NotebookForums.com › Forums › General Notebook Discussions › Notebook Forums - General › Gaming: Turion 64 vs Pentium M
New Posts  All Forums:Forum Nav:

Gaming: Turion 64 vs Pentium M - Page 2

post #21 of 65
4/3/05 at 6:58pm
Quote:
Originally Posted by HardBall
Encoding has always been AMD's weakest point, so that's not that surprising. In most other tasks, the A64 will run away in a competition with top clocked PM.
If you look at the gamepc review, the stock 2.13GHz Dothan running on a modern chipset beats the 2.2GHz A64 in gaming, 3D rendering, LAME encoding, WMV encoding, Photoshop and Flash. It only loses in ScienceMark.
Reply
Reply
post #22 of 65
4/3/05 at 6:58pm
Quote:
turion still get the upper hand, remember as soon as MS finally launches XP64, 64bit apps will come out and the PM will now be obsolete
No. 32-bit computing may be obsolete in 5-6 years when no applications are made for it.
Reply
Reply
post #23 of 65
4/3/05 at 7:56pm
Quote:
Originally Posted by azote
man ! I'm so glad that i cancel my xps gen2

I saw that Windows XP 64 Bit went GOLD

then I saw the benchmarks between Pentium M and Turion 64 in gaming ..

gaming performance: http://www.theregister.co.uk/2005/03...iongaming4.pdf

Digital media performance:
http://www.theregister.co.uk/2005/03...iondigital.pdf

Office performance:
http://www.theregister.co.uk/2005/03...rionoffice.pdf

after seen this I'm very happy... of not getting the XPSG2

some turino laptops out there:
http://www.amdboard.com/turion_64_notebook.html

I know is old news but man .. no one told me to wait on the dell forums !
Me too Although, I think that the AMD benches are not accurate. I cancelled because of questions with the screen. The mobile AMD chip may not dominate the PM.
Reply
Reply
post #24 of 65
4/3/05 at 8:28pm
Well, if in fact the Turion does not compete with the PM, then I won't regret getting a Z71V when the time comes around. Besides, the prices are to drop when I should be getting my lappy, and the technology should be getting better. I might end up with a V6V body pumping out the power of a Z71V if I wait long enough That'd be hot
Reply
Reply
post #25 of 65
4/3/05 at 8:54pm
Quote:
Originally Posted by accord1999
If you look at the gamepc review, the stock 2.13GHz Dothan running on a modern chipset beats the 2.2GHz A64 in gaming, 3D rendering, LAME encoding, WMV encoding, Photoshop and Flash. It only loses in ScienceMark.
2.2GHz A64 has been there since September of 2003. A recently released speed upgrade finally beating an 18 month old chip; whoooohooo!!! And that was the only review where PM 2.13 beat the 2.2GHz A64 in gaming. If you compare the FX55, or the soon to be released 4200+ and FX57, then you would really see the A64 architecture pulling away in gaming and scientific apps, while still leading by significant margin on most other programs.
Reply
Reply
post #26 of 65
4/3/05 at 8:58pm
Quote:
Originally Posted by HardBall
2.2GHz A64 has been there since September of 2003. A recently released speed upgrade finally beating an 18 month old chip; whoooohooo!!! And that was the only review where PM 2.13 beat the 2.2GHz A64 in gaming. If you compare the FX55, or the soon to be released 4200+ and FX57, then you would really see the A64 architecture pulling away in gaming and scientific apps, while still leading by significant margin on most other programs.
So? Neither of these are notebook CPUs. Versus the expected top-end Turion, the P-M handily beats it while using less power.
Reply
Reply
post #27 of 65
4/3/05 at 9:03pm
Quote:
Originally Posted by accord1999
So? Neither of these are notebook CPUs. Versus the expected top-end Turion, the P-M handily beats it while using less power.
Exactly, but the people who support the so called "prowess" of PM are claiming that somehow the Dothan can beat the top desktop CPUs. PMs are great for notebooks, especially those < 5lbs, and that's where they belong. They won't become respectable desktop CPUs unless it undergoes some serious architectural changes, including having much more robust FPU, branch prediction, and get rid of the yoke of that tiny mem/IO bandwidth.

And don't make claims about PM vs Turion yet. No reliable bench has really been done head to head. Turions are based on K8 architecture, so I expect them to perform at least as well as the PM clock for clock. They initially, only come out up to 2.2GHz, so if Intel races to 2.4 PM, then they may have a slight lead. But we really have to wait and see on this. I don't really expect either one of them to blow the other one away.

