Originally Posted by Svarog
Quoting myself here.
I gave F5 another try and set the PCI-E to Gen2, so far i havn't had a Black Screen issue anymore.
So whatever changed between F4 and F5 it doesn't allow me to run Gen3 with my current Videocard.
Atleast it works.
I told this to the engineer, I am able to run 3.0 mode but I am using a AMD GPU 7970, I have a GTX670 coming from RMA tho.
Originally Posted by SonDa5
Still working on it.
In theory I should be able to. I'm sure the lower temps have helped but a load is still a load and digital power control programmed to perform X voltage based on load is stil going to need the same voltage.
Originally Posted by SonDa5
I don't think I have come to the wrong conclusion. Why would Gigabyte visibly put the 60amp rated ultra durable chokes with the neat looking lightning bolts visible to the eye then hide the VRM components underneath the heat sinks and then call this MB a "OC" series mother board using VRM parts found on much less expensive Z87-UD3 board with 8 phases and then the Z87X-UD4 board with 16 phases of the same IR3553M part?
To me it is all marketing designed to sell a board marketed as an enthusiast level over clocking board to make the most money possible.
I really think Gigabyte cut big corners on cheapening the VRM and offering minimal features with the Z87X-OC. Seems crystal clear to me that I am at the true conclusion.
I haven't seen any great reviews for the Z87X-OC or the Z87X-UD3H and they both have the same CPU VRM phase part and count of 8 of these IR3553M parts.
Those 7GHZ over clocks with this board really don't count because that is with xtreme cold cooling which eliminates a great amount of need for high quality digital power control.
I'm stuck with the board for the time being and the only other boards that I think would be a good replacement would be an MSI Z87 Mpower MAX or an
Asus MAXIMUS VI HERO.
Of course I will still have fun tweaking this MB and spend more time tweaking but I can't get over the physical limitations of the weaker IR3553m part used instead of the IR3550 part used and the idea that I paid a premium price for this mother board with this part that is found on mother boards that cost alot less money.
This was my first Gigabyte MB and it hasn't been a good experience.
Good reviews for the OCing of the OC board:
(even tho he thinks the color scheme and lack of 3 and 4-way SLI support at $200 is an issue lol)
Let me say something, as an electrical engineer who has been studying motherboard VRMs since X58 days which fascinated me so much that is what I am going to work on for my masters thesis, I was like you at some point, I didn't understand the VRM well enough to actually make good conclusions based on it, but the padwan becomes the master over time. Just don't read reviews for VRM information unless they actually know what they are talking about, AKA read design guides, read thesis, read other things. here are some links:
(i can't believe they gave him a degree without having him take English first)
or stuff from my site or youtubre channel.
I have maybe 3-4 offers from huge review sites to just do VRM explanations/IC reports on the boards for a reason, that is because I am getting my masters in electrical engineering, you would think I would know what i am talking about. The VRM is fascinating, but you come to one conclusion soon, that is that heat is the issue. However if you aren't able to increase your CPU OC by cooling your VRM with a legit water block like the one you have then you cannot come to the conclusion they gave you cheap parts. The VRM is NOT throttling the CPU or keeping it behind on the OC board, heck I bet if you ran without a heatsink it wouldn't either. BTW setting the VRm to extreme perf mode will disable any temperature compensation, that means that the digital PWM will operate in a mode where it provides the most and cleanest power regardless of its powerstage tempearture which can easily go into the 100s and be considered normal. 65C or 70C or 80C is awesome but mostly due to the IR3553 package and design which is based on directFET technology when it comes to body conduction between high and low side FETs as well as uses other things such as gold internally for gate driver connections to reduce driver losses. They are quite amazing. Whatis more amazing is the size compare, IR3550 can do 60A and 50A at 90%, IR3553 can do 40A and 25A at 90%. Yet the peak is higher on the little one b/c ithas that basic technology in it.
IR3550 won an award for actually being able to output its rated 60A spec with a heatsink on it, no one else could come close so IR charges whatever the heck they want. I have met the VP of the Power division from IR and they are nuts about this crap.
If you change to a Hero or an Extreme you will get that same as you say "false marketing with 60A chokes" as ASUS is doing this as well, they are giving higher current capacity chokes than their D87350 can provide which is 25A continous and 40A peak, same as IR3553. They also use IR3563B as the 8 phase PWM, and they also use 60A choke with a 40A/25A part, same as GBT is doing. It is not a crime to put a higher rated choke on a board, in fact is it the common practice to do so as the choke is the last bottleneck of CPU current.
Look at all the great boards, and you will see 1 thing in common, Choke power ratings NEVER match output power current of the VRM.
Take a look at the Ramapge III Extreme, it used 35A chokes with DirecFETs capable of 50A, in that case 60A chokes would have helped but no one has used anything higher than 40A chokes until GBT came out with 50A ones for the X58 OC board and these 60A ones with Z77.
Take a look at the famous "server" Volterra grade 10 phase VRMs found on the X58 classified, the rating on the power stages are 45A and the inductor/choke is 40A, this is a good design, but the current is also limited by chokes.
You can look at the Z77X-UP5 TH and UP4 TH it is the only board I have seen that could be capable of 60A output at the power stages and 60A output at the inductor, the closest match.
ASRock's Z87M-OC Formula uses the 35A chokes with the same NextFet(40A peak/25A continuous=same ratings as IR3553 but not including driver losses).
Also these higher quality chokes with higher current capacity will run cooler and provide better ripple current than if they used 35A chokes with the IR3553.
The original X58A-OC used 12 phases of some 35A DrMOS which are not as good as IR3553, and it used 50A chokes which at that time was the highest choke output on a motherboard.
For $200 you really can't ask for much more man, no one is going to give you a board with 8 IR3550 for $200.
Also server parts are usually always the best, especially when it comes to VRM. People loved that volterra 10 phase VRM on the X58 classified, you find that really only on servers in lower phase count now.
When it comes to power density you can't beat IR3550, its impossible to do what it does in the space it does with its power loss numbers, b/c of that IR can charge whatever the heck they want, even $5 a pop if they chose too, while they could sell that IR3553 which has the NexFET as a competitor for probably much less.Edited by Sin0822 - 7/31/13 at 10:07am