24 CPU power phases and SATA III (GIGABYTE P55 mobo)

[DeltZ]

L = N x phi/ I
L = NBA/I to an approximation
= number of turns on inductor (toroid) x magnetic flux density due to each turn (U0UrNI/2pi x radius of toroid) divided by the current passing through each coil.



Therefore...more inductors means more awesomeness?

I'm so confused.

Edit: though it is interesting to see inductors is a function of the the square of the number of turns on the toroid (choke)

[Macadamia]

that! Bring back ing parallel interface!!

that too! I want my ISA slots back

[Blacky]

I was wondering why so much phases for i5...isn't this an overkill? is ironic how their i7 models doesn't have this amount of phases, an i7 would need this more than i5 imo, I could be wrong tho.

[570091D]

I was wondering why so much phases for i5...isn't this an overkill? is ironic how their i7 models doesn't have this amount of phases, an i7 would need this more than i5 imo, I could be wrong tho.

i don't think you're wrong man, the i5 should have a lower tdp, so all this extra crap around the socket is just a waste of money. and that's sad, it feels like gigabyte is trying to rip us off with all these "benefits" that add cost but not performance. some of their boards are rippin' fast, but i don't think that has to do with the virtual phases or the extra copper in the mobo. good engineering and electrical efficiency is what makes a great board. of all the mobo manufacturers, i think evga is headed in the right direction.

[Raja@ASUS]

L = N x phi/ I
L = NBA/I to an approximation
= number of turns on inductor (toroid) x magnetic flux density due to each turn (U0UrNI/2pi x radius of toroid) divided by the current passing through each coil.



Therefore...more inductors means more awesomeness?

I'm so confused.

Edit: though it is interesting to see inductors is a function of the the square of the number of turns on the toroid (choke)

that edit line say it all - you can achieve the same in a single package, which is what most other manufacs do. Might be some small currnet handling/ DCR gains in using 2 chokes, but that would still not make anything one could call a virtual phase.

[Richard Dower]

L = N x phi/ I
L = NBA/I to an approximation
= number of turns on inductor (toroid) x magnetic flux density due to each turn (U0UrNI/2pi x radius of toroid) divided by the current passing through each coil.



Therefore...more inductors means more awesomeness?

I'm so confused.

Edit: though it is interesting to see inductors is a function of the the square of the number of turns on the toroid (choke)

I didn't understand any of that - -

[Richard Dower]

Personally i am waiting for P57, as am i right in assuming this will have support for 32nm CPU?

[Shintai]

Personally i am waiting for P57, as am i right in assuming this will have support for 32nm CPU?

P55 also supports 32nm. P57 just got 3 new features P55 doesnt. (One of them is onboard flash).

PCH southbridges also doesnt matter about what CPU type. They all use DMI. You cant compare it to the regular old chipsets.

X58 also supports 32nm btw.

[Chrono Detector]

As usual Gigabyte always comes up with the nicest looking boards with plenty of rich features. Liking this board. Couldn't care less about the colour scheme, I'll still get it.

[saaya]

24 chokes = higher clocks/stability/loads

24 beefy chokes @ that

and you base that on what?
can you show me some results?

Can someone explain then, with all the "future proofing" WHY Gigabyte choose not to adopt NEC USB 3.0 chipset that launched recently?....wouldn't that be a HUGE PR, marketing, sales, fanboy coup for them?!

actually... no... that would be a useful and nice feature and really make it stand out! good point really, i hope at least some board makers will add usb 3.0, if its really as fast as its supposed to be, then itll be a very nice feature to have!

As usual Gigabyte always comes up with the nicest looking boards

Couldn't care less about the colour scheme, I'll still get it.

thats kinda contradicting... kinda...
its like a color blind saying he loves red roses... eh huh?

[NapalmV5]

and you base that on what?
can you show me some results?

lol gosh saaya i hope you dont think im basing that on a dream i had the other day..

i got ep45/q9650 results all over xs

ep45 vs ep45: i can tell you without a doubt the inductance/capacitance of ep45-extreme/dq6 easily beats "2 oz copper pcb" of ep45-ud3p @ high loads

the gigabyte design does wonders for air cooling.. even water but if we'd all live in a -100C world.. then no one would give so much crap about phase/choke/cap/2 oz copper pcb/etc.

temps vs components

[saaya]

so if we were to remove half the chokes, and only have 1 per phase... then what?
vcore would be less stable at high settings so we would get worse cpu overclocks at the same vcore, right?

[Raja@ASUS]

so if we were to remove half the chokes, and only have 1 per phase... then what?
vcore would be less stable at high settings so we would get worse cpu overclocks at the same vcore, right?

Lol. I think the 2 boards mentioned by Napalm-5 differ due to the number of physical phases rather than the chokes..

[A2C]

Lol 24 CPU power phases are too much, no? maybee been better 8 digital phases no?

[Hornet331]

actually... no... that would be a useful and nice feature and really make it stand out! good point really, i hope at least some board makers will add usb 3.0, if its really as fast as its supposed to be, then itll be a very nice feature to have!

same problem as with sata3, before the both are mainstream i have gone at least thorught another boards, so i could care less if the p55 boards have em or not.

