.

But let me run a possible situation, lets pretend that intels chipset has TDP that is 3 watts great then that inside the turion setup(for the memory controller and other design differences). now I have no idea if how much power consumption a chipset takes, but lets prettend it is 13watts for intels and only 10 for ATIs(that is porbably nowhere near accurate, intels is probably more memory power efficent then ATi, but just pretend)

then the difference in chipset + cpu is

27+13=40
22+10=32
35+10=45

at which point it would be obvious that comparing the ML an the PM would be a more accurate represenation of total power requirements vs Speed. Since I do not have the numbers for the chipset all i can do is make guesses, and shots in the dark. that is why i think these number are vital to the a fair comparision, you maybe right, they maybe less then 1 watt, but it may not be.

hopefully this made sense.

Sorry, the rude words, I hope no hard feeling.
post #32 of 49
Thread Starter 
I understand you now. But you need to know that the memory controller is already factored into the chipsets TDPs:

AMD/ATi: CPU (active transistors + MC + 1MB cache) + NB + SB
vs
Intel/Intel: CPU (active transistors + 2MB cache) + NB (with MC) + SB

In both instances there are 5 components that produce heat for the chipset and the CPU, and while AMD's integrated memory controller certainly increases performance and decreases chipset complexity (and theoretically power draw), Intels additional L2 cache and ability to switch off almost all of it's L2 cache also increases performance and decreases power draw. The components are just shifted around. Now, if the chipset in the AMD notebook was an nForce4Go, for example, which would not have a NB as ATi has it, then I can understand the lack of another chip to feed making a difference in total TDP and power draw. In this case however it is almost too minimal to be measured.
post #33 of 49
Excuse my nOObiness, but i've just discovered reps...what implications do they have?

Karma...love the Fanboy letter...cracked me up.
post #34 of 49
Hmmm...the specs on the CPU's they compare though are quite different and the

Pentium M 750 1.86ghz

is put against the 1.8ghz MT 34, In msot results that 0.06 might have made the difference :/

They use different hard drive, different manufacturers, different dvd, different batteries...

Also you gotta consider that laptoplogic review used an ML 37 and 760 2.0GHz wich both consume more energy and are both 2.0 ghz
post #35 of 49
Yeah, I wouldn't care which processor Dell uses. They use the P-M right now, which is what I would go with. If they used the Turion, I would be fine with it. All I'm saying is, if I were to actually build my own laptop, that is, if I could, I would go with P-M over Turion.

"P-M makes for a terrific grandpa/grandma notebook! Email/Internet/Photos, that kind of things as battery last about 30 min extra when doing such stuff."
Gee, that's funny. I'm not a grandpa, nor a grandma. I guess I'd still be considered a teen, and I use e-mail. I use the Internet. I do stuff with photos. Wait a minute, so do all my friends, who, I might add, are also much younger than any grandma or grandpa. That's really odd. I guess I should stop using the internet and e-mail, and I should tell all my friends to do the same, as that sort of thing is reserved for grandpas and grandmas.
post #36 of 49
...last I checked this forum was in english...

Plus, Dell is hardly as good as it used to be....
post #37 of 49
Quote:
Originally Posted by runninkyle17
I cannot believe what I am seeing and hearing from you guys. First of all most people "in the know", know that Tom's Hardware is in Intel's pockets big time! So no wonder the tests are set to where everything is equal and the DDR and DDR2 tests are in Intels favor. Yes, I will admit that the DDR memory uses up more power than the DDR2 and this is why AMD is switching to DDR2, but c'mon people you need to look at the history behind the tests and who is doing the tests first.

It is no wonder that this review came out right after the Laptoplogic review (hmmmm...how convenient for Intel). Intel is rightfully scared of the Turion chips, even though Intel still makes and sells 85% of the CPU market (and I think the moble market is something like 95% to 98% Intel CPU's). Anyway, I am all about benchmarks and tests and everything else, but try not to read too much into the reviews of major websites and the like. Take PCWorld for example, their top 15 notebook PC's are usually pieces of crap (Dell, Compaq, HP, to name a few), but this is only because the magazine and website are deep into those companies pockets.

Let AMD have its fair share of the crown and in the mean time tell Intel to lower their prices and still churn out chips that can compete with AMD. Thank you very much, I will be here all week!!!
For someone who's "deep" in Intel's pockets, they seem to love AMD's desktop chips. Why don't you start making some sense before you say anything? Also, PCMagazine isn't exactly "enthusiast" level, and they're rating it based on productivity and price--not gaming performance. Champ.

