Pentium M beats up Turion64?

[Zebo]

Personally I though we knew the PM was maybe a lsight better clock for clock versus the A64 already in a desktop platform...

I never knew that. In AT's review a 2.0Ghz M only wins 11 out of 42 test to A64 2.0 Ghz 3200. Similar ratio in TR. (I hav'nt computed it but remeber talking about it then)

In GraftPC's review:
A 1.6 dothan wins every test to a 1.6 A64. And wins big.
:roll:

And dual channel's not the the excuse either since almost every site has shown DC and giving up 513kb L2 is about the same single channel and having 1MB L2 as happens here when making the comparison. Plus GraftPC review shows the 1.6 Dothan beating the desktop 3200 2.0Ghz chip most of the time too... LMAO. WHAT- EVAR!!


What I know about P-M is excellent low power. Could easily match or best A64 with some work like on board mem controller. When Intel gives up Netburst chips AMD coould be in deep trouble. When they fix these huge flaws:
http://images.anandtech.com/graphs/pent...20m%20desktop%203_02060570233/6221.png
http://images.anandtech.com/graphs/pent...20m%20desktop%203_02060570233/6231.png
http://images.anandtech.com/graphs/pent...20m%20desktop%203_02060570233/6232.png
http://images.anandtech.com/graphs/pent...20m%20desktop%203_02060570233/6235.png
http://images.anandtech.com/graphs/pent...20m%20desktop%203_02060570233/6236.png
http://images.anandtech.com/graphs/pent...20m%20desktop%203_02060570233/6237.png
http://images.anandtech.com/graphs/linux%20pentium%20m_12220491256/5861.png
http://images.anandtech.com/graphs/linux%20pentium%20m_12220491256/5839.png

Mainly to do with data sets and workstation performance Dothan could put the hurt on AMD. Not nesssarly because it's better clcok for clock. But even and that's all Intel needs for us to switch back.

One thing for sure is I won't be going to GraftPC for those reviews when they come.

 

[Insomniak]

Nothing there not to respect, really. The Pentium M is the one decent thing Intel has done in the past two years....considering they've been crashing and burning everywhere else, it's nice to see one team still has some chops.

 

[Zebo]

Originally posted by: Duvie
Remember the recent article about "sites to buy"..."reviews to be bought"...This is it PPL!!!!!

These fvckers have most likely neutered the Turion and A64 platform like most of the sites seem to be doing recently...Almost like a directive was sent from the mother ship telling them how to do it....Hmmmm INtel????

Edit: I would have to say looking at the encoding page 11 I think it looks about right for the apps they tested.....Remember they are showing 2.13ghz iin this article and 2.0ghz in the other reviews....That is 130 more mhz....

I agree they didn't some of them ore popular encoding like Xvid and Divx which likely would have gone to the A64's.....The 3400+ is a single channel sckt 754 chip....Had they used a 3500+ 2.2ghz DC even with 512kb it could have been different in a few of those test...Also they should have used a clawhammer to give the l2 cache which could have made a percent or 2 more....


I mean the bottom line is I can pick chips, drivers, other hardware components to know the way to manipulate the results...these sites do it all the time..compare ones best against someone elses weaker chip....

LOL, I've followed GraftPC's three Dothan tests very carefully, not by choice but because it was a sensational title in these forums, and everytime, it causes a big stir. Like when they had a OCed 2.4 Dothan whoopin a FX-55 while every other site showed the exact opposite. I don't know what to say anymore man it's just discusting.

 

[Duvie]

Zeb...I have to agree after a more thorough analysis of the data the 2.4ghz on average is less then a 4000+ 1mb clawhammer.....better then a 3500+ and probably closer to a 2.3ghz chip.....I used TR review for that...

here is the issue...notice the speech recognition performance and they said it was due to the weaker memory subsystem...Also probably why inthe other reviews gaming shows the P-M falls a little further back against the same speed AMD chip that has a better memory subsystem.....so it does make the gamePC review little a little weird like either the neutering of the memory subsystem with cas and using a single channle controller is just that manipulation I was talking about above to spin the results into what they really wanted to show....

