Pentium M Dothans on a 478 MicroATX board?

L

Lucidguy_3000

n00b
Joined
Nov 10, 2004
Messages
58
I know people run P4M's on 478 boards, is the same possible with the dothans? im trying to make a portable rig and the need for less cooling is very attractive. Anyway, thanks for any input! :)
 
Dothans are Socket 479, which is more than Socket 478 with an extra pin. Here's a link to the AOpen board, and one to the DFI board. Those are the only ones I know of at the moment.
 
DFI is now selling mobos specifically for the Dothans. They are expensive, but their overclocking potential is pretty good and they are darn fast when overclocked.
 
BigBadBiologist said:
DFI is now selling mobos specifically for the Dothans. They are expensive, but their overclocking potential is pretty good and they are darn fast when overclocked.
Click to expand...
You have been beaten like a red-headed stepchild ;) AOpen also sells Dothan boards, and I trust AOpen more than DFI for some reason ...
 
Well then Newegg needs to fix their site, the dothans are listed as Pentium M 478's :rolleyes: thanks for the info though!
 
evildre said:
You have been beaten like a red-headed stepchild ;) AOpen also sells Dothan boards, and I trust AOpen more than DFI for some reason ...
Click to expand...

Well, I just haven't seen any reviews of the AOpen board yet. I have seen a review of the DFI.
 
evildre said:
Dothans are Socket 479, which is more than Socket 478 with an extra pin. Here's a link to the AOpen board, and one to the DFI board. Those are the only ones I know of at the moment.
Click to expand...
no, dothans are 478 and 479, in notebooks they are 478, and on desktops they are 479. the ones newegg has are 478, and the s479 ones are rather hard to find

i imagine intel did that just so poeple wouldnt put them in their P4 boards now that there are desktop boards for p-m
 
lithium726 said:
no, dothans are 478 and 479, in notebooks they are 478, and on desktops they are 479. the ones newegg has are 478, and the s479 ones are rather hard to find

i imagine intel did that just so poeple wouldnt put them in their P4 boards now that there are desktop boards for p-m
Click to expand...
Really now? This is the first time I've heard that ... the Centrino notebooks I've built (Compal CL51, Asus M5N, etc.) have all had Socket 479 motherboards :confused:
 
evildre said:
Really now? This is the first time I've heard that ... the Centrino notebooks I've built (Compal CL51, Asus M5N, etc.) have all had Socket 479 motherboards :confused:
Click to expand...
you might wanna go back and count again :p cuase the mobile versions of the p-m are in a 478pin package. they are different than the p4's socket IIRC, but there are 478 pins. the 755, 745, 735, 725, and 715 dont even have a s479 counterpart

http://processorfinder.intel.com/scripts/list.asp?ProcFam=942&NoNav=NO&CorSpd=ALL&SysBusSpd=ALL&MfgTech=ALL&step=ALL&cache=ALL&PkgType=ALL&btnFOS=Filter+on+selections

edit: after looking over the thign i linked too, it doesnt seem that the packages have any specific platform differences (ie, dekstop or notebook, there are 478 ad 479 versions of banias chips evedently..) so i have no idea what the hell intel is smoking :confused:
 
Then how, pray tell, did that review of the 2 GHz Dothan overclocked to 2.53GHz on a DFI board happen, when there is no socket 479 755 on intel's site?
 
robberbaron said:
Then how, pray tell, did that review of the 2 GHz Dothan overclocked to 2.53GHz on a DFI board happen, when there is no socket 479 755 on intel's site?
Click to expand...
because they used the 2.1ghz chip, it was brand new when they tested it. i remember that review too, and i rememebr thinking, now when the hell did they come out with a 2.1ghz??
 
lithium726 said:
the review i read on DFI's site when the board was announced said 765

if it was a 755, i dont know. call DFI and ask them

edit: you know, it couldve been an unreased ES..
Click to expand...

