OBR source:"Why are Gigabyte boards worst ocers then other?"

[massman]

sorry but i think everybody here missing the point. first question everybody should ask here is "if intel has a spec of 43 capacitors why gigabyte doesnt obey this spec? especially while as everybody said that components are dirty cheap for them.". yes article is aiming us to gigabyte cant overclock well idea which is completely wrong but come on this OBR leave that not fact silly ideas to them. But question still stays we should get a logical answer. "if intel has a spec of 43 capacitors why gigabyte doesnt obey this spec? especially while as everybody said that components are dirty cheap for them."

Yeah, I like my overclocking boards to follow Intel spec as closely as possible. That's why I only bench and overclock with

[kromosto]

Kromosto,

The spec does not state 43 caps, rather a nominal resistance/capacitance value as a guide not a hard set rule. It is up to the builders to design to spec or improve/tweak as they see fit. Some could argue that a 4 phase VR is just as good as a 24 phase VR, it is how the builder implements the design.

Attachment 129340

ok example is hitting the bullseye . so that part is a lie too.

[lutjens]

I think that the one thing that can be concluded from this thread is that you get what you pay for. If you buy a cheap board, you may not only get fewer features, but less overall quality (inferior design, poorer quality capacitors, etc). A more expensive board may cost more, but should be better quality. Folks should keep that fact in mind when buying a board and not make a decision based solely on a board's feature set.

[kromosto]

Yeah, I like my overclocking boards to follow Intel spec as closely as possible. That's why I only bench and overclock with

you like that skull arent you . dont get me wrong following intel specs is not important to me but if there is a spec (and Fugger said there is no spec like this so not for this example) i expect them to make it the minimum or if they lower the for example capacitor numbers then the spec for a reason then i want to learn that reason

[rintamarotta]

Hell with Asus, Gigabyte, MSI or ASRock. EVGA and Intel boards forever, if there is spec then Intel will make board to that spec and it will work flawlessly, if some intel's board have got bad overclocking options then just buying EVGA will fix it.

Cheap boards are ment only for normal usage or minimal overclocking or not for heavy overclocking atall so its understandable that it might be missing few caps here and there, but i dont need worry since im not buying any low end crap.

[lutjens]

Hell with Asus, Gigabyte, MSI or ASRock. EVGA and Intel boards forever, if there is spec then Intel will make board to that spec and it will work flawlessly, if some intel's board have got bad overclocking options then just buying EVGA will fix it.

Cheap boards are ment only for normal usage or minimal overclocking or not for heavy overclocking atall so its understandable that it might be missing few caps here and there, but i dont need worry since im not buying any low end crap.

I think the high-end versions of the boards made by most manufacturers are good. The Asus ROG and WS category boards are examples of this. It's when you get into the barebones boards with little to no features that are meant to hit a budget price point that corners begin to get cut. I'd surmise that the point at which corners begin to get cut varies by manufacturer, but it no doubt happens.

My $0.02...

[dinos22]

Hey Splave why would I take anything that OBR copies and pastes from a marketing slide seriously. I see that even on XS though are some gullible types, it's sorta funny

I cant speak for Gigabyte but I simply see this as a compliment, when your competitor feels the need to superficially attack you to score some points with the masses. People aren't stupid and this sort of thing never works because its BS but keep at it if you feel it helps you I say. Remember that everyone compares themselves to the top players..... Btw this is part of a much bigger slide that we've seen before Computex and even respectable employees of this competitor of ours were apologetic for these moronic PPTs they must be feeling the heat in that region

GUYS GO BACK TO POST#75 THAT FUGGER WROTE, intel sets a nominal capacitance value guideline and not the amount of capacitors

Thanks for clearing that up FUGGER!

[Splave]

Hey Splave why would I take anything that OBR copies and pastes from a marketing slide seriously. I see that even on XS though are some gullible types, it's sorta funny

I cant speak for Gigabyte but I simply see this as a compliment, when your competitor feels the need to superficially attack you to score some points with the masses. People aren't stupid and this sort of thing never works because its BS but keep at it if you feel it helps you I say. Remember that everyone compares themselves to the top players..... Btw this is part of a much bigger slide that we've seen before Computex and even respectable employees of this competitor of ours were apologetic for these moronic PPTs they must be feeling the heat in that region

GUYS GO BACK TO POST#75 THAT FUGGER WROTE, intel sets a nominal capacitance value guideline and not the amount of capacitors

Thanks for clearing that up FUGGER!

/thread :P that wasnt too hard was it! lmao

[Vinas]

The amount of components just means the Asus boards are less efficient IMHO. The amount of caps mean next to nothing, the magic is all in the circuit design. Some would argue that better designs need fewer components. (and before any flaming - I run a ROG board currently, but have also had DQ6 boards).

[cowie]

yawn all mb's suck never seen one i could not kill.
dont matter the price or brand

[naokaji]

There are currently 170 S1155 boards listed on the Gigabyte website, even if some of them would not be too suitable for oc'ing but are rather just designed to work fine at stock, who cares? plenty of other models to choose from.

