Tuesday, September 29th 2020

GIGABYTE Offers Statement Regarding the SP-CAP and MLCC Capacitor on GeForce RTX 3080 Graphics Cards

In response to the recent reports speculating that the use of POSCAP capacitors on the GeForce RTX 3080 graphics cards could lead to stability issues and crashes, GIGABYTE would like to clarify the issue with the following statement. It is false that POSCAP capacitors independently could cause a hardware crash. Whether a graphics card is stable or not requires a comprehensive evaluation of the overall circuit and power delivery design, not just the difference in capacitor types. POSCAPs and MLCCs have different characteristics and uses, thus it is not true to assert that one capacitor type is better than the other.

The GIGABYTE GeForce RTX 30 graphics cards are designed in accordance with NVIDIA specifications, and have passed all required testing, thus the product quality is guaranteed. GIGABYTE GeForce RTX 3080 GAMING OC and EAGLE OC series graphics cards use high-quality, low-ESR 470uF SP-CAP capacitors, which meet the specifications set by NVIDIA and provide a total capacity of 2820u in terms of GPU core power, higher than the industry's average. The cost of SP-CAP capacitors is not lower than that of MLCCs. GIGABYTE values product integrity highly and definitely does not reduce costs by using cheap materials.
NVIDIA has released a driver (version 456.55) on September 29, 2020 that improves stability. Users are advised to update to the latest driver for optimized performance. For users who have any concern, please contact our local service centers or representatives.

GIGABYTE has been constantly improving and optimizing product quality, especially in terms of thermal designs, to provide the best gaming experience to the consumers for decades. For the latest AORUS GeForce RTX 30 graphics card series, we have also paid extra attention to the cooling performance and introduced industry-leading solutions such as MAX-Covered Cooling to ensure that the operation of each component is stable.
Add your own comment

20 Comments on GIGABYTE Offers Statement Regarding the SP-CAP and MLCC Capacitor on GeForce RTX 3080 Graphics Cards

#1
Flanker
Reminds me a decade ago how every manufacturer of motherboards and psus advertise using only Japanese capacitors :p

Honestly I don't care where the capacitors are made or what type they are as long as they are issue free. I have grown rather fond of gigabyte products. So I will probably keep buying their stuff as long as they don't have serious issues or have alternatives from other manufacturers that gives me more for my money.
Posted on Reply
#2
Prima.Vera
BS statement is BS. Nice try Gigabyte, try again.
Posted on Reply
#3
phanbuey
Prima.Vera
BS statement is BS. Nice try Gigabyte, try again.
After owning a few gigabyte cards, I don't think I will buy another.
Posted on Reply
#4
SIGSEGV
OOPS! I agree with you Gigabyte. :rockout:
Posted on Reply
#5
Hossein Almet
To date, there is no reporting of Gigabyte cards CTD.
Posted on Reply
#6
Verpal
Thanks to Nvidia at least people will start to realize not all big black cap with strange word on it is just POSCAP with different capacity.

From now on, people will start to use SP-CAP/POSCAP interchangably and completely unaware of the tens and thousand of different caps out there.
Posted on Reply
#7
GorbazTheDragon
> provide a total capacity of 2820u in terms of GPU core power, higher than the industry's average
And what's the impedance/frequency curve like :D

> The cost of SP-CAP capacitors is not lower than that of MLCCs.
Yeah but there's 8(?) MLCCs in an array where you would put a SP-CAP? And then the production time costs on top of that?

nice try guys
Posted on Reply
#8
TheLostSwede
Who ever approved that press release should be fired. Since when is capacitance measured in μ (micro)? It's Farad or F, or in this case μF. Talk about a clueless writer and manger that approved it.
provide a total capacity of 2820u
That should be capacitance, not capacity. Also, a higher capacitance ≠ better, it's not how this works.

Also, this statement is as clear as mud and doesn't really say if there's an issue or not.
Posted on Reply
#9
Flanker
TheLostSwede
Who ever approved that press release should be fired. Since when is capacitance measured in μ (micro)? It's Farad or F, or in this case μF. Talk about a clueless writer and manger that approved it.


That should be capacitance, not capacity. Also, a higher capacitance ≠ better, it's not how this works.

Also, this statement is as clear as mud and doesn't really say if there's an issue or not.
Hey it's a wall of text in English, good enough:laugh:

Edit: the Chinese version has the exact same sloppiness. Disgraceful AF man
technews.tw/2020/09/30/nvidia-rtx-3080-poscap/
Posted on Reply
#11
ratirt
So it appears it was a driver issue?
Posted on Reply
#12
Ferrum Master
Such a low quality statement.

Capacity without frequency/voltage has no meaning. They should have refrained of posting that rubbish table. One does simply does do ONLY electrolytic capacitor without decoupling with ceramics for high frequencies, because they can't do it by design, no matter how good are they. There must be Ceramics as close as possible to IC pin and connected to a properly designed low impedance ground plane to short the HF.