I emailed the author of the review asking him if it was an engineering sample. I'm not saying you don't know what you're talking about, I just find it odd that intel wouldn't have anything on their site about it.
 
robberbaron said:
I emailed the author of the review asking him if it was an engineering sample. I'm not saying you don't know what you're talking about, I just find it odd that intel wouldn't have anything on their site about it.
Click to expand...
good man. its also quite possible that intel just hasnt updated the s-spec chart, and in that case the 90nm 479 chips might be avaliable. the entire point of my posts was to say that there are indeed 478 pin chips and that the ones that newegg has are the 478 version.
 
I just bought a dfi 855gme mobo from newegg along with a pm 1.6 does this mean they wont work together? I checked intels website s spec finder and you are right the ones newegg sells are 478. A 478 chip should work in a 479 board right? well if not ill be sending the stuff back to newegg. I'll find out soon enough it should be delivered today. I can do a overclocking and mobo review if they work.
 
Exocet956 said:
I just bought a dfi 855gme mobo from newegg along with a pm 1.6 does this mean they wont work together? I checked intels website s spec finder and you are right the ones newegg sells are 478. A 478 chip should work in a 479 board right? well if not ill be sending the stuff back to newegg. I'll find out soon enough it should be delivered today. I can do a overclocking and mobo review if they work.
Click to expand...
be sure and tell us if it works! if it does that would be awesome :D
 
OK, to clear up some confusion here, there are 2 different types of P-M processor. There is the 478 pin FCPGA package, which is what the Aopen and DFI boards both accept and then there's the 479 pin FCBGA (flip chip ball grid array) package, which won't fit those boards. I'm not positive, but I think the FCBGA procs are meant to be soldered directly to the board (correct me if I'm wrong there).

You can look this information up in chapter 4 in the P-M on 90 nanometer data sheet, which is around page 40 . Hopefully this will ease some people's worries here.

BTW, I read a review or 2 of the Aopen board around a month or 2 ago and it looked pretty good for a board with no vcore or vdimm options. It has a neat little windows overclocking app that lets you adjust the multiplier and fsb speed in windows.
 
well so far it is running awesome. At stock this dothan 1.6 is equaling or beating my p4 3.06 gig. Im prime95 testing @ 2.14 right now. This thing is so damn fast it scares me. Hell just the preliminary tests show a crazy numbers. This board booted up first time. i even swapped boards with my 3.06 gig did a os repair install didn't have to reinstall a single app or driver i only had to load one new driver for the onboard nic. Ive just to begun to find out what this chip is capable of. Im still experimenting with the memory dividers. Im running 16x133 at 1:1 right now. its been running prime for about seven hours now so im fairly confident its stable at this speed. IM going to see if i can get this chip to run at 2.2 or 2.3. It would get into windows but would either lock up or be really slow. The chip i think can do it i just may have had the wrong memory dividers. temps do not seem to be a problem. even with the tiny ass heatsink and 40mm fan on top. it actually seems to get cooler when its overclocked. i could use the name of another app that i could use to monitor the temps that supports this board. the numbers im getting are anywhere from 25-45c. right now under prime 95 im getting 45c that seems to be the highest it ever gets. although it was lower earlier so I don’t know if the app is reading the temps accurately. O h well got to go to work now so ill keep you guys informed. cya
 
Last I read, all Dothans are actually 478 pins, if you count the actual pins. Intel uses 479 as a means of identifying them as electrically different.

If you want to know more about overclocking Dothans, xtremesystem.org has a ton of posts on them. The new Alviso chipset, with dual channel support and a higher FSB should be out soon.

Try here: http://processorfinder.intel.com/scripts/default.asp
 
The DFI 855GME is mPGA479M. I could see how an mpga478m proccessor would be compatible with that. I am going to get this board asap.
 
Well I’m back and I have some results for you guys. First let me explain how I am going to conduct this review. I will first cover my testing methodology. I realize that my aged AIW 9800 pro is going to be the bottleneck in these benchmarks but I have accounted for this and in addition to modern games I have run older engine games at a low resolution to remove this bottleneck. I ran my standard benchmark load out that I use to evaluate all systems that I own by.
 
