Guru3D Rig of the Month - February 2021
ASUS GeForce RTX 3060 STRIX Gaming OC review
EVGA GeForce RTX 3060 XC Gaming review
MSI GeForce RTX 3060 Gaming X TRIO review
PALIT GeForce RTX 3060 DUAL OC review
ZOTAC GeForce RTX 3060 AMP WHITE review
Fractal Design Meshify 2 Compact chassis review
Sabrent Rocket 4 PLUS 2TB NVMe SSD review
MSI Radeon RX 6900 XT GAMING X TRIO review
Guru3D Q1 Winter 20/21 PC Buyer Guide
Gigabyte Releases Statement Regarding the SP-CAP and MLCC Capacitor on GeForce RTX 3080
Gigabyte is the first to offer an extensive press-release on the aforementioned CTD issues. With this statement they hope to clarify a thing or two, they deny the root issue to be POSCAP and MLCC related. Basically, they claim that the boards are meeting the speciations set by NVIDIA.
-- Gigabyte --
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, we 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.
| Model | NVIDIA RTX 3080 Reference Design | GIGABYTE RTX 3080 GAMING OC & EAGLE OC |
| GPU core Power | SP-CAP 220u/9m OHM *5 MLCC 22u *10 | SP-CAP 470u/6m OHM *6 |
| Total Capacity(u) | 1320u | 2820u |
GeForce RTX 3080 VISION OC 10G from Gigabyte has silver design and triple fan cooler - 09/29/2020 06:52 AM
From Gigabyte's "VISION" series for creators, a GeForce RTX 3080 graphics card equipped with a silver-colored triple-fan cooler "WINDFORCE 3X" has surfaced online....
Gigabyte Outs A Duo of new Mini-ITX motherboard A520I AC motherboards - 09/28/2020 07:45 AM
And with A520, these should not be expensive. Meet the A520I AC and A520M S2H. Two motherboard models that use the AMD A520 chipset have been released. Sales will start on October 2nd (Friday)....
Gigabyte GeForce RTX 3070 Series Renders Find their way to the www - 09/21/2020 07:28 AM
Gigabyte apparently has unveiled their GeForce RTX 3070 graphics cards. The cards are based on Ampere GA104 GPU and the two products are RTX 3070 GAMING OC and RTX 3070 EAGLE OC. ...
Gigabyte shows RTX 3090 TURBO: With a dual-slot Blower Style cooler - 09/19/2020 08:49 AM
We're not sure what to make of it, other then it's going to make quite a bit noise with a 350W GPU. Gigabyte will be outting a 1500 USD version RTX 3090 based on a blower-style cooler. ...
Review: Gigabyte B550 Vision D - 09/16/2020 07:03 AM
Meet the Gigabyte B550 Vision D motherboard. This product from the Vision series (which includes also a Z490 motherboard) is an AMD AM4 socket product targeted at creators. Built-in the ATX form-facto...
JonasBeckman
Senior Member
Posts: 17333
Joined: 2009-02-25
Senior Member
Posts: 17333
Joined: 2009-02-25
#5836857 Posted on: 09/30/2020 07:23 AM
The release notes only say a few titles were optimized for 3000 series stability with the new driver.
My assumption is that a few titles now have more stringent boost behavior so they can boost above 2 Ghz and NVIDIA hasn't locked it but it requires more.
Similar to AMD boosting the 5700's aggressively with the non-Adrenaline 2020 drivers and then the average went down a bit in later releases but with a few more conditions met the GPU's can still boost just as before.
Different systems behind these boost algorithms but similar in results but yeah I don't know any of the specifics here either.
EDIT: Comparisons would be needed though with the 456.20 drivers and the 456.55 drivers through GPU-Z or similar maybe use GeForce Experience to up the wattage and power draw and see if that lifts the boost restriction but the tweaks could be more than simply thermals or power.
Could even be game specific if there's a particular set of titles that are more finicky and prone to crashing than others and here NVIDIA could have done some other stuff that might not be about boost at all, hard to say without further details on what this actually is or someone more skilled being able to see what's been changed.
Guessing it's not as simple as profile comparing via NVIDIA Inspector for some of these games.
(The new Call of Duty perhaps?)
EDIT:
Nothing regarding boost apart from a small ns delay in switching between steps is done driver side, the steps are all on the cards vbios.
Ah so NVIDIA's is almost entirely bios side, interesting.
Hmm wonder what was tweaked then, well there's probably ways to handle this in the driver to resolve stability in various ways.
The release notes only say a few titles were optimized for 3000 series stability with the new driver.
My assumption is that a few titles now have more stringent boost behavior so they can boost above 2 Ghz and NVIDIA hasn't locked it but it requires more.
Similar to AMD boosting the 5700's aggressively with the non-Adrenaline 2020 drivers and then the average went down a bit in later releases but with a few more conditions met the GPU's can still boost just as before.
Different systems behind these boost algorithms but similar in results but yeah I don't know any of the specifics here either.
EDIT: Comparisons would be needed though with the 456.20 drivers and the 456.55 drivers through GPU-Z or similar maybe use GeForce Experience to up the wattage and power draw and see if that lifts the boost restriction but the tweaks could be more than simply thermals or power.
Could even be game specific if there's a particular set of titles that are more finicky and prone to crashing than others and here NVIDIA could have done some other stuff that might not be about boost at all, hard to say without further details on what this actually is or someone more skilled being able to see what's been changed.