And please back up your statement when you say:
Quote:
Originally Posted by accord1999
top-end Turion, the P-M handily beats it
Without even seeing a shred of evidence in head to head bench. Otherwise, you just look biased and bigoted and will be in danger of being labled an Intel cronie who trolls the forums.
Reply
Reply
post #28 of 65
4/3/05 at 9:25pm
Quote:
Originally Posted by HardBall
And don't make claims about PM vs Turion yet. No reliable bench has really been done head to head. Turions are based on K8 architecture, so I expect them to perform at least as well as the PM clock for clock. They initially, only come out up to 2.2GHz, so if Intel races to 2.4 PM,
If the 2.13GHz Dothan beats the 2.2GHz 512KB L2/S939 in those various fields, it's going to beat the 2GHz 1MB L2/S754 Turion, let alone the slower 25W models, which are the only ones close to the P-M's real power dissipation.
Reply
Reply
post #29 of 65
4/3/05 at 9:34pm
Quote:
Originally Posted by HardBall
Exactly, but the people who support the so called "prowess" of PM are claiming that somehow the Dothan can beat the top desktop CPUs. PMs are great for notebooks, especially those < 5lbs, and that's where they belong. They won't become respectable desktop CPUs unless it undergoes some serious architectural changes, including having much more robust FPU, branch prediction, and get rid of the yoke of that tiny mem/IO bandwidth.
What? This is a notebook fourms, and nobody has claimed that PM is the best desktop processor.
Reply
Reply
post #30 of 65
4/3/05 at 9:36pm
Quote:
Originally Posted by accord1999
If the 2.13GHz Dothan beats the 2.2GHz 512KB L2/S939 in those various fields, it's going to beat the 2GHz 1MB L2/S754 Turion, let alone the slower 25W models, which are the only ones close to the P-M's real power dissipation.
Beats it at what, mp3 encoding? Where are you getting your information anyhow? Why do you call the 25w models slower? Sounds like quite a bit of hearsay, and last i checked the sonomas were 27.5w right?
Reply
Reply
post #31 of 65
4/3/05 at 9:48pm
Quote:
Originally Posted by gozi
Beats it at what, mp3 encoding? Where are you getting your information anyhow? Why do you call the 25w models slower? Sounds like quite a bit of hearsay, and last i checked the sonomas were 27.5w right?
From the gamepc review, games, 3d rendering, audio encoding, WMV encoding, Photoshop and Flash. 27W is the family TDP for the 533MHz FSB Dothans, set at a level for future CPUs. x86-secret.com measured the 2GHz/400MHz FSB Dothan at 16W running Prime95.
Reply
Reply
post #32 of 65
4/3/05 at 10:07pm
Quote:
Originally Posted by accord1999
If the 2.13GHz Dothan beats the 2.2GHz 512KB L2/S939 in those various fields, it's going to beat the 2GHz 1MB L2/S754 Turion, let alone the slower 25W models, which are the only ones close to the P-M's real power dissipation.
Again, you are talking about the old Newcastle CPUs, which has already pretty much bit the dust, even the Winchester will be replaced later this month by San Diego and Venice cores, which will have SSE3, new stepping, and other improvements along the way. It would be like comparing the old Banias CPUs with today's A64s, which is simply anachronism.

By the way, you havn't shown us a single link about what you claim to be an advantage of Dothan over the 2.2GHz (presumably 3400+ or 3500+) A64.

And I have used many laptops in the past (my own or others), including several PM laptops, they are great at business apps, fair on encoding, OK at integer statistics, but absolutely sucks at gaming, and even worse at compiling code. Please don't try to tell us how PM is a great all around CPU, because it's simply not true; it can competing in float calculations to save its life.