First off let me say that between the two windows applications ClockGen and RightMark Cpu Clock rate Utility give you complete control of: FSB, multipliers, and vcore in windows. RightMark also allows notebook type power management by dynamically adjusting clock speed, multipliers and vcore. As the cpu needs more power it adjusts the clock speed to give additional power when needed. When power is not needed it clocks the processor down to save power. Unlike the notebook version, which does not work with these desktop motherboard it adjusts far faster than my notebook. Plus it has different profiles so you can choose your levels of performance. In automatic mode it functions without user intervention but for those who have to have complete control you can manually set different power states. Clockgen is pretty self explanatory it allows you to adjust the FSB in windows and yes the FSB does affect power savings. I.E. at 133fsb instead of idling at 600mhz my 1.6 Dothan idles at 800mhz. However be very careful it does not allow memory divider adjustment in windows so you are stuck with what you set in the bios. If I wanted to run a different divider at stock the instant I upped the FSB to where I knew it should be stable it crashed.
 
The testing setup is as follows:
Engine Demo Resolution Quality Sound

Quake 1 Bigass1 640x480
Quake 2 Crusher 640x480
Quake 2 Massive1 640x480
Serious Sam 1 MetroCoop 1280x1024 HQ NO
Serious Sam 1 MetroCoop 1280x1024 HQ YES
Serious Sam 1 MetroCoop 1024x768 HQ NO
Serious Sam 1 MetroCoop 1024x768 HQ YES
Serious Sam 2 Grand Cathedral 1280x1024 HQ NO
Serious Sam 2 Grand Cathedral 1280x1024 HQ YES
Serious Sam 2 Valley of the Jaguar 1280x1024 HQ NO
Serious Sam 2 Valley of the Jaguar 1280x1024 HQ YES
Serious Sam 2 Grand Cathedral 1024x768 HQ NO
Serious Sam 2 Grand Cathedral 1024x768 HQ YES
Serious Sam 2 Valley of the Jaguar 1024x768 HQ NO
Serious Sam 2 Valley of the Jaguar 1024x768 HQ YES

3Dmark2k1
3Dmark2k03

Subjective test: Half-life 2 Game play.
 
Benchmark results

Reference System
________________________________________________________________
P4 3.06 gig Northwood (23x133) 166 MHZ DDR single channel

AVG FPS Engine Demo Resolution Sound

445 Quake1 Bigass1 640x480
206.1 Quake2 Crusher 640x480
322 Quake2 Massive1 640x480
120.3 SS1 MetroCoop 1280x1024 NO
119.2 SS1 MetroCoop 1280x1024 YES
140.4 SS1 MetroCoop 1024x768 NO
137.9 SS1 MetroCoop 1024x768 YES
82.5 SS2 GrandCath 1280x1024 NO
80.4 SS2 GrandCath 1280x1024 YES
51.9 SS2 Valley 1280x1024 NO
52.2 SS2 Valley 1280x1024 YES
92.6 SS2 GrandCath 1024x768 NO
89.3 SS2 GrandCath 1024x768 YES
68.6 SS2 Valley 1024x768 NO
68.1 SS2 Valley 1024x768 YES

15253 3Dmark2k1
5550 3Dmark2k3
___________________________________________________________________
 
________________________________________________________________
PM 1.6 gig Dothan (16x100) 100 MHZ DDR single channel

AVG FPS Engine Demo Resolution Sound

579.7 Quake1 Bigass1 640x480
247.9 Quake2 Crusher 640x480
407.7 Quake2 Massive1 640x480
123 SS1 MetroCoop 1280x1024 NO
122.2 SS1 MetroCoop 1280x1024 YES
161.1 SS1 MetroCoop 1024x768 NO
160.9 SS1 MetroCoop 1024x768 YES
86 SS2 GrandCath 1280x1024 NO
85.1 SS2 GrandCath 1280x1024 YES
53.4 SS2 Valley 1280x1024 NO
53.3 SS2 Valley 1280x1024 YES
103.3 SS2 GrandCath 1024x768 NO
100.2 SS2 GrandCath 1024x768 YES
70.1 SS2 Valley 1024x768 NO
69.6 SS2 Valley 1024x768 YES