[NapalmV5]

so if we were to remove half the chokes, and only have 1 per phase... then what?
vcore would be less stable at high settings so we would get worse cpu overclocks at the same vcore, right?

1 choke per phase is beefy/efficient enough.. fine by me.. as long as it keeps temps low @ loads on air cooling

get my drift?

i dont care what design it is as long as it keeps high clocks/loads stable @ air cooling

even crapy designs @ subzero temps perform

[Glow9]

Can someone explain the point of P55 and their processor buddies to me again? It's not like i7's are that big of a leap from current processors why stick something in between? This seriously still seems stupid to me.

[570091D]

Lol 24 CPU power phases are too much, no? maybee been better 8 digital phases no?

i agree, 8 digital phases would be better. take up less board space, probably cost less, but would produce more heat. the point is, the core i5 probably won't justify any of this. it might be capable of some crazy clock speeds, but intel isn't going to let the i5 cannibalize i7 sales and performance.

[Caveman787]

I think they don't have a choice. If the lessen the performance or overclockability just to make it not compete with it's other products when overclocked they would hurt their business even more.

I believe that happened before and their was backlash from the enthusiast/ overclocking community. But i5 is the more mainstream cheaper chip and budget people aren't as likely to be included in the power overclocking crowd as it's been stated before were still only a small portion of overall sales.

[saaya]

same problem as with sata3, before the both are mainstream i have gone at least thorught another boards, so i could care less if the p55 boards have em or not.

yeah but i think a lot of people would care... even if you plan to sell the board again in 6 months or 12 months, if it has usb3.0 and its mainstream by then, or its about to become mainstream, then youll get more money for the board than other p55 boards cause its already usb3.0...

think about 775 boards with and without 45nm cpu support...
sure, back then it didnt really matter cause once 45nm cpus come out most people here would buy a new board anyhow, but the boards that supported 45nm could be sold 2nd hand much better

1 choke per phase is beefy/efficient enough.. fine by me.. as long as it keeps temps low @ loads on air cooling

get my drift?

i dont care what design it is as long as it keeps high clocks/loads stable @ air cooling

even crapy designs @ subzero temps perform

hmmmm not really...
FlamingBlade has a more basic pwm than BloodRage, still beefy, but more basic, and cheaper... the same cpu that did 5.3 on BR only did 4.9 on FB... on ln2... so theres def a diference in pwms when you go subzero... i think on air theres not such a big diference between good and bad pwms, but once you go water or ln2, theres def a diference.

and i dont think adding chokes helps to improve a pwm... i never heard that

[NapalmV5]

of course component specs have a lot to do with the way pwms perform

when it comes to loads the 4.9 vs 5.3 could mean nothing

no matter how you slice/dice it.. higher inductance/capacitance + higher quality of that inductance/capacitance will always offer higher loads

[saaya]

sure, i just dont see how extra chokes help, unless you use crappy chokes, then doubling them might boost the pwm efficiency and output, but thats more like fixing a limited design than improving a good design... so again, i dont know much about pwms but worked with people who do, and all im saying is, i never ever heard of adding extra chokes to boost performance, efficiency or stability

so im rather sceptical about this... it looks like a cheap attempt at fooling average end-users into believing the board has loads of phases and hence is great...

[NapalmV5]

pit your flamingblade against giga-ex58-extreme

@ same overclock/cooling/cpu/ram/video card/audio card/pcie raid controller/storage system/psu/os

just by looking at pics of the two mobos i bet you this.. your flamingblade will cry for mercy

add another video card for sli/xfire and your flamingblade will curse you

prove me wrong and ill buy your flamingblade

[nfm]

I didn't understand any of that - -

Physics II .
24 chokes? They better be shielded well if they are scattered around the board to avoid crosstalk since they store energy in magnetic field around them. Just like raju said, this is purely marketing BS. The more chokes, the more ringing caused from LC circuit since it tunes to specified frequency. But generally, more chokes should distribute the heat better around the board and more inductance is never bad, it resists they change in current over time (RL circuit), so it should help with voltage spikes/drops, but again all of this could be accomplished with 6-phases.

[saaya]

pit your flamingblade against giga-ex58-extreme

do you even know what board im talking about?
FB is a sub 200$ board, the ex58 extreme is a 300$ board... and i dont work for foxconn anymore, and i dont want to pimp their boards, i just mentioned it compared to the BR since thats two diferent pwms to compare.

just by looking at pics of the two mobos i bet you this.. your flamingblade will cry for mercy

no offense, but the ex58 extreme doesnt look that nice... at all...

add another video card for sli/xfire and your flamingblade will curse you

pf, whatever man... your apparently biased way beyond reasoning, its no use trying to talk to you...

prove me wrong and ill buy your flamingblade

its not mine, and i dont care what board you buy... if you would buy it and then post comments about FB like you post comments about the gigabyte board now id rather you stick to the board you have now and dont talk nonsense about a board in which dev i was involved with