An indepth look at the Pentium M aka Dothan

[TriniXaeno]

http://www.sudhian.com/showdocs.cfm?aid=635

    
If you’re an Intel enthusiast, the last eleven and a half months haven’t been especially bright, especially when compared to the heady days of 2002 and 2003, when Northwood hadn’t met a clockspeed it didn’t like, and Canterwood could do no wrong.  The retail market’s pre-launch anticipation of what was widely seen as Pentium 4’s pre-emptive strike against the Athlon 64 was badly deflated by the higher temperatures, decreased efficiency, and mediocre scaling Prescott actually delivered.

The launch of Alderwood and Grantsdale in June hasn’t done much to change the situation.  While these chipsets are fine, stable products in and of themselves, they’re Prescott-only unless you’re one of the rare individuals who can afford an Extreme Edition.  For users replacing older systems the chipsets themselves could be attractive options, but they don’t offer much in the way of compelling features to the enthusiast market, and they’re saddled with Prescott’s thermal baggage.

Prescott may be as attractive as the idea of belly flopping into a concrete-filled swimming pool, but another star has been steadily rising over Santa Clara.  Intel first debuted the Pentium M (codenamed Banias) when they launched the Centrino mobile platform back in 2003, but enthusiasts didn’t really begin giving it a serious eye until Banias transitioned to 90nm, picked up an extra meg of L2 cache, and became Dothan.  Its easy to see why consumers turned off by Prescott’s thermal signature and power requirements might be attracted to the ultra-cool, low-wattage Dothan—but does Intel’s mobile processor have what it takes to compete on the desktop market?

If we accept oft-repeated, marketing-driven IT “wisdom” as fact, Dothan on the 855 platform can’t possibly offer the performance that it does. AGP 4x and single-channel DDR333 couldn’t possibly offer the same robust performance as a brand-new DDR2-533 chipset with Prescott (or Athlon 64) support, PCI-Express, and integrated gigabit!  Look at the bigger numbers?  Everyone knows that bigger numbers mean better performance!

Except, of course, when they don’t.

While it wouldn’t be fair to paint all the new platform technologies of the last few years as worthless, it’s fair to say that much of their value to the typical (or even not-so-typical) end user has been severely over-hyped.  Looking at our results, it’s hard to argue that AGP 8x or single-channel PCI-E does a damned thing for gamers.  The old “future games will show a benefit” shtick isn’t working either; AGP 4x is a four-year old standard that’s been the basis for every DX7, DX8, and DX9 part.  While I’m willing to believe that the bi-directional interface of PCI-E may allow for some performance tricks that aren’t in use yet, it’s obvious that the last four years of “advances” to the AGP / PCI-E interfaces haven’t bought us jack in terms of advanced gaming performance. DDR2 has a certain degree of promise, but until Prescott can synchronize its FSB and memory busses on at least DDR2-533, that benefit will remain hidden.  Intel’s decision to use a 2 meg L2 cache in the Pentium 6xx series coming next year will cut the real-world benefit even farther.

Ironically, the real improvements of the past few years are the ones you don’t hear marketed so aggressively, probably because they’re harder to explain to your average consumer.  The performance benefits of 875 Canterwood were only slightly explained by its dual channel DDR400 interface; Intel’s CSA and PAT technologies made substantial contributions of their own by removing Ethernet communication from the PCI bus and cutting RAM latencies.  If NVIDIA’s numbers are valid, it’ll be TurboCache that benefits the low-end user; your typical gamer or user with a single video card won’t see a benefit from PCI-Express over AGP.  (High end gamers, of course, can use SLI).

Do platform improvements matter?  They do—but often not in the ways or for the reasons marketing departments would like you to think.  Going by the specs, Dothan and the 855GME platform are a joke.  In reality, Prescott isn’t laughing.  

Read the full story here.

http://www.sudhian.com/showdocs.cfm?aid=635

Here are a few charts to get you drooling.




                   

[TrinireturnofGamez]

INTEL is supposed to be souping up the dothan core and using them to replace the presscot ( hot) .. ... till then i will stay miles away from a pentium
    the dothan can overclock to 2.8 ghz , at that speed it beats out EVERY processor in existance for gaming, though its  application performance is horrible, intel will fix it sooner or later .                      

[Synth]

2.8 ghz ? What are you smoking? Thats even worse than you when you said XP mobiles hit 2.7 or so.

Dothan weakness is memory bandwidth plain and simple. So its worthless to real AV applications. Other than that its design efficiency is on same level as an a64.                    

[TriniXaeno]

which is simply amazing for a chip based on the P6 core and to see those scores on a platform using such old tech besting todays latest and greatest tech is a slap in the face to our "progress" all these years.

so much for agp 8x, pci express, 800fsb, hyper threading, 64bit, etc....

Seems they do jack for gamers.                    

