Chamidorix
New Member
- Total Posts : 36
- Reward points : 0
- Joined: 2020/09/24 05:24:12
- Status: offline
- Ribbons : 1
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 11:23:05
(permalink)
My friends, they've ignored this for 2+ months on the FTW3. Time to stop being an apologist (hopefully wwxww can learn to read his own definition), and understand that EVGA (moderated for language) up hard this generation and hardcoded an analog ratio in the load balancer chips that they cannot change on these initial/1st generation PCBs. For multiple generations now Nvidia PCB designs have included chips that dynamically route power from 8-pins + pci-e slot to VRM stages, to enable even loading of VRMs and uniform pull from pci-e psu cables. Some chips have had digital controllers that could be adjusted via memory scan or bios/firmware updates, but EVGA appears to be using analog chips that take hard-set input voltages on chip pins to control the ratio (bearded hardware manually hardmodded this voltage to adjust ratio as part of his FTW3 unlocking). Note that this is separate from the voltage controller, which is digital on the kingpin and analog on the FTW3. Now, each 8-pin and pci-e slot have an individual power limit, measured via shunt, and the core (nvvdd+msvdd) and vram (fbvdd) each have an individual power limit, separate from the overall board power limit. The pci-e slot limit is around 80W. The hardcoded ratio on the balancer chips is set to low, that when the pci-slot is at 80W, the 8 pins are far below their individual power limits of 150W each. The reason Furmark etc can pull more is they pull way more power on VRAM, and the VRAM VRM stages do not appear to be balanced against pci-e slot like core does. On a solution note, it looks like either RMA until you get a batch where they've adjusted the ratio on the chip, or since we've got the standard 7 shunt assortment you can just shunt the pci slot 5MO resistor like FTW3 (the blue circled on in bottom image). It's a wider shunt to accommodate the well isolated traces and I'm working on figuring out the best resistor to buy from digikey etc to stack or replace.
post edited by Delirious - 2020/12/18 18:12:12
|
Cadillac94pimpin
Superclocked Member
- Total Posts : 182
- Reward points : 0
- Joined: 2019/03/26 23:51:35
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 11:50:56
(permalink)
Pretty solid explanation you give here. Do you know why some cards seem to be affected while others are not?
i9-9900K EVGA 3090 KPE (PUSH PULL) MSI Z390 MEG ACE 32 GB 3200 MHZ CORSAIR H150i Pro WD BLACK SN750 SEASONIC 850W Focus Gold+ Lian Li O11D XL + 10x LL120s+3xML 120s
|
glocked89
Superclocked Member
- Total Posts : 187
- Reward points : 0
- Joined: 2019/01/16 23:48:00
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 12:01:10
(permalink)
Chamidorix My friends, they've ignored this for 2+ months on the FTW3. Time to stop being an apologist (hopefully wwxww can learn to read his own definition), and understand that EVGA up hard this generation and hardcoded an analog ratio in the load balancer chips that they cannot change on these initial/1st generation PCBs. For multiple generations now Nvidia PCB designs have included chips that dynamically route power from 8-pins + pci-e slot to VRM stages, to enable even loading of VRMs and uniform pull from pci-e psu cables. Some chips have had digital controllers that could be adjusted via memory scan or bios/firmware updates, but EVGA appears to be using analog chips that take hard-set input voltages on chip pins to control the ratio (bearded hardware manually hardmodded this voltage to adjust ratio as part of his FTW3 unlocking). Now, each 8-pin and pci-e slot have an individual power limit, measured via shunt, and the core (nvvdd+msvdd) and vram (fbvdd) each have an individual power limit, separate from the overall board power limit. The pci-e slot limit is around 80W. The hardcoded ratio on the balancer chips is set to low, that when the pci-slot is at 80W, the 8 pins are far below their individual power limits of 150W each. The reason Furmark etc can pull more is they pull way more power on VRAM, and the VRAM VRM stages do not appear to be balanced against pci-e slot like core does. On a solution note, it looks like either RMA until you get a batch where they've adjusted the ratio on the chip, or since we've got the standard 7 shunt assortment you can just shunt the pci slot 5MO resistor like FTW3 (the blue circled on in bottom image). It's a wider shunt to accommodate the well isolated traces and I'm working on figuring out the best resistor to buy from digikey etc to stack or replace.