We need to expose these sites for what they are and make sure ppl dont view sites like this as a definitive review site.....

I like Intel and god Knows I want them to do well and keep pushing the envelop...I would never think to even run an INtel system at anything but optimal. I am ashamed of some of the sites....At think other then test choices I think AT and TR are by far the best sites around, PERIOD!!!!

 

[Duvie]

Sidenote... Ever notice for how the P4-M even oc'd to 2.4ghz in TR review and suck soooo bad in cpu test in 3dmark05 (beaten by all the p4 chips) can be the only chip INtel has to compete with the top end Athlons??? I mean what is going on here....ooops I guess I forgot it is FutureMock....Another reason why synthetic benches blow chunks....

Why would they run 1024 res anway in games other then running 640 on them all....That is the only true cpu test.....





Why were the 755 and 745 in the AT review 2.00ghz and 1.8ghz and now the new chip seems to be 2.13ghz, 2.0ghz, and 1.86ghz...

IN the AT article it is 333ddr and 100mhz cpu for 400fsb....

IN the TR article it is same thing except the 2.4ghz has been OC'd to 533fdb abnd 33ddr ...both are using PC3200 but obviously downclocked.....

In the GamePC article it is 533fsb chips on all of them (at least the 2.13ghz and 1.86ghz and they are running 400ddr...

Basically this is a product of all enhanced 533fsb and then using 400ddr versus the old way using 400fsb chips and 333ddr....This would be huge in bandwidth and therefroe we cant be looking at the other reviews as a guideline...This could have made huge difference in games that love bandwidth....That was the p-Ms weakspot, right???

 

[Zebo]

Originally posted by: Duvie
Zeb...I have to agree after a more thorough analysis of the data the 2.4ghz on average is less then a 4000+ 1mb clawhammer.....better then a 3500+ and probably closer to a 2.3ghz chip.....I used TR review for that...

here is the issue...notice the speech recognition performance and they said it was due to the weaker memory subsystem...Also probably why inthe other reviews gaming shows the P-M falls a little further back against the same speed AMD chip that has a better memory subsystem.....so it does make the gamePC review little a little weird like either the neutering of the memory subsystem with cas and using a single channle controller is just that manipulation I was talking about above to spin the results into what they really wanted to show....

We need to expose these sites for what they are and make sure ppl dont view sites like this as a definitive review site.....

I like Intel and god Knows I want them to do well and keep pushing the envelop...I would never think to even run an INtel system at anything but optimal. I am ashamed of some of the sites....At think other then test choices I think AT and TR are by far the best sites around, PERIOD!!!!

Sure do need to get after them. But you know people will forget, they always do. I've done this with toms before too, who's not near as bad, still the #1 computer hardware site on net. Has'nt made any diff.


Re: synthetics. They are mainly made for P4 Which is why p4 does so well in them, and so average in real world. Dothan is totally different chip, operates real world fast like Athlon and p3s, so I expect a full rewrite of these synthetic programs when p4 disappears.

 

[Lonyo]

Originally posted by: DAPUNISHER
AMD is rumored to be going to a dual cored Turion so I imagine Merom or whichever Intel plans on to counter it with ain't far off. Inq says P-M gets a EE version next with much higher TDP that can ramp to 3ghz or beyond. Should keep things interesting

Yonah single and dual core. - 2006
There's a recent XBitlabs article with some info on it, but the site is down atm, so I can't link.

They say dual core Yonah with 27W TDP, and possibly 1467MHz core clock + architectural improvements, (as mentioned in the AT article on Dothan - multimedia accelerator etc).
Single core could be even less, maybe <20W?!
(They also say 65nm, so those power figures could be realistic)

 

[Zebo]

Yeah Lonyo it appears xbit was hacked...too bad it did'nt happen to you know whom.
http://www.xtremesystems.org/forums/showthread.php?t=60498&highlight=CBBID

 

[Leper Messiah]

well, when you use an A64 with 3-4-4-9 2T, of course its going to get its ass handed to it...

