P67A-UD7 Most In-Depth Preview/Review!!

[bingo13]

hahah Gary you sound stressed mate

I am not stressed, just not going to lay down on this one after getting a shot across the bow. If he had actually tested the board before posting up some of "his" (well we know where they came from) conclusions or reworded the innuendos (FUD) or false statements about our design then we would not have a problem here.

[massman]

I do agree with what you are saying, yes extreme cooling is required to be competitive in crowds that HWbot caters to. But HWBot isn't the majority of people who buy the ROG series or the highend BigBAng or UD7,UD9 boards, that number is the real question here. Is it worth it to just market a board to the extreme users?

Excellent remark! This is where we leave planet 'ratio and logic' and enter planet 'marketing'.

From my point of view, I see that GBT has been one of the first to acknowledge the power enthusiast community as market-leading group. Although the extreme overclocking community is a niche market, it has an influence on other parts of the IT market such as gamers and high-end PC enthusiasts. I would in fact go this far and state that the overclocking community has the ability to make or break a product, although that would require a very large amount of people to spread a deeply negative opinion. Nevertheless, for manufacturers it's not a bad move to invest in the quality of their most high-end products to make sure the enthusiast market approves or okay's it, the difference being that approving might create a hype and okay'ing will avoid public dissing.

Examples can be found quite easily. DFI's NF2/4 series were legendary and, although being delivered with a BIOS crammed with hard-to-understand options even the gamers and occasional overclockers bought those. A more recent example would be the Classified series; an excellent product also delivered with that massive BIOS of which only the technical people know what the settings do. Nevertheless, a lot of people purchased those boards and if it wasn't for the poor delivery (at least here in EU), there would have been a lot more. I recently dugg up some stats on mainboards on HWBOT and you can clearly see the Classified being used more than other boards. The same goes for for graphics cards, by the way. The Lightning series have been doing so well over the past generations (let's say from GTX275) that even the 'normal' people, who will never use liquid nitrogen or even change the stock cooler, are now looking forward to the next release of this graphics card. As a sidenote, HWBOT statistics show that in January MSI was the 4th most used VGA brand in the current-gen market (NV400/500 and HD5000/6000). Since June they are first. Yes, beating ASUS and all that.

Extreme overclockers (LN2) influence die-hard overclockers (SS to air, but still pushing for limit), who influence occasional overclockers (running 24/7 OC and sometimes pushing higher just to see what happens), who influence performance enthusiasts (not overclocking, but buying high-end hardware), who influence the mainstream audience (not high-end, but still want quality).

Having that established, we need to understand that vendors are companies that want to make profit. So, they will focus on the needs of the extreme overclockers to make them happy, but also need to make sure that every single penny put into research has to be earned back. Given that hypes are very often based on single products, it makes sense for vendors to put all their best stuff in the most high-end product and label it 'best of best'. If it's good, extreme overclockers will jump on it; occasional overclockers might follow; performance enthusiasts perhaps go for the product just below that and the mainstream audience will figure 'that brand is very good'. Imho, having a very clear product line helps here; UD9-UD7-UD6-UD5-... people know exactly where the product they are buying is located within the vendor's product line.

Like you mentioned, the extreme overclockers require a whole different board that what we're getting nowadays. Couple of USBs, PS/2, one PATA, 3x DIMM, etc. At first sight, it would be quite difficult for vendors to sell this to the mass-market. At least, if you still think in terms of the overclocking community being a part of the gaming community (which it certainly was at a certain point). If you'd ask me, I'd place the group of extreme overclockers more in the workstation/end-user server market as the main requirement (stability) is similar. Connectivity, which I would relate to gamers and general enthusiast market, is certainly not a concern of the extreme overclockers.

Especially the part where the overclocking community is moving away from the gaming community, as overclocking/benchmarking is becoming more and more a game in itself, is interesting. At one of the ASUS tech seminars, I tried to bring this up (again), but I'm not sure if I got my point across. The idea is simple: overclockers need stability, and hardly any connectivity. This means lower cost on peripherals, higher(?) cost on r&d. The product could very well be a small form factor mainboard with plugin connectors for USB/LAN/etc or a large XL-ATX board with 7 PCI-e lanes. In case of the small-form factor board, I would definitly argue that the market is greater than just the extreme overclocking community. In fact, I would argue that a lot of people would be interested in a to-the-point product with keywords 'stability', 'flexibility' and 'cut-down peripheral'.