[ice_chill]

The questions still remains and no-one has given a valid explanation:
Why is Gigabyte's design inferior to Intel Specification ???

A good manufacture creates a motherboard that meets the specifications, and then improves it. Gigabyte has Degraded the Design.
-------
Also my personal experience with Gigabyte, are that they reset the smallest overclocks (+5%) on Cold boot, and are unable to run some Linux OS's, never had any of such problems with ASUS.

[M.Beier]

The questions still remains and no-one has given a valid explanation:
Why is Gigabyte's design inferior to Intel Specification ???

A good manufacture creates a motherboard that meets the specifications, and then improves it. Gigabyte has Degraded the Design.
-------
Also my personal experience with Gigabyte, are that they reset the smallest overclocks (+5%) on Cold boot, and are unable to run some Linux OS's, never had any of such problems with ASUS.

I am confident that they have tested it, and evalutated that it wasnt needed, that there was no benefit.... Maybe I am just naive but I believe they (R&D) know exactly what they are doing, and I do not think this is to save a few cents, then it would not be applied to the UD series, but only D.....

But I am biased on this one, and dino would normally say "Marc you are talking out of your arse again!"

[Splave]

[[XC] Oj101]

FUGGER has given a valid explanation, read more and post less

[sin0822]

The questions still remains and no-one has given a valid explanation:
Why is Gigabyte's design inferior to Intel Specification ???

A good manufacture creates a motherboard that meets the specifications, and then improves it. Gigabyte has Degraded the Design.
-------
Also my personal experience with Gigabyte, are that they reset the smallest overclocks (+5%) on Cold boot, and are unable to run some Linux OS's, never had any of such problems with ASUS.

it isn't inferior in any way. It just used different way of decoupling the output. Power doesn't have to go through the capacitors to get to the CPU, the capacitors just have 1 leg in between the inductor and the CPU, and the other to the ground so that they can store energy for when it is needed. The inductors along with the capacitor bank create the output filter which job is to provide a steady stream of power to the CPU. The purpose of the output capacitor bank which are made of the can-type and the MLCC, which are being discussed, is to source and sink the output, so that if you take a 100A load step and the inductors cannot respond quick enough then the capacitors provide all the power for the transient. However during this discharge you have to make sure that the voltage drop isn't significant to deviate away from what you want. Thus you need high and low frequency decoupling, the low frequency is usually done by the can-types and the high frequency by the MLCC. The ESR of the can-types is important as is the ESL of the MLCC, however if you have very good capacitors then I could possibly see the need for some of the MLCC to go away.

Using higher quality MLCC could also reduce component count.

If GIGABYTE's design was inferior how come their LLC is so dead on? measure with a DMM, as the UD5H has points for and see for yourself. During a large load the voltage would drop significant if the CPU couldn't handle it.

[Hondacity]

Do you guys have the Intel CPU power design guide for IB?

[lutjens]

Intel does make their boards to spec and their boards are traditionally very solid and durable, with good quality parts. Their boards function well at spec and have a decent amount of headroom above the spec. Where Intel is less than ideal, is in designs that exceed the spec by an excessive amount. Many heavy duty, high end motherboards (for all manufacturers) would be considered extremely overbuilt for operation at spec. Instead, they are built for the rigors of extreme overclocking (a la LN2) in mind and are the most desirable parts, IMHO... They can take whatever is thrown at them with no issue, and their individual components end up being worked less hard when they are run at spec (which is much less than that which they are built for). Stressing components less is good for their life expectancy and makes them more reliable in the long run.

[l0ud_sil3nc3]

lol surely there must be more to this, maybe sin0822 will chime in

fixed

[Slappa]

it isn't inferior in any way. It just used different way of decoupling the output. Power doesn't have to go through the capacitors to get to the CPU, the capacitors just have 1 leg in between the inductor and the CPU, and the other to the ground so that they can store energy for when it is needed. The inductors along with the capacitor bank create the output filter which job is to provide a steady stream of power to the CPU. The purpose of the output capacitor bank which are made of the can-type and the MLCC, which are being discussed, is to source and sink the output, so that if you take a 100A load step and the inductors cannot respond quick enough then the capacitors provide all the power for the transient. However during this discharge you have to make sure that the voltage drop isn't significant to deviate away from what you want. Thus you need high and low frequency decoupling, the low frequency is usually done by the can-types and the high frequency by the MLCC. The ESR of the can-types is important as is the ESL of the MLCC, however if you have very good capacitors then I could possibly see the need for some of the MLCC to go away.

Using higher quality MLCC could also reduce component count.

If GIGABYTE's design was inferior how come their LLC is so dead on? measure with a DMM, as the UD5H has points for and see for yourself. During a large load the voltage would drop significant if the CPU couldn't handle it.

Thank you for this. It seems some people have forgotten the function of a capacitor.