Basically all this fuss is around nothing... it is just a retarded PCB/VRM design. You can blame anything in there... caps are just like a plaster to a corpse now. Why there is such an RF creeping in should be fixed in the first place. There are techniques to cancel it it with smart PCB design, there for the may be cases when everything just works without additional components. This ain't the case.

Bravo for the show in these late days. Any card owners, use your rights and return your cards, wait for new more mature PCB revisions. Driver fix to mitigate switching noise, thus the RF creep is a workaround. It is a step back.

Many AIB makers are just scared of the useless dud stock they have made now. They can only blame themselves. Those who made a delay and braced the issues, respect to them.
Posted on Reply
#13
maxitaxi96
well who to believe here. der8auer soldered 10 MLCC on the backside, replacing one POSCAP and his oc of 100 MHz became stable...
i think it does make a difference, but only on factory-OCed cards therefore it think the whole issue is mostly drama...
Posted on Reply
#14
kayjay010101
maxitaxi96
well who to believe here. der8auer soldered 10 MLCC on the backside, replacing one POSCAP and his oc of 100 MHz became stable...
i think it does make a difference, but only on factory-OCed cards therefore it think the whole issue is mostly drama...
If you'd listened to the video, he said it made around +40MHz OC stable by replacing TWO full SP-CAP with TWO arrays (that's 20) of MLCC's. This is around what every generation previous also had with cap-modding, so nothing exclusive to the 30 series, nor indicative of any particular issue that would result in any crashing, especially at stock, like people have reported. Removing two SP-CAPs (not POSCAPs...) resulted in a loss of ~60MHz of headroom, so his +70 core OC was no longer stable. Then adding TWO arrays (not one) of MLCCs brought it up to a +100 core OC being stable, while 120 still crashed. So comparing 6x SP-CAP to 4x SP-CAP 2x MLCC he got a whopping +30 MHz OC with MLCC over SP-CAPs.

Undoubtedly the problem was the aggressive brief boost to 2070MHz+ that made the crashes, as he outlines in the video. The driver was supposed to fix this, and it seems it has for a lot of people.
Posted on Reply
#15
ExcuseMeWtf
GIGABYTE values product integrity highly and definitely does not reduce costs by using cheap materials.
I definitely have bridge to sell too...

Said as owner of GTX 670 Windforce 3x working to this day BTW.
Posted on Reply
#16
maxitaxi96
kayjay010101
If you'd listened to the video, he said it made around +40MHz OC stable by replacing TWO full SP-CAP with TWO arrays (that's 20) of MLCC's. This is around what every generation previous also had with cap-modding, so nothing exclusive to the 30 series, nor indicative of any particular issue that would result in any crashing, especially at stock, like people have reported. Removing two SP-CAPs (not POSCAPs...) resulted in a loss of ~60MHz of headroom, so his +70 core OC was no longer stable. Then adding TWO arrays (not one) of MLCCs brought it up to a +100 core OC being stable, while 120 still crashed. So comparing 6x SP-CAP to 4x SP-CAP 2x MLCC he got a whopping +30 MHz OC with MLCC over SP-CAPs.

Undoubtedly the problem was the aggressive brief boost to 2070MHz+ that made the crashes, as he outlines in the video. The driver was supposed to fix this, and it seems it has for a lot of people.
Yeah sorry 20 MLCCs... but still the higher clock was not stable. It's not that outlandish to assume that some (not all) models that have a factory-OC AND 6 POSCAPs/ SPCAPs are prone to not being stable.

The new driver is effectivly a minimal downclock to bring those cards under the threshold where they would crash. So problem solved
Posted on Reply
#17
kiriakost
Whether a graphics card is stable or not requires a comprehensive evaluation of the overall circuit and power delivery design, not just the difference in capacitor types.

Finally some one which agrees with my own theory of conspiracy :D
MSI when you will be ready to point out issues at GPU design, I will offer you a warm hag.
But this will not happen, because its best to keep NVIDIA in the list of your friends at any cost.
Posted on Reply
#18
EarthDog
randomUser
Here
Buildzoid knows it better
lol, thats at least 5 times this video has been posted at tpu, lol.
Posted on Reply
#19
CmdrLaw
EarthDog
lol, thats at least 5 times this video has been posted at tpu, lol.
Well he's now done one specifically about this Gigabyte press statement ;)
Posted on Reply
#20
EarthDog
CmdrLaw
Well he's now done one specifically about this Gigabyte press statement ;)

TY!

It's a shame that it is 40 mins long... ZZzzzzzzzzzzzzzzzzzzzzzzzzzzzz. Why can't he be like Fornite youtubers and post like 10-12 min vids? Him and Techjesus... vids are WAY too long.
Posted on Reply
Add your own comment