Perhaps it still is a bridge too far at the moment and vendors are afraid the ROI will not be great enough in the end. We've seen vendors come with extreme oc bioses in the recent past, though. I'll go out on a limb and state that I think there's a correlation between extreme overclocking competitions and vendor eagerness to listen to the specific requirements of extreme overclockers. By which I want to say: I guess it will take more pure marketing exposure to get those special X-OC product lines we all want

[Johnny87au]

@massman sorry im not reading that lol

[Buckeye]

Very good Massman, I think what you just wrote is pretty accurante

The Classy was a top notch board that pleased just about everyone, except the budget minded people. It really hit a sweet spot and that was a direct result of a very good designer, Shimano, who pleased everyone with that design. When you add Shimano and Kingpin together you have a marketing power house and a great product.

So much of this is marketing tho.

[KaOSoFt]

sin0822, this is one of the most thoroughly reviewed computer component I've seen so far... congratulations on that!

In fact, I've been a Gigabyte fan for years now, and right now I'm waiting precisely for these boards, so I can finally upgrade to Sandy Bridge, coming from a Q6600. I plan to get a nice rig, and sell this one. Do you know if Gigabyte will be releasing a board just like the UD7, but with only one graphics card slot? I really don't like multi-card setups, especially because I mostly cool with air, and the added noise levels are not worth to me.

In any case, without further ado, aside from the review, I was waiting for these boards, and after such a nice review, I'll be sure getting them.

[Leeghoofd]

It really hit a sweet spot and that was a direct result of a very good designer, Shimano, who pleased everyone with that design. When you add Shimano and Kingpin together you have a marketing power house and a great product.

Too bad Shimano is all into bike gear now

I really wonder how these high end SB boards will differ from the mainstream ones. As most of the OCing potential is in the CPU multiplier.. not in the board... or it's features...

ow yes well spoken Dr Massman !!

[M.Beier]

Too bad Shimano is all into bike gear now

I really wonder how these high end SB boards will differ from the mainstream ones. As most of the OCing potential is in the CPU multiplier.. not in the board... or it's features...

ow yes well spoken Dr Massman !!

This is best post on last 2 pages

Easy and simple to handle, doesnt spoil my ability to eat christmas cookies, and yet makes them taste even sweeter.

---
Guys, holy cow, what the heck happend? This was a simple preview with a bunch of pictures, and it turned into a bunch of arguementation that makes the European parlament look like rookies when it comes to stuffing in useless content in text.

[sin0822]

Thanks massman, i agree with you on your post! i am glad we are having this discussion, its something to waste time until NDA is up! hahaha. i really enjoy motherboards, and thanks everyone for you input!

[Lsdmeasap]

Awesome review sin, keep em coming!

Looking forward to your further testing!

[sin0822]

thanks man!

[Holst]

Thank you for your kind words! I do a lot of research and I went to school for this this stuff. I am not an electrical engineer but I went to school with them and I have taken a few classes. I really enjoy doing this and that is why I do it.

once again I appreciate the kind feedback!

I was very surprised with this board as well, I really like what they have done with the heatsinks and I wanted to give them a test run to see how they perform. I am confident that these heatsinks not only look very appealing they will perform up to par.

The board has been designed like a masterpiece, from the high quality VRM(uses the same driver MOSFETs as the UD9, different controller(a better one)), to the way they incorporated the NF200 chipset is just great. All these ports as well as the USB3 turbo mode are very impressive. I was looking around at other motherboards, I haven't seen one with as much connectivity and as well implemented.

I know the reviews are long, but this isn't ram we are talking about here, motherboards are as complicated as they come, and thus need to be technically evaluated.

Just wanted to say thanks for the review. Very interesting and well researched. If you get the chance to write any more I will be very interested in reading them.. manufacturers take note

It will be interesting to see if the expensive power regulation gives a good overclock. When I was with Epox boards that were expected to overclock well didnt always live up to there specs.

[sin0822]

Thank man, I will have a full performance evaluation when NDA is up as well as some other interesting articles on other hardware.