[crash5s]

Intel does make their boards to spec and their boards are traditionally very solid and durable, with good quality parts. Their boards function well at spec and have a decent amount of headroom above the spec. Where Intel is less than ideal, is in designs that exceed the spec by an excessive amount. Many heavy duty, high end motherboards (for all manufacturers) would be considered extremely overbuilt for operation at spec. Instead, they are built for the rigors of extreme overclocking (a la LN2) in mind and are the most desirable parts, IMHO... They can take whatever is thrown at them with no issue, and their individual components end up being worked less hard when they are run at spec (which is much less than that which they are built for). Stressing components less is good for their life expectancy and makes them more reliable in the long run.

Malarky. Intel boards are great. Granted, they aren't really made to be OC'd, but for a board you plan to run 24/7 for years on end they do better than the competition. They aren't really making boards for people here. Intel is making workstation boards and tossing in some OC features and other stuff to make them a viable desktop purchase.

You could lob the same comment at Tyan and Supermicro, both of which make incredible products as well. Intel's "extreme" boards just seem to blur the line.

[l0ud_sil3nc3]

^nice gents

[STEvil]

it isn't inferior in any way. It just used different way of decoupling the output. Power doesn't have to go through the capacitors to get to the CPU, the capacitors just have 1 leg in between the inductor and the CPU, and the other to the ground so that they can store energy for when it is needed. The inductors along with the capacitor bank create the output filter which job is to provide a steady stream of power to the CPU. The purpose of the output capacitor bank which are made of the can-type and the MLCC, which are being discussed, is to source and sink the output, so that if you take a 100A load step and the inductors cannot respond quick enough then the capacitors provide all the power for the transient. However during this discharge you have to make sure that the voltage drop isn't significant to deviate away from what you want. Thus you need high and low frequency decoupling, the low frequency is usually done by the can-types and the high frequency by the MLCC. The ESR of the can-types is important as is the ESL of the MLCC, however if you have very good capacitors then I could possibly see the need for some of the MLCC to go away.

Using higher quality MLCC could also reduce component count.

If GIGABYTE's design was inferior how come their LLC is so dead on? measure with a DMM, as the UD5H has points for and see for yourself. During a large load the voltage would drop significant if the CPU couldn't handle it.

Yeah no kidding.. my asus board (p8p67 pro) has a disgusting ability to control CPU vcore (ie: it sucks).

[lutjens]

Malarky. Intel boards are great. Granted, they aren't really made to be OC'd, but for a board you plan to run 24/7 for years on end they do better than the competition. They aren't really making boards for people here. Intel is making workstation boards and tossing in some OC features and other stuff to make them a viable desktop purchase.

You could lob the same comment at Tyan and Supermicro, both of which make incredible products as well. Intel's "extreme" boards just seem to blur the line.

Did I not say that? What specific thing that I said can be classed as "Malarky"? I said that Intel's boards are very solid and durable for operation at spec, but they aren't overbuilt to the extent that the heavy duty boards from many major manufacturers are.

A heavy duty, overbuilt motherboard will do just as well as an Intel board when operating at spec, but since it's overbuilt (a greater number of better quality capacitors, stronger power regulation, etc) it's often superior. Mid and lower range boards definitely don't apply and are often inferior to Intel's boards as the manufacturer strives to maintain the maximum amount of features while hitting a low price point and often compromises are made to do this. I'd imagine that Intel's Extreme series boards probably go beyond spec a fair bit to ensure reliability in an environment where some overclocking (and hence greater stress) is expected, but I doubt they are as robust as an Asus Rampage IV Extreme or an EVGA X79 FTW. Intel's workstation and enterprise boards go beyond their standard desktop boards and the base spec for such boards should be significantly higher than that of a regular desktop board to ensure total reliability in all conditions.

[punx223]

why is an OBR thread still being discussed?

dont we have something better to do like sit in a corner and watch paint dry? I felt dirty just clicking the link man.

all boards as of recent overclock very close to the same and even on extreme overclocks the difference is marginal here. of course some have features or settings to make clocking easier but hell all of them can still do it.

I have used ASUS, MSI Gigabyte and even biostar man.... all of them can clock and lets face it if the board is 130-150 bucks likelihood of us strapping it to a test bench (except for the morbid curiosity or the "lets see what happens" factor)

Otherwise we all have our brand of choice for one or many reasons...

I like gigabyte boards because they feel solid as hell and are generally stable but the bios has been buggy at times for me.

I like asus because I think at this time they have a solid feature set and a even more solid BIOS.

I like Biostar because its cheap and just works every time I power it on and I can build for my parents without worry of post loops.

I like MSI cause I love military *pew pew* haha

seriously can we all stop bickering like little pansies and agree that every board can have issues and probably does if you look hard enough.


My rule is if it powers on and works as expected at stock then you got what you paid for....

NO overclock is guaranteed. Overclocking used to be a skillful thing you had to figure out to do it let alone be good at it... now taht overclocking is easier everyone es and moans unless the board jumps outta the box preps itself and clocks to a world record all while making them a sandwich...

ok i want a sandwich now..... DAMN YOU maximus V Gene why you no make me sammich?!?!?!

if you still feel like ing in this thread please just shut down your computer and walk away.