[TrinireturnofGamez]

sythn when i say something , don't discount it for nonsense , LOOK IT UP before you complain , and  i DID say that the pentium M had horrible application performance  

I then reached a maximum stable speed of 2707Mhz, 245MHz FSBx 11, with 2.06V. Not much higher, but good nonetheless. It was the highest I've seen yet from these processors. I was able to take a screenshot at 2804Mhz, but it froze right after I had saved the *.bmp. I have run 3dmark2001, with EXTRA high 9800Pro overclock to 526/425, and currently hold the record air cooled AMD XP/9800Pro score: Click Here That is the 5th highest score for AMD Athlon XP and 9800 Pro overall. I was pretty stoked about those runs.

: athlon XP mobile on AIR cooling , mines would reach 2.6ghz @ 1.9vcore .  
http://www.fastlanehw.com/reviews.php?i=102&page=4

http://www.fastlanehw.com/reviews.php?i=102&page=5

After a few people had heard about my acheivement with this CPU on air, a friend from the forum, who lives nearby, asked me if I'd like to borrow his Prometia Mach I cooler. Like any serious overclocker, I immediately jumped on the offer, and he came by my house that weekend. He was excited to see what I could do with the cooler and this CPU, so I set out to get as high as I could.

I installed the Prometia CPU kit, and it's ready to go!
I was able to attain 2912MHz perfectly stable with 2.1V for all my benchmarks, and those will follow on the next page.,

http://www.gamepc.com/labs/view_content.as...cookie%5Ftest=1

  DFI even let us know that some users are seeing overclocks up to 2.7 GHz, which is amazing given the nature of the Pentium-M processor. We were able to get a glimpse of these speeds by using ClockGen, a windows overclocking utility which supports the DFI motherboard. Using this utility, we were able to boost clock speeds up to 2.68 GHz on our Pentium-M processor, although at this level, we started to see major instability. At 2.53 GHz, the system was completely stable for all of our benchmarks and did not show a single hint of instability.

 
  The DFI website itself had an article of it @ 2.8ghz , can't find the link right now .




so synth... go suck some salt    and get off my case already .  
   

     which is simply amazing for a chip based on the P6 core and to see those scores on a platform using such old tech besting todays latest and greatest tech is a slap in the face to our "progress" all these years.

so much for agp 8x, pci express, 800fsb, hyper threading, 64bit, etc....

Seems they do jack for gamers.

: no game uses more than AGP 4x bus speeds , even doom III , few games are multi threaded to begin with so hyperthreading is a waste .....                    

[TriniXaeno]

lol @ rebuttal

yeah, it is quite a shame that we have all these "advancements" and don't get to see the benefits of them.

At least in the video card arena, all the new bells and whistles tend to get leveraged pretty early on.                    

[coldstorm]

please do remember that the 2130 cost $645 US and the 1700 cost $245 .  a athlon $3500 cost 259 us
What save the day for the pentium M is the 22 w TDP and the 10 clock L2 cache that is the best in the current generation of processors. When  intel decide to use that L2 in the chips they may reclaim the gaming crown they lost to amd                    

[TriniXaeno]

That ain't save the day at all.

At that price, the Athlon 64 is the better bet!

The point was just that the technology so old yet kicking so much brass.

Personally, I would still recommend an Athlon 64 939 platform as a no brainer decision for buying a gaming rig today.                    

[Synth]

Its very hard not to automatically think all that comes from your mouth is crap. And I am not the only one who thinks of your words like this.

Anyway since you need once again a lesson in logic here I go:

I then reached a maximum stable speed of 2707Mhz, 245MHz FSBx 11, with 2.06V. Not much higher, but good nonetheless. It was the highest I've seen yet from

these processors. I was able to take a screenshot at 2804Mhz, but it froze right after I had saved the *.bmp. I have run 3dmark2001, with EXTRA high 9800Pro

overclock to 526/425, and currently hold the record air cooled AMD XP/9800Pro score: Click Here That is the 5th highest score for AMD Athlon XP and 9800 Pro

overall. I was pretty stoked about those runs.

Is this realistic? Who cares if this can acheived with extreme method? Thats like saying one should worship a Pentium 1 Classic cause if I feed it 10Volts and nitrogen cool it, it can reach 1Ghz!

Is it 100% stable 24/7? Get my point?

And even if his can do this it is by no means representative of real world results. There haven't been enough of these produced and tested to be of any statistical value to your "facts".


If you wanna drool over unrealistic "rare" events like this be my guest, just don't post like its a common thing and in the process misleading the public. And he didn't even hit 2.8 stable for that matter.

And don't tell me that your not misleading since anyone who reads your post will be inclined to believe that its a miracle chip which can hit 2.8 easy.

Get my point?



And while you at it comparing a 2.8GHz dothan ("beats out EVERY processor in existance for gaming") which btw isnt stable to begin with so I don't know how you can even use that as "evidence"...(theres those mystical "facts" again)...

