PDA

View Full Version : Serial ATA + AMD K8 Pics


TooMuchCoffeeMan
04-06-2002, 12:59 PM
Well, these are the first I have seen so I thought I better hook you guys up too.

http://idrinktoomuchespressoand.ismoketoomuch.com/cpxsata1_optimised.jpg

The rest can be found here (http://forums.hardwarezone.com/showthread.php?s=&threadid=234643).

I.R
04-06-2002, 02:39 PM
Wow look at the "branding" on that chip :p kind makes all the other chips other there look like crap ;)

Nice to see that the mounting holes are still there :D

Wibber
04-06-2002, 02:40 PM
YAY!

_N_
04-06-2002, 03:35 PM
excellent,
nice pics of that ASUS A7V8V board

about time, anyone know how much speed increase is there?
will there be more than 7200rpm Serial ATA HDD's?

I can't believe they got socket 754 up and running, seems like a dream.

Gh0s7 L3mUr
04-06-2002, 05:44 PM
*drools* :p

I think I'd better start saving. :D

Geek4Life
04-06-2002, 06:03 PM
I don't think the speed increase will be all that great. Untill they speed up the actual hard drives them selves.

Viper_NZ
04-06-2002, 07:15 PM
mmmm... I want one.....

It's not the speed increase of Serial ATA that interests me.. It's those thin cables :D

KingJackal
04-06-2002, 07:25 PM
Yeah - serial ATA isn't driven by SPEED, but rather by AIRFLOW.

Most of the large OEM's ( HP, IBM, etc ) were very keen to see ribbon cables dissapear out of their systems. Cos more airflow for them = machines that don't die so fast ( = less RMA's = more $$$s ;) ).

As for the board - WTF?!! Only 3 DIMM slots? Aye? This thing is 64 bit people. As in 'I can use ub3r amounts of RAM' bit processing.

HELLO???

Anyone hooommmeeee??

Geez - the AMD showoff board at was it Computex ( ? ) had 8 DIMM slots. Now THAT's a board :p :cool:

[H]arls
04-06-2002, 08:11 PM
Very cool hardware, be even better if they got rid of the old I/O ports on that mobo though.
I understand it would be a stupid business decision to build a board with only serial ATA + legacy free ports at this point in time, (seeing as theres bugger all ATA drives out there) but still, we need someone to lead the pack. :)

Humantuckshop
04-06-2002, 08:23 PM
Originally posted by KingJackal

As for the board - WTF?!! Only 3 DIMM slots? Aye? This thing is 64 bit people. As in 'I can use ub3r amounts of RAM' bit processing.

HELLO???

Anyone hooommmeeee??

Geez - the AMD showoff board at was it Computex ( ? ) had 8 DIMM slots. Now THAT's a board :p :cool:



Well, it is an ASUS, not an Epox or Abit - also check out the 2 phase power supply on that POS. :mad: :(

_N_
04-06-2002, 09:04 PM
hey don't dis the 2 phase power on asus,
if it wasn't for 2 phase power on my A7V266E, i wouldn't be able to use my 5 year old 250W ATX PSU. :P :D

BTW
Tried using a 300W but the V were lower on it, so don't bother trying to get a better PSU.

Viper_NZ
04-06-2002, 09:15 PM
You can't diss Asus's 2 phase power. It's actually damn good...

Remember you can have good 2 phase, or crap 3 phase, and the 2 phase will be better...

Humantuckshop
04-06-2002, 09:32 PM
I thought 3 phase power was supposed to be the standard now tho...... I always blamed ASUS's 2 phase power for the crap OC'ing with my old A7M266.

KingJackal
04-06-2002, 10:05 PM
Originally posted by Viper_NZ
You can't diss Asus's 2 phase power. It's actually damn good...

Remember you can have good 2 phase, or crap 3 phase, and the 2 phase will be better...

<JUSTTOANNOYVIPER>
But what about GOOD 3-PHASE power?
*innocent look*
</JUSTTOANNOYVIPER>

Deviant
04-06-2002, 10:14 PM
Ok teach me somthing new? What's the diff between 2 phase and 3 phase power on a board? How can you tell?