What is a safe average power draw from the pcie slot assuming I shunt it? My x299 board has additional power going to the pcie lanes coming from the psu, will this alleviate the dangers of excessive power draw from the pcie slot?
post edited by Delirious - 2020/12/18 18:12:42
|
degenerate
SSC Member
- Total Posts : 614
- Reward points : 0
- Joined: 2013/08/23 05:38:31
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 12:01:36
(permalink)
Chamidorix My friends, they've ignored this for 2+ months on the FTW3. Time to stop being an apologist (hopefully wwxww can learn to read his own definition), and understand that EVGA ****ed up hard this generation and hardcoded an analog ratio in the load balancer chips that they cannot change on these initial/1st generation PCBs.
Imagine having to shunt-mod a Kingpin to make it worthwhile... If true, I'll be returning mine. Sucks because I cancelled my other back-order that I was waiting 3 months for. Good god this GPU launch is enough to turn a man to another hobby... I mean really...!
|5900x|MSI B550 Unify-X|2x16GB G.Skill 3800 C14 @ 1900 FCLK|2TB WD SN850 M.2|EVGA 3090 KPE|EVGA 1600T2|Alienware '25' & LG 48" CX OLED|
|
Dabadger84
CLASSIFIED Member
- Total Posts : 3426
- Reward points : 0
- Joined: 2018/05/11 23:49:52
- Location: de_Overpass, USA
- Status: offline
- Ribbons : 10
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 12:09:56
(permalink)
I would not shunt-mod the card (especially don't shunt mod the PCIe power, that is actually a very bad idea itself), if it doesn't hit the power targets it should be hitting, RMA it - that's my opinion, as someone who's planning on buying one.
ModRigs: https://www.modsrigs.com/detail.aspx?BuildID=42891 Specs:5950x @ 4.7GHz 1.3V - Asus Crosshair VIII Hero - eVGA 1200W P2 - 4x8GB G.Skill Trident Z Royal Silver @ 3800 CL14Gigabyte RTX 4090 Gaming OC w/ Core: 2850MHz @ 1000mV, Mem: +1500MHz - Samsung Odyssey G9 49" Super-Ultrawide 240Hz Monitor
|
JZegers
New Member
- Total Posts : 57
- Reward points : 0
- Joined: 2010/06/30 10:07:43
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 12:10:44
(permalink)
How exactly do you plan to just return it?
|
wwxww
iCX Member
- Total Posts : 462
- Reward points : 0
- Joined: 2019/01/13 09:30:51
- Location: New Orleans
- Status: offline
- Ribbons : 1
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 12:12:54
(permalink)
@Camidorix It seems I misinterpreted your first post. Thanks for the informative post for the uninformed. i.e. me. How would the card react to the shunt modification when not under load? Would the power draw be any different from the entire card with little load?
|
braxtonjames
New Member
- Total Posts : 6
- Reward points : 0
- Joined: 2020/04/19 16:15:36
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 12:45:35
(permalink)
I am unfortunately having the same issue :(
|
thegreattonge
New Member
- Total Posts : 39
- Reward points : 0
- Joined: 2019/09/23 15:30:36
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 12:54:16
(permalink)
Dabadger84 I would not shunt-mod the card (especially don't shunt mod the PCIe power, that is actually a very bad idea itself), if it doesn't hit the power targets it should be hitting, RMA it - that's my opinion, as someone who's planning on buying one.
^ this If the product has a design issue, give EVGA the opportunity to fix it. this isn't an apologist POV, but more so a level headed consumer.
|
Chamidorix
New Member
- Total Posts : 36
- Reward points : 0
- Joined: 2020/09/24 05:24:12
- Status: offline
- Ribbons : 1
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 13:50:13
(permalink)
Cadillac94pimpin Pretty solid explanation you give here. Do you know why some cards seem to be affected while others are not?