Quote:
Since performance differences are so small I would go w/ what ever is your preference. However, there are some things you might want to consider before making the Turion 64 vs P-M (centrino platform) choice. If you are into future stuff like 64-bit generation of software and OS or SSE3, then you would simply go wrong if you get the P-M since it is incompatible with any of that. If you want to play it safe and save a fiew bucks go w/ a Turion 64 notebook.
And yet nobody has yet to make a compelling argument as to why you should go with a Turion. Most people won't even use SSE3 unless they're doing a lot of media, and considering the equivalent Pentium M processor costs about $30 more than the Turion MT-37 (which isn't available for purchase by itself), you really WILL only be saving a couple of bucks. $30 for longer battery life and performance isn't exactly going to put anybody out. Additionally, Windows XP 64... has shown little to no performance increases. Once more... wasted money, especially if you have to buy x64 separately--which most people will have to do if they're ordering it from somebody like Acer.

Also, karma, another reason I'd be interested to see them test a notebook with DDR2 and Intel instead of just DDR is because the DDR2 RAM uses 1.9v compared to DDR's 2.5. So there's quite a difference there... the 2.5 is almost twice as much as the processor, so I'm guessing RAM is probably going to be a big factor as well.

Quote:
Intels additional L2 cache and ability to switch off almost all of it's L2 cache also increases performance and decreases power draw.
Just to put some numbers with that (for other people's purpose), it shuts down all but 1/64 of its L2 cache, whereas the Turion has to completely power its entire cache regardless. That's why I'm wondering what would happen if Turion used some of these same ideas and put them into its design. Also, the hard drives shouldn't make much of a difference; since they're both 5400, and have identical cache, spin, etc., the only real difference is that one is 20gb more than the other--specifically, the AMD system. So if anything, the only performance difference--aside from a faster read speed--would be additional power drain on the Turion system. I'm actually surprised that they didn't bother changing out the drive... however, they have the same speeds in general, so the difference is not going to be a test breaker.

About the speeds... if anything, the additional FSB of the Intel chip would use MORE power. But that being said, the difference should be so subtle, it's literally going to translate into a minute of battery life.

Currently, the only real argument for the Turion (its still a nice processor, don't get me wrong) is a very, very small price difference, and the fact that its 64-bit. Since most software doesn't take advantage of SSE3 instructions--of which Intel will be adding in 2006--and the 64-bit edition software that ramificatio seems to hold so highly holds no performance differences except for software specifically designed to take advantage of it (read: no consumer level programs), the performance difference and battery life of the Pentium M simply makes it more acceptable than the Turion. Like I said, I love AMDs (my next desktop is [hopefully] going to have an X2), but while they reign supreme in the desktop market, AMD is rather infantile in the processor market.

Don't worry... I really have faith that AMD will overcome INtel's current hold given time. In fact, I would even say that within the next year, AMD's hold on the laptop market will have strengthened ten fold. Its just not reasonable to say that they're more powerful than the Intel for now, though...
post #38 of 49
To me it's like Turion 64 is just fighting of the P-M with one hand, yawning...

32-bit scenario
Performance: Similar per clock speed (some reviews show T64 faster some P-M)
Battery life in gaming: Turion 64
Battery life in office: P-M

64-bit scenario and SSE3
Turion 64: increased performance overall.
P-M: non-functional, error, nada, null, (you get the picture...) .

Just the facts, no?
post #39 of 49
I got a compelling reason why you should get Turion over Pentium-M....

<.<O.O>.>

Wisper*Intel kills puppies...*

AI got a nice compelling reason actually, if 2 things procide same performance and does not hurt you in any way. Get the Turion, why? so Intel will get scared and AMD gets more funding to compete and create better technology and then Intel and AMD will work harder and better processors due to competition.
post #40 of 49
WEll, lets see. 11 minutes of extra battery life when playing games (the Sempron definitely wins that one anyway), a whole 12 minutes with DVD playing (which would also depend on the optical drives), and 35 minutes below the Pentium M on battery life. Considering I have the exact same spec'ed machine only with a 100gb drive and 16mb of cache, 2GB of memory, AND a 2.0ghz processor, and I got 3:44 for battery life, I'm a bit confused as to why their battery times are so low. Perhaps the batteries are just low quality? Who knows?

Also, who knows what, for example, the video cards underclocked to? Also given that they could have tested the Pentium M using DDR2 (which is what most current Pentium Ms run on), that's nearly a 30% drop in energy consumption alone. Its also not counting into effect the different kinds of frequency drops when on the battery--which would skew both the AMD _and_ Intel results. For example, on my notebook, Power4Gear drops the processor speed to 500mhz--CHC only drops it to 800mhz. How is that going to be any different with all three notebooks?

Also, once more, you can give no compelling argument as to why anybody should pick 64-bit over 32-bit processors--and the same goes for SSE3. That's like saying, "Well, my car has a much bigger antenna. That makes it a faster car!" But it doesn't does it? Unless you have a reason to specifically support 64-bit processing, there's no reason to jump on the bandwagon.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Notebook Forums - General
NotebookForums.com › Forums › General Notebook Discussions › Notebook Forums - General › Extremely good Turion / P-M technical comparison