Humantuckshop
04-06-2002, 10:31 PM
OK, here is a picture of a 2 phase mobo:

Humantuckshop
04-06-2002, 10:37 PM
And a 3 phase mobo:

Viper_NZ
04-06-2002, 10:40 PM
That's not ALWAYS true... You can get 3 phase power with only 2 sets of mosfets

But it's usually a damn good indication

Humantuckshop
04-06-2002, 10:48 PM
Yep that's true, you can but that would be crappy 3 Phase power you referred to in your earlier post, would it not?

swiftynz
04-06-2002, 11:05 PM
so does that mean that the GA7DXR+ is a 3-phase?
http://www.gigabyte.com.tw/products/img/7dxr+.jpg
sorry, i'm just not sure what a mosfet is....:(

/edit
(btw :( i reckon is the clostest thing to :I from the old forum)

Humantuckshop
05-06-2002, 12:24 AM
Yup that mobo is definitely 3 phase. I have crudely concocted another picture (based on the 7DXR+) to show what a MOSFET is:

Deviant
05-06-2002, 05:12 PM
how do you know they are MOSFETS and not Voltage regulators?

I recon ASUS would design better with 2 stages than some mobo manufacturers with 3 stages. Anyway I still don't unserstand, They aren't phases, they are probably 1.75, 3.3V and 5V, so where does the phases thing come from?

dumass
05-06-2002, 05:41 PM
Well, atleast the serial cables look tidy, no more rounded IDE cables :)

Humantuckshop
05-06-2002, 05:58 PM
Originally posted by Deviant
how do you know they are MOSFETS and not Voltage regulators?

I recon ASUS would design better with 2 stages than some mobo manufacturers with 3 stages. Anyway I still don't unserstand, They aren't phases, they are probably 1.75, 3.3V and 5V, so where does the phases thing come from?


Because they are MOSFETS. That's what they are......

The stages are more to do with current than voltage rails. Most newer CPU's draw 50 amps of current or more, so therefore a 3 phase power should be used.
2 phase power should only be used with CPU's that draw a current of 40A or less.

Hans' review of the MSI K7T266 Pro2 has a bit about it: http://www.overclockers.co.nz/ocnz/mobo/k7t266pro2/2.shtml

mird-OC
05-06-2002, 07:06 PM
yeah like tuckshop said, it's about the current... it just depends on the requirements for the CPU, and the limits of the FETs used. so a 2-phase circuit using high quality FETs can often be better than a 3-phase circuit using low quality FETs.

on another note - it's good to see ASUS are still supporting the socket A mounting holes :D

Deviant
05-06-2002, 08:45 PM
Have you ever seen a 1000 amp mosfet? Yes they exist, they don't use 20 in series, the use one power one.

You do not need 2 or 3 stages to supply 50amps, but you do need different stages to supply different voltages.

Do you even know what 2 phase and 3 phase power is? I seriously doubt it!

How do you know they are Mosfets? More likely Voltage regulators in my book. The only way you can set different voltages with a Mosfet is in a switching mode configuration, like a switch mode power supply (IMO), this is unlikely as those inductors look like filtering inductors, not the big inductors required to produce the currents you guys are talking about. Do you know what a Mosfet is and how they work?

I think you guys have been misled, or don't really know what they are. I will look into this more closely to see if I can come up with any answers.

2 phase for 40a, and 3 phase for 50a is definately wrong in my book.

*waits for KJ's 0.02c*

Deviant
05-06-2002, 09:05 PM
have a read of this:
http://www.pcguide.com/ref/mbsys/mobo/comp.htm

He seems to agree with me, that they are voltage regulators. Although old pentium mobo, same principles apply, click on the voltage regulator link on the page to find out more. Backs my argument up, but please find more if you think I am wrong.

_N_
05-06-2002, 09:32 PM
ok, looks like back to basics,

first of all, we all heard of phase shift?
when you have a C or L in an AC circuit? you get phaseshift.

This is US based, they have 120V 60Hz AC.

No matter what it is that spins the generator, all commercial electrical generators of any size generate what is called 3-phase AC power. To understand 3-phase AC power it is helpful to understand single-phase power first. Single-phase power is what you have in your house. You generally talk about household electrical service as single-phase 120 volt AC service. If you use an oscilloscope and look at the power found at a normal wall-plate outlet in your house, what you will find is that the power at the wall plate looks like a sine wave, and that wave oscillates between -170 volts and 170 volts (the peaks are indeed at 170 volts; it is the average (rms) voltage that is 120 volts). The rate of oscillation for the sine wave is 60 cycles per second. Oscillating power like this is generally referred to as AC, or Alternating Current. .

The power plant produces AC. However, it produces three different phases of power simultaneously, and the three phases are offset 120 degrees from each other. Out of every power plant come four wires: the three phases plus a neutral or ground common to all three.

There is nothing special or magical about 3-phase power. It is simply three single phases synchronized and offset by 120 degrees.

Why three phases? Why not one or two or four? One big advantage that 3-phase power has over 1-phase or 2-phase power is the fact that, at any given moment, one of the three phases is nearing a peak. In 1-phase and 2-phase power there are 120 moments per second when the sine wave(s) cross zero volts. High-power 3-phase motors (used in industrial applications) and things such as 3-phase welding equipment therefore have even power output. Four phases would not significantly improve things but would add a fourth wire, so 3-phase is the natural settling point.