An a64 at 3.4Ghz will beat it...EASY. Or even a lowly 5.5Ghz Prescott would for that matter.

Unfair you say? To that I say - Its all extreme and none are realistic.

Get my point YET?




 

: athlon XP mobile on AIR cooling , mines would reach 2.6ghz @ 1.9vcore .
http://www.fastlanehw.com/reviews.php?i=102&page=4

http://www.fastlanehw.com/reviews.php?i=102&page=5

After a few people had heard about my acheivement with this CPU on air, a friend from the forum, who lives nearby, asked me if I'd like to borrow his

Prometia Mach I cooler. Like any serious overclocker, I immediately jumped on the offer, and he came by my house that weekend. He was excited to see what I

could do with the cooler and this CPU, so I set out to get as high as I could.

I installed the Prometia CPU kit, and it's ready to go!
I was able to attain 2912MHz perfectly stable with 2.1V for all my benchmarks, and those will follow on the next page.,

Your xp mobile can run 2.6 @ 1.9v? 24/7? If so then why arent you running it
Your comparing a vapor phased result...and extreme voltaged


None of your past posts suggested that your facts are based on realistic conditions...realistic expectations. Not the xp mobile post and not this one.

Just read any major o/cing forums and you will see that 2.6Ghz and above is very rare...muchless on realistic environments.

2.4 maybe even at a stretch 2.5. BuT DEFINITELY NOT 2.6, 2.7 or 2.8!


I should know since I live on those types of forums and I don't trust techsites.


Its all misleading; As is much of your posts.
 



Just a tip, dont base your "facts" so much on tech reviews from tech sites, which are often technically wrong or unrealistic.
Use actual public results on forums...and based on realistic stable environments.




http://www.gamepc.com/labs/view_content.as...cookie%5Ftest=1

 

  DFI even let us know that some users are seeing overclocks up to 2.7 GHz, which is amazing given the nature of the Pentium-M processor. We were able to get

a glimpse of these speeds by using ClockGen, a windows overclocking utility which supports the DFI motherboard. Using this utility, we were able to boost

clock speeds up to 2.68 GHz on our Pentium-M processor, although at this level, we started to see major instability. At 2.53 GHz, the system was completely stable for all of our benchmarks and did not show a single hint of instability.

 
MAJOR INSTABILITY AT 2.68Ghz...and DFI themselves in their  own high tech labs using there own mobo? Should I say more?

Even if these were realistic results it still shows that a STABLE 2.8Ghz is a MIRACLE AT BEST. And there goes all your "proof".



Just a little history note...Intel used to promise Prescott tech could hit 5Ghz "easy"...what happened in reality?
The moral of story? Until a certain state of statistical usage has been acheived any results are mere entertainment at best.

so synth... go suck some salt  and get off my case already .

You really wouldn't like to aggravate me more. I already have a lecture prepared to post (about you) should you continue to defend your illogical misleading and often just plain wrong ways. I didn't post till now since I decided to give you another chance.

no game uses more than AGP 4x bus speeds , even doom III , few games are multi threaded to begin with so hyperthreading is a waste .....

hahaha this coming from a man who released upon on the gatt public not too long ago a magical "hyperthreading" tweak...Let me break it down now for you:

So then what you did was proclaim to the public about a tweak, to attain a state, which in your own words is "a waste"?

What does this now say about you? Nothing we didnt already know, I suppose.                    

[TriniXaeno]

whoa synth!!

cut him some slack.

It's called "enthusiasm!" and "passion"

lol. Ok, so he isn't exactly the pillar of technical integrity today, so what? Let the man make his post. You and I are here to correct him when he is offbase (as we just did) without having to resort to degrading him.

I have a good feeling that he will surpass us when he get's to our age. The sharp edge of his passion just needs to be tempered with a sheat of real world experience. That will come in time, fear not.

Together we aspire, Together we achieve!                    

[TrinireturnofGamez]

i'v run at 2.6ghz for 12 hours , and prime 95 showed no errors , though the temps would load up to 65  degrees........ which i consider very unhealthy ....... i don't ALWAYS run it at that speed because i don't need the speed , and i would think 55 idle and 65 load vs 41 idle and 57 load will kill my machine much faster. ...
    i see your point that some chips may not reach reported speeds....  but you see up to now i have always gotten decent , perhaps even exceptional overclockable hardware..... so i trust the reports of the 'extreme' overclocks to be a realistic report ...
  i got my 8500 up to 300mhz/580mhz , my 6600gt to 560/1.2GHZ and the athlon XP mobile to 2.5/2.6ghz  .
   i also notice something else , over time your parts give you better and better overclocks under the same conditions  ..... (question for synth) is it the thermal paste 'curing' better , or is it 'burning in'  ( i never understood that term )