It is quite simple, they revised and fixed the problem on newer manufactured pcbs. As stated, the load balancing chips take an input voltage to determine the ratio so they simply adjusted this input voltage, or similar. It's already appearing more and more with newer FTW3s (an increasing number of reports that do 500W with the XOC bios), the very simple way to check is if you pci-e slot is drawing ~75+W or 50-60W under time spy extreme load. PCBs are manufactured in batches and there certainly are multiple batches of KPE PCBs that have been created, so it follows that you can reasonably assume the newer ones have the correct balance while the older ones made before the whole FTW3 XOC debacle (or whenever EVGA became aware) still have the problem. My KPE edition finally arrived today, so I will be testing the PCI-E draw and asking for an RMA if I have this issue. I'd recommend all other KPE owners do the same (run your card under TSE on 520W bios and observe PCI-E slot power draw in GPU-Z or HWInfo), to hopefully pressure EVGA into providing more support or clarifying if there is a misunderstanding somewhere.
post edited by Chamidorix - 2020/12/18 13:56:23
Attached Image(s)
|
Cadillac94pimpin
Superclocked Member
- Total Posts : 182
- Reward points : 0
- Joined: 2019/03/26 23:51:35
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 14:02:37
(permalink)
braxtonjames I am unfortunately having the same issue :(
Can you describe your issue?
i9-9900K EVGA 3090 KPE (PUSH PULL) MSI Z390 MEG ACE 32 GB 3200 MHZ CORSAIR H150i Pro WD BLACK SN750 SEASONIC 850W Focus Gold+ Lian Li O11D XL + 10x LL120s+3xML 120s
|
braxtonjames
New Member
- Total Posts : 6
- Reward points : 0
- Joined: 2020/04/19 16:15:36
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 14:15:27
(permalink)
Cadillac94pimpin
braxtonjames I am unfortunately having the same issue :(
Can you describe your issue?
For some reason, I am getting power-limited in the 460-470W range across multiple workloads/benchmarks. Maximum PCIe slot power draw I saw across Time Spy Extreme, Port Royal, etc. was 66.88w according to HWiNFO64. I am running the LN2 BIOS, confirmed 520W/121% PL slider.
|
Kylearan
iCX Member
- Total Posts : 288
- Reward points : 0
- Joined: 2013/12/26 04:04:40
- Status: offline
- Ribbons : 2
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 14:44:17
(permalink)
Everyone here posting about a "low" TDP% in Timespy Extreme Please use HWINFo64 and compare the TDP% and TDP Normalized %, please. GPU-Z does not show TDP Normalized. Also if you have time, do the same run on "Superposition 4k--custom--extreme shaders" and compare it to "Superposition 4k optimized" (again look at TDP vs TDP Normalized!!) Report back. I have a theory, but I need you guys to test this for me because I do NOT have an eVGA card. I have a shunt modded 3090 FE, and I can draw up to 600W without hitting the power limit, but there is "something" going on with both Timespy Extreme and "Superposition 4k--extreme shaders (custom)" that does NOT happen with Timespy nor 4k Optimized. Also note that 4k extreme shaders in Superposition has a MUCH MUCH MUCH lower FPS than 4k optimized. I need you guys to do some testing and REPORT your TDP Normalized vs TDP % vs "TDP Slider % set". I am Falkentyne on overclock.net, by the way.
post edited by Kylearan - 2020/12/18 15:08:01
|
nycx360
Superclocked Member
- Total Posts : 212
- Reward points : 0
- Joined: 2009/08/23 14:46:09
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 15:53:11
(permalink)
Never hit 520. Pwr cap. Timespy extreme
|
markuaw1
iCX Member
- Total Posts : 329
- Reward points : 0
- Joined: 2017/12/15 17:17:39
- Status: offline
- Ribbons : 1
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 16:02:52
(permalink)
Stickboy46 I don't see 500+ on PR or anything in 3dMark. Moment I fire up Furmark and run the short little 4k test, it pulls 500-520 consistently. I'm not seeing any downclocking though .. it's a rock solid clock throughout PR or even when gaming. https://www.3dmark.com/pr/653446
https://www.3dmark.com/pr/648784 STRIX, EK Block & Kingpin 520W bios :) good Bios thanks EVGA !