So there you go, If you looked at a proper power cable (I did this this morning and asked the lecturer why they have 4 wires, and like it says above, 3 for 3 phase power, and one ground).


So now I hope everyone know's what 3 phase power is.

theYANK
05-06-2002, 09:54 PM
Holee molee, I can actually feel the neurons puuting out new pathways....
And I can see I got 3-phase power now! w00t!

mird-OC
05-06-2002, 10:05 PM
it's come down to economics and common sense. YES there are big MOSFETs but they're not suitable for motherboards given the voltages, the price, and the effects of running big FETs.

i can tell you right now, it's not about the different voltages... voltages such as VCORE, VIO, VAGP, VMEM etc are all output via PMW regulators - have a look on a motherboard and you'll see little PMW regulator ICs next to the CPU socket, the RAM, the AGP port, etc etc. i know this, because i mod this :)

and 3-phase AC is a totally different kettle of fish. NOT THE SAME THING.

theYANK
05-06-2002, 10:10 PM
Originally posted by mird-OC
...and 3-phase AC is a totally different kettle of fish. NOT THE SAME THING.
does that mean i was just confused?

Deviant
05-06-2002, 10:20 PM
Price and economics you say. Well one 50 amp Mosfet is cheaper than 3x20 amp ones!

"voltages such as VCORE, VIO, VAGP, VMEM etc are all output via PMW regulators" Those little chips can't do that by them selves, the need voltage regulators, which is what those funny things are circled in the pictures.

What kind of new 3 phase power have you invented? It being DC and all. What is this 3 phase you are talking about?

Humantuckshop
06-06-2002, 01:12 AM
Seems that no one really knows what a VRM is. Here is a picture of what one website calls a VRM, quite a bit different from a MOSFET, aye? What is going on - is there no industry standard for this or something?

Humantuckshop
06-06-2002, 01:18 AM
And I really don't care either, the little black chips have always been called MOSFET's as far as I am aware, and every other computer phr33k I know call them MOSFET's aswell - I will continue to call them MOSFET's, basically because I always have, and I'm not going to waste any more of my time looking that crap up on google - life is short.

Cheers. :D

mird-OC
06-06-2002, 10:05 AM
Originally posted by Deviant
Price and economics you say. Well one 50 amp Mosfet is cheaper than 3x20 amp ones!
thank you for quoting me out of context. if you go back and re-read my post i stated three reason why they use multi-phase power circuits. notice how i said the effects? - the main one being heat. show me a big FET capable of supplying the same amount of current as a 3-phase circuit that runs cooler and is cheaper. you won't find one.

here's some literature for you:

ABIT First to Implement "3-Phase" Motherboard Power (http://www.duxcw.com/news/releases/abit/3phase.htm)

The 1GHz “Classic” Athlon alone requires 37 amps under continuous heavy loading. With Thunderbirds, those demands have increased: The 1.2GHz model requires a maximum current of 39A, the 1.3GHz requires 41A, the 1.4GHz 43A, and the 1.5GHz requires an unprecedented 46A! ...

In order to support these high frequency CPUs, a motherboard must be able to handle 46 Amps, or the system will not be able to run stably or may be subject to heat damage. Motherboards handle current through synchronous pulse width modulation DC to DC conversion circuits. These circuits fall into two categories: single-phase and multi-phase. Single phase power solutions were designed for current up to 20 Amp, Multi-phase power solutions, which utilize MOSFE transistors to share the current loading, can handle up to 40 Amps. Therefore ABIT has gone one step further to innovate 3-Phase Power on the KT7 and KT7-RAID for 1.3GHz+ Thunderbirds.

At 40A, ABIT’s 3-Phase Power stays a cool 70° Celsius while the transistors on other motherboards climb to 90°. At 46Amps, ABIT’s 3-Phase Power remains within the bounds of safety at 88° while other power solutions hit 120°and higher! At 120°, not only do you have an excess of heat in your case; you also risk damaging the mainboard’s PCB, destroying the transistors (rendering the board useless), or even damaging the CPU.

happy now?

Deviant
06-06-2002, 08:28 PM
Thanks for the good link Mird. It clears alot up, like when you said it's not that knid of 3 phase power. It makes more sense now.

They do say you need 3 phase, cause 2 phase will get too hot, but that would depend on the rating of the MOSFETs. If they get too hot, then you can put a bigger one in or use a bigger heatsink or both. That may be what other mobo manufacturers do, but abit obviously decided to do it this way for whatever reason.