Guessing it's not as simple as profile comparing via NVIDIA Inspector for some of these games.
(The new Call of Duty perhaps?)
EDIT:
Nothing regarding boost apart from a small ns delay in switching between steps is done driver side, the steps are all on the cards vbios.
Ah so NVIDIA's is almost entirely bios side, interesting.
Hmm wonder what was tweaked then, well there's probably ways to handle this in the driver to resolve stability in various ways.
Maddness
Senior Member
Posts: 1459
Joined: 2014-10-24
Senior Member
Posts: 1459
Joined: 2014-10-24
#5836865 Posted on: 09/30/2020 07:50 AM
Appreciate the time it took you to type all that.
Appreciate the time it took you to type all that.
JonasBeckman
Senior Member
Posts: 17333
Joined: 2009-02-25
Senior Member
Posts: 17333
Joined: 2009-02-25
#5836869 Posted on: 09/30/2020 08:01 AM
Yeah but it's kinda invalidated by what Astyanax posted if it's handled deeper from the GPU bios rather than the drivers controlling the behavior so it's something else that's been done here.
But without any details on as to what's been changed and there's a lot of stuff some of the display driver overrides can be capable of even for DirectX 12 and Vulkan and the resulting fixes, optimizations or tweaks for stability or game engine issues and more.
Hopefully though whatever NVIDIA did tweak does at least improve the stability situation with the current GPU models that do have problems here until newer bios updates can alter the algorithm making it a bit less unstable even if it means stricter criteria for how the GPU boosts and as a result a lower overall clock speed in some systems.
AMD did that with their AGESA code updates for the Ryzen CPU's as a bios example I am aware of (Mainly the temperature thresholds lowered by 5 - 10 degrees or so Celsius.) though GPU wise it's often been handled by Wattman and Overdrive and their power play API in software which would differ from NVIDIA if theirs is primarily from the bios data on the GPU.
(For AMD there were Bios updates to set a higher speed for the 5600 GPU's memory though that's a little bit different.)
Some hooks do exist though like what can be done with Afterburner or the new GeForce Experience but NVIDIA directly stated they aren't capping the clock speeds and that probably also means no tweaks to power draw to limit GPU max boosts either.
EDIT: Hmm that or NVIDIA or the manufacturer for cards that do have confirmed issues open up for RMA's for refreshed models with hardware fixes but that's costlier than a bios update.
Yeah but it's kinda invalidated by what Astyanax posted if it's handled deeper from the GPU bios rather than the drivers controlling the behavior so it's something else that's been done here.
But without any details on as to what's been changed and there's a lot of stuff some of the display driver overrides can be capable of even for DirectX 12 and Vulkan and the resulting fixes, optimizations or tweaks for stability or game engine issues and more.
Hopefully though whatever NVIDIA did tweak does at least improve the stability situation with the current GPU models that do have problems here until newer bios updates can alter the algorithm making it a bit less unstable even if it means stricter criteria for how the GPU boosts and as a result a lower overall clock speed in some systems.
AMD did that with their AGESA code updates for the Ryzen CPU's as a bios example I am aware of (Mainly the temperature thresholds lowered by 5 - 10 degrees or so Celsius.) though GPU wise it's often been handled by Wattman and Overdrive and their power play API in software which would differ from NVIDIA if theirs is primarily from the bios data on the GPU.
(For AMD there were Bios updates to set a higher speed for the 5600 GPU's memory though that's a little bit different.)
Some hooks do exist though like what can be done with Afterburner or the new GeForce Experience but NVIDIA directly stated they aren't capping the clock speeds and that probably also means no tweaks to power draw to limit GPU max boosts either.
EDIT: Hmm that or NVIDIA or the manufacturer for cards that do have confirmed issues open up for RMA's for refreshed models with hardware fixes but that's costlier than a bios update.
kapu
Senior Member
Posts: 4420
Joined: 2006-12-22
Senior Member
Posts: 4420
Joined: 2006-12-22
#5836899 Posted on: 09/30/2020 09:11 AM
Guys, due to Der8auer's test, just 2 sets of MLCC conbination make 30MHz difference. What if we replace all 6 SP-CAPs? The hardware is been set when it off the streamline, the only thing to make it work properly is to tweak the driver software. If there's a defect in component aspect, software can do noting to make hardware better, the only thing software can do is, adapt its defect.
Also, Gigabyte said 'SP- CAPs are not cheaper than MLCCs', that's really interesting. Your engineer chose SP-CAPs with the same cost only to get much worse result. I guess some people will lose their job.
YOU have asus TUF with 6 mlcc's , do they boost higher ? do they overclock better ? Should be quite simple to test right now.
Guys, due to Der8auer's test, just 2 sets of MLCC conbination make 30MHz difference. What if we replace all 6 SP-CAPs? The hardware is been set when it off the streamline, the only thing to make it work properly is to tweak the driver software. If there's a defect in component aspect, software can do noting to make hardware better, the only thing software can do is, adapt its defect.
Also, Gigabyte said 'SP- CAPs are not cheaper than MLCCs', that's really interesting. Your engineer chose SP-CAPs with the same cost only to get much worse result. I guess some people will lose their job.
YOU have asus TUF with 6 mlcc's , do they boost higher ? do they overclock better ? Should be quite simple to test right now.
Click here to post a comment for this news story on the message forum.
Senior Member
Posts: 1459
Joined: 2014-10-24
Did they give any tidbits on what was done to make the cards more stable.