[Systemlord]

I think that this is the most well researched and most in depth review I have ever read on any forum even if it took three lunch breaks to finish, well done!

-Systemlord

[MACMAC]

If you have an SB chip on hand, find all the voltage read points, it will save me some time

Great overview by the way, but you know that already

[Andrea deluxe]

Thank man, I will have a full performance evaluation when NDA is up as well as some other interesting articles on other hardware.

from your ic's analisys........

one phase generate 57 watt


UD7 24 phase's= 1.368 watt
UD5 20 phase's= 1.140 watt
UD4 12 phase's= 684 watt


right????

[sin0822]

um more like 1 driver MOSFET can push 35amps continuos at a mximum 1.9v lets say we push into the CPU, so take amergae to 30amps to be conservative, there is only su much 12v power put into the vrm system. 30x1.9=57 watts yes, in reality the phases can push 35 amps maimum continous as thdata sheet says, and in multple phase array they can produce over 90% efficiency, there are a few more factors, but maye this would help you. Watch it until the second test they do 1.88vx700amps= 1316 watts.
http://www.youtube.com/watch?v=YsFG0Pv-7bM

[Leeghoofd]

The question is : is this all required for SB ? everything looks good on paper and on PR slides... tests will show what's right and what's not... at the moment I remain in the dark sceptical corner

[Hondacity]

um more like 1 driver MOSFET can push 35amps continuos at a mximum 1.9v lets say we push into the CPU, so take amergae to 30amps to be conservative, there is only su much 12v power put into the vrm system. 30x1.9=57 watts yes, in reality the phases can push 35 amps maimum continous as thdata sheet says, and in multple phase array they can produce over 90% efficiency, there are a few more factors, but maye this would help you. Watch it until the second test they do 1.88vx700amps= 1316 watts.
http://www.youtube.com/watch?v=YsFG0Pv-7bM

the moment he increased the load..(time1:40)the vcore 2.01v.....dropped to 1.76v....pathetic regulation lol

pretty much useless...

no cooling on the vregs at full power? i wonder what temps they were at...

[sin0822]

well yes voltage is going to drop for increase in load. in reality you could never pull that much amperage from one chip.

It is just showing max output of the board. As someone asked.

IDK how its going to pan out, all i can say is that both types of system shoudl work fine for sandy bridge as the CPU will more than likley have lower TDP and TDC so even 12 phase analogue woudl be more than enough for a single chip. So would any digital vrm system.

i really have no idea why ASUS has their panties in a knot, just a market manager wants to come and prove his VRM system is better, in transient responce, great, what about max output, efficiency and so forth, tranisent responce is just one area.

[Leeghoofd]

There's more to a board than just VRMs mate... especially in the SB category, overhyping stuff is maybe not such a good idea... only thing that interests me with these boards is the efficiency in raw performance... we got most of the boards in the house now, time to start cracking at it... gonna be some interesting weeks...

[sin0822]

yea maybe i shouldn't have talked about it so much, but fee dback I have recieved on other sites, was that people really wanted to knoe the difference between digital and analog vrm, they wanted to knwo what serial vid was and they want to know what vrd 12 certified means, whch obviously i don't know exactly, at least i tried to spell out digital versus analogue vrm. I doubt max OC will be affected by one board or another.

[Leeghoofd]

Well it's all a bit too technical for me (One of the reasons I don't really comment on the technical part of this (p)review. As I will make mistakes either way :p)

I want pretty rock solid voltages/amps, all the required settings to stabilise things and some headroom for LN2 sessions... How it's done, sweet, but I'm a (ab)user...nothing more...

[sin0822]

haha aren't we all (ab)users! hahaha I like the way you type. This board deffinitly puts out solid stable voltages and a hell of a lot of amperage, do you need ti all even under LN2? deffinitly not, is it nice to knwo you have it? yes.

I need to install spell check on this computer

[Raja@ASUS]

digital vrms are not slower at transient response or error correction than analog ones per se - it all comes down to open and closed loop performance rather than the digital "delays". Without scope shots such statements of global performance potential are grossly misleading.

[zsamz_]

alot of effort in the review i like it
but alot of mistakes
now we just need to see the gigabut in action
damn sexy board