16597 3Dmark2k1
5794 3Dmark2k3
___________________________________________________________________
 
________________________________________________________________
PM 2.14 gig Dothan (16x133) 133 MHZ DDR single channel

AVG FPS Engine Demo Resolution Sound

775 Quake1 Bigass1 640x480
355 Quake2 Crusher 640x480
555 Quake2 Massive1 640x480
129.4 SS1 MetroCoop 1280x1024 NO
129.2 SS1 MetroCoop 1280x1024 YES
176.6 SS1 MetroCoop 1024x768 NO
175.7 SS1 MetroCoop 1024x768 YES
94.2 SS2 GrandCath 1280x1024 NO
90.5 SS2 GrandCath 1280x1024 YES
57.3 SS2 Valley 1280x1024 NO
55.4 SS2 Valley 1280x1024 YES
123.6 SS2 GrandCath 1024x768 NO
114.4 SS2 GrandCath 1024x768 YES
75.5 SS2 Valley 1024x768 NO
74.1 SS2 Valley 1024x768 YES

19445 3Dmark2k1
6031 3Dmark2k3
___________________________________________________________________
 
Subjective Gaming:

The Dothan provides higher frame rates and smoother game play in Half-Life 2.


Overclocking is painless and as you can see this chips overclocks well a 514 Mhz overclock is nothing to sneeze at as you can see in the raw cpu numbers this chip is a p4 killer and should rival athlon fx in gaming prowess. Hell if a 1.6 gig Pentium M can equal my 3.06 gig you can bet that when overclocked this chip is a hell of a desktop cpu let alone a mobile cpu. The only thing holding this cpu back is the lack of platform choices and lack of full size motherboards. If anything that is my biggest issue with this board I happen to love lots of expansion capabilities and while it may be adequate for some it leaves me wanting. The pci-x slot is an odd animal. Only newer pci cards will fit in the dual keyed slot although it is backwards compatible. A quirk I never could fix was my ata 133 controller would function properly but would refuse to detect any drive attached to it no matter its configuration. I moved it to a regular pci slot and viola it detects perfectly. The majority of all my pci cards are all older single keyed variety so essentially I’m limited to 2 pci slots. Another peculiarity of this board which even with the latest bios is not “fixed”, is the fact that I cannot get any of the other memory dividers in the bios to be stable at overclocked speeds. Plus even though I have some corsair value select ddr 400 dimms in this board I cannot get it stable past 133 FSB. The chip I am pretty sure can go higher and the ram had better be able to so I guess it’s a bios or chipset thing. One word about temps they are insane. The chip idles at 23-25c tops out at around 40c even overclocked. The chip has prime95ed for 17 hours straight so I know it will run all night and day at 2.14 at 133 FSB. I do kind of miss hyperthreading but the advanced power saving techniques and unlocked multipliers make up for it.
 
yea i saw that thread i realize my box is not 100% tuned yet but i have simply run out of time to do it. After work i have very little time during the week and my weekends are devoted to anime watching so i realize sacrafices have to be made. The performance i got now is by no means unacceptable plus until i get a better video card more clockspeed will not help as mush as a video card upgrade. I wonder if ati will release any more aiws for agp? i hope so. :)
 
this information is so exciting that it makes me weak in the knees... seriously.
Thank you for being the subjective guinea pigs for this little experiment, I would grant you my life, but I kinda need it... on occasion. :cool:
 
New bios for Aopen 1.07 ..flashed it straight away and it seems to work ok.

Did the Vdimm mod so now i could free the other mem slot which was formerly occupied by DDR Booster. Now running 2 X 512MB OCZ3700EB CL2,5-2-2-5 on the board. Divider 3:4, FSB 179MHz MEM 239MHz, Multi X16 = 2690MHz. 512MB for HL2 wasn´t clearly enough but now this thing flies. Tomorrow will try some KHX BH5 to get CL2 and more performance :cool:
 
You must log in or register to reply here.
Back
Top