Whether the Turion is better or worse, it's anyone's guess at this point. Most likely it will be better on some applications, while slower on others. And since it's based on K8, I have somewhat of an idea how it will perform, but am by no means certain about the outcome.
Reply
Reply
post #33 of 65
4/3/05 at 10:13pm
Quote:
Originally Posted by B Nietsnie
What? This is a notebook fourms, and nobody has claimed that PM is the best desktop processor.
Just look at some of the other threads, such as one titled:
Proper review of Pentium M on the desktop

http://notebookforums.com/showthread.php?t=62363

Or the:
New Dothan Gaming Review

http://notebookforums.com/showthread.php?t=49943

I'm certainly not the one who started it. And just read these and other similar threads, you will see people claiming left and right that the Dothan is the future of Desktop CPUs.
Reply
Reply
post #34 of 65
4/3/05 at 10:51pm
Whats the deal with the Turions having 512k of L2 Cache? Is AMD trying to pull a Celeron or some bullshit? Celerons have 512k L2 Cache, sure the 2 MBs in the Dothan is a bit excessive, but I'd rather have too much than not enough
Reply
Reply
post #35 of 65
4/3/05 at 11:04pm
all thest test are useless in the end i find
each user has there own preference
and each chip is designed to handle information differently.
example:: amd might be better at games

or intel might be better at encoding etc

that is why i say have the best of both worlds !
Reply
Reply
post #36 of 65
4/3/05 at 11:08pm
Quote:
Originally Posted by MichaelX30
Whats the deal with the Turions having 512k of L2 Cache? Is AMD trying to pull a Celeron or some bullshit? Celerons have 512k L2 Cache, sure the 2 MBs in the Dothan is a bit excessive, but I'd rather have too much than not enough
Only some of the Turions have 512 K, others have 1MB, and they will be rated the same way A64s have been, with the clock speed and the cache playing a role.

You have to remember, the only reason the PM need such a large cache is because of the insufficient memory bandwidth, which is at best 533 MHz shared with I/O.

Turions, on the otherhand, has independent memroy channels, and should experience no contention problems. And since th controller is on die, there should be much lower latency than the PM. So having smaller cache is simply the lack of necessity for larger caches, for the Turion. And if anything, I would expect the overal memory performance to be better than the Sonoma, since the A64s have the same cache are outperforming the P4s with 800MHz FSB and 1-2MB L2.
Reply
Reply
post #37 of 65
4/3/05 at 11:18pm
In the case of gaming on a laptop. the CPU should be the last thing on the list. Pay more attention to the kind of GPU you can get with the size, weight and battery performance that you want. Even back when I have my Aspire 2025 with a 1.8G P-M and 128MB or MR9700. I have never seen it run at 100% during gaming as I run out of GPU power way before then
Reply
Reply
post #38 of 65
4/3/05 at 11:29pm
is there any proof that 64bit desktop applications will out perform 32bit desktop applications anytime soon? Making the assumption that 64bit will double your performance is as flawed as thinking a 32-way server will out perform an 8-way server by a magnitude of 4, which is not the case because of software development not writting applicable threading for such an environment. That means that current applications even when ported to 64bit should see little benefit.
Reply
Reply
post #39 of 65
4/3/05 at 11:43pm
Quote:
Originally Posted by Michelangelo
is there any proof that 64bit desktop applications will out perform 32bit desktop applications anytime soon? Making the assumption that 64bit will double your performance is as flawed as thinking a 32-way server will out perform an 8-way server by a magnitude of 4, which is not the case because of software development not writting applicable threading for such an environment. That means that current applications even when ported to 64bit should see little benefit.
64-bit systems won't outperform the 32-bit ones by a single nano-second by itself. But what it affords is the enormous latitude for the developers, where much more accurate numbers can be stored in registers and much more addressed system memory can be used when compiling programs, and eventually, when running programs on consumer machines (when memory gets cheap enough).

What do you mean by 8-way servers??
"8-way' server" = server that can run 8 threads simultaneously; i.e. with 8 logical cores.
Reply
Reply
post #40 of 65
4/4/05 at 2:08am
Quote:
Originally Posted by HardBall
64-bit systems won't outperform the 32-bit ones by a single nano-second by itself. But what it affords is the enormous latitude for the developers, where much more accurate numbers can be stored in registers and much more addressed system memory can be used when compiling programs, and eventually, when running programs on consumer machines (when memory gets cheap enough).
Increased address space isn't the only thing that 64-bit brings to the table; remember, with the new 64-bit programming model comes 8 extra GPRs and 8 extra SIMD registers. This alone has the potential to contribute a noticable performance boost to many applications compiled 64-bit native, though mileage will obviously vary from app to app.
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Notebook Forums - General
NotebookForums.com › Forums › General Notebook Discussions › Notebook Forums - General › Gaming: Turion 64 vs Pentium M