post edited by markuaw1 - 2020/12/18 16:08:41
|
edgeofsanity
New Member
- Total Posts : 17
- Reward points : 0
- Joined: 2009/01/13 21:43:37
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 16:06:11
(permalink)
Can't believe I have shunt mod a "LN2" card to get the advertised power limit. 110% power target is the upper limit with a 460-470w board power draw in timespy extreme GPU test 1. GPU test 2 will see some spikes over 500w but stays well below 99.9% of the time. PCIE power draw typically 55-60w. Strongly considering an RMA, will wait for my ftw3 block first I think. Can't go back to air cooled noise hell with this right now.
|
Sajin
EVGA Forum Moderator
- Total Posts : 49137
- Reward points : 0
- Joined: 2010/06/07 21:11:51
- Location: Texas, USA.
- Status: online
- Ribbons : 199
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 16:18:06
(permalink)
edgeofsanity Can't believe I have shunt mod a "LN2" card to get the advertised power limit. 110% power target is the upper limit with a 460-470w board power draw in timespy extreme GPU test 1. GPU test 2 will see some spikes over 500w but stays well below 99.9% of the time. PCIE power draw typically 55-60w. Strongly considering an RMA, will wait for my ftw3 block first I think. Can't go back to air cooled noise hell with this right now.
Lol. Shunt modding isn’t going to help you hit the max power limit. Shunt modding removes power limits. To hit the 520w limit on the kingpin card you need to increase your voltage using the classified tool.
|
edgeofsanity
New Member
- Total Posts : 17
- Reward points : 0
- Joined: 2009/01/13 21:43:37
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 16:32:51
(permalink)
Sajin
edgeofsanity Can't believe I have shunt mod a "LN2" card to get the advertised power limit. 110% power target is the upper limit with a 460-470w board power draw in timespy extreme GPU test 1. GPU test 2 will see some spikes over 500w but stays well below 99.9% of the time. PCIE power draw typically 55-60w. Strongly considering an RMA, will wait for my ftw3 block first I think. Can't go back to air cooled noise hell with this right now.
Lol. Shunt modding isn’t going to help you hit the max power limit. Shunt modding removes power limits. To hit the 520w limit on the kingpin card you need to increase your voltage using the classified tool.
Shunting moves the power limit up, doesn't remove it. Was referring to shunting pcie only in this instance. Classified tool isn't going to change this.
|
ffrgtm
New Member
- Total Posts : 15
- Reward points : 0
- Joined: 2016/07/13 23:50:46
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 16:33:49
(permalink)
From the 2080ti kp xdevs page: "Make sure to max out power limiters to ensure best performance, even when reported power levels looks same. Now you may ask, how can I get card to show advertised 520W in Precision X1 or OLED? Well, that’s a lot of power and Precision X1 will give you only averaged value. The performance will be capped significantly way before you’d be able to reach averaged 520W power reporting in PX1." Following TiN's word's above: is 520W not a limit, rather than target? Also, I'm afraid I can't lend much credibility to the hardware rev. theory being crafted in this thread: I believe I have one of the first cards shipped (processed before thanksgiving) and my pcie slot draws are well under the spec.
|
Sajin
EVGA Forum Moderator
- Total Posts : 49137
- Reward points : 0
- Joined: 2010/06/07 21:11:51
- Location: Texas, USA.
- Status: online
- Ribbons : 199
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 16:37:57
(permalink)
edgeofsanity
Sajin
edgeofsanity Can't believe I have shunt mod a "LN2" card to get the advertised power limit. 110% power target is the upper limit with a 460-470w board power draw in timespy extreme GPU test 1. GPU test 2 will see some spikes over 500w but stays well below 99.9% of the time. PCIE power draw typically 55-60w. Strongly considering an RMA, will wait for my ftw3 block first I think. Can't go back to air cooled noise