 

[Keysplayr]

Originally posted by: Zebo
GamePC is full of it. Some commercial PC sales site..yea right. Thier numbers totally disagree with anands. Where same speed A64 beat dothan in every game, see for yourself. GamePC has dothan winning by 25% in some games and winning every game.
http://www.anandtech.com/printarticle.aspx?i=2342

Whatever, I'll wait for some real laptop reviews from a site I trust not micheals computers or dell or gamepc..

Don't even bother with the current Dothan's Zebo. Wait and see what Yonah might bring. Whether it be band aids that mask the weakness of the Dothan, or true architectural core refinements. If all you do is game, then a Dothan is the best way to go in a desktop rig. Unfortunately, people do a lot more than just game.

 

[imported_michaelpatrick33]

I thought I saw dualcore Turions at 1.6 or 1.8 with 25 or 35watts for early 2006. This should get interesting.

 

[Zebo]

Originally posted by: keysplayr2003

Originally posted by: Zebo
GamePC is full of it. Some commercial PC sales site..yea right. Thier numbers totally disagree with anands. Where same speed A64 beat dothan in every game, see for yourself. GamePC has dothan winning by 25% in some games and winning every game.
http://www.anandtech.com/printarticle.aspx?i=2342

Whatever, I'll wait for some real laptop reviews from a site I trust not micheals computers or dell or gamepc..

Don't even bother with the current Dothan's Zebo. Wait and see what Yonah might bring. Whether it be band aids that mask the weakness of the Dothan, or true architectural core refinements. If all you do is game, then a Dothan is the best way to go in a desktop rig. Unfortunately, people do a lot more than just game.

Yeah I don't know what Intels problem is. They obviously have a chip that's superior to P4, maybe even superior to A64 with some desktop tweakage. It's low power req and hence low noise are insane. It's just more elegant all around than P4's brute force Mhz high power approach which I also appreciate. Bring it definity.

 

[Jeff7181]

Originally posted by: Zebo

Originally posted by: keysplayr2003

Originally posted by: Zebo
GamePC is full of it. Some commercial PC sales site..yea right. Thier numbers totally disagree with anands. Where same speed A64 beat dothan in every game, see for yourself. GamePC has dothan winning by 25% in some games and winning every game.
http://www.anandtech.com/printarticle.aspx?i=2342

Whatever, I'll wait for some real laptop reviews from a site I trust not micheals computers or dell or gamepc..

Don't even bother with the current Dothan's Zebo. Wait and see what Yonah might bring. Whether it be band aids that mask the weakness of the Dothan, or true architectural core refinements. If all you do is game, then a Dothan is the best way to go in a desktop rig. Unfortunately, people do a lot more than just game.

Yeah I don't know what Intels problem is. They obviously have a chip that's superior to P4, maybe even superior to A64 with some desktop tweakage. It's low power req and hence low noise are insane. It's just more elegant all around than P4's brute force Mhz high power approach which I also appreciate. Bring it definity.

It wouldn't make sense from a business standpoint... and Intel is definately about business first and foremost. But it seems like pride is holding them back. They've had two opportunities to abandon the Netburst architecture... instead they created Prescott, and then a dual Prescott. I really hope 65nm does something good for them because they're looking kinda bad right now.

 

[Fenuxx]

Originally posted by: Insomniak
Nothing there not to respect, really. The Pentium M is the one decent thing Intel has done in the past two years....considering they've been crashing and burning everywhere else, it's nice to see one team still has some chops.

That's exactly what I think as well. I was an avid Intel user until the company came out with that craptastic chip called Prescott, and since Intel insists on pushing the Prescott design on everyone, even though the design is flawed, that drove me to what really is my first AMD system (in my sig). I've been more than happy with this system, and, at least until Intel moves away from Prescott, I'll be staying with AMD for the time being. The P-M is a damn good CPU, and thats why I wasn't upset about a craptastic CPU when I ordered my Dell Inspiron 9300. It should be a damn good laptop, and, paired with a GeForce Go6800, it shouldn't do too shabby with games either. I would have, truthfully, liked to get a Turion64-based laptop, but for what my pricerange was, Turion64 was out of the question. Both CPU's have their strong-and-weak points, and IMO, they're pretty much on the same level.


EDIT: Damn Spelling gets me every time

 

[Suspicious-Teach8788]

Originally posted by: mamisano
Odd, the Pentium-M also smacked the Athlon 64 around quite handily!

And the Athlon 64 also smacked the Pentium 4 around quite handily! =P

 

[Vette73]

Originally posted by: Zebo

Originally posted by: carlosd
There is no deal with a Pentium M CPU for the desktop. Besides P-M is really weak in some tasks, for that reason it is not able to compete with A64 CPUs.

You got that right, CHERRY picked. Every test the M did surpurb on in techreports/anandtechs review GamePC includes. Every test it did poorly on they exclude.

Yea something does not add up. Sorry but I don;t trust that review at all.

 

[UzairH]

Intel are FOOLS for still using the Prescott core. Why don't they put an on-die mem controller on the Dothan, remove its weak points like branch prediction, and release a 3+ GHz desktop chip. THAT would rock the A64 boat.

 

[miketheidiot]

Originally posted by: UzairH
Intel are FOOLS for still using the Prescott core. Why don't they put an on-die mem controller on the Dothan, remove its weak points like branch prediction, and release a 3+ GHz desktop chip. THAT would rock the A64 boat.

i don't think that intel is cabable of relesing a 3 ghz dothan based chip(seeing as they are at 2.13 ghz or whatever currently), just like amd is incapable of releasing a 3.5 ghz k8 chip.

 

[Suspicious-Teach8788]

Originally posted by: UzairH
Intel are FOOLS for still using the Prescott core. Why don't they put an on-die mem controller on the Dothan, remove its weak points like branch prediction, and release a 3+ GHz desktop chip. THAT would rock the A64 boat.

Intel stated that it "is not a member of the HyperTransport consortium." Yea yea continue with that MCH crap.

 

[Viditor]

Just compare these 2 benchmarks from GamePC...

Turion Review
P4 EE Review

Note that according to them, a stock P-M blows the doors off of a P4 EE that has been overclocked to 4.0GHz on the exact same benchmark. In fact, it performs the same as an FX-55...

There be a foul stench here...

 

[Brunnis]

Originally posted by: Zebo
Guys almost everyone of thier benches disagrees with AT and techreport.

Not only do the benchmarks disagree, but GamePC's power consumption numbers are incorrect. The n00bs actually ran the Turion MT-34 on 1.35V, instead of it's specified 1.2V. This makes the numbers match up with the 35W ML-34, not the low power MT-34. Irritating is what it is.

Well, I've sent them a mail about this issue now.

 

[classy]

Let me clarify Gamepc on Dothan. Reading their review on Dothan is useless. They run Athon64 at really low memory timings Cas333. Its their way of making things "equal"? Obviously they skipped the classes on architecture. They are the only website that would lead you to believe Dothan is the best thing since sliced bread. Shame the other websites don't get the same results.

 

[Keysplayr]

Originally posted by: classy
Let me clarify Gamepc on Dothan. Reading their review on Dothan is useless. They run Athon64 at really low memory timings Cas333. Its their way of making things "equal"? Obviously they skipped the classes on architecture. They are the only website that would lead you to believe Dothan is the best thing since sliced bread. Shame the other websites don't get the same results.

Why are you all putting so much emphasis on the GamePC review? Other review sites will offer reviews that we can fight over.

 

[Zebo]

Originally posted by: Viditor
Just compare these 2 benchmarks from GamePC...

Turion Review
P4 EE Review

Note that according to them, a stock P-M blows the doors off of a P4 EE that has been overclocked to 4.0GHz on the exact same benchmark. In fact, it performs the same as an FX-55...

There be a foul stench here...

Heehee good find V, Even more so since the video card is the same and setup is the same. F Frauds.

 

[68GTX]

Originally posted by: Viditor
Just compare these 2 benchmarks from GamePC...

Turion Review
P4 EE Review

Note that according to them, a stock P-M blows the doors off of a P4 EE that has been overclocked to 4.0GHz on the exact same benchmark. In fact, it performs the same as an FX-55...

There be a foul stench here...

They used the same graphics drivers in This Review.

I'll just have to assume that the core & clock were the same between the AGP and PCI-E versions of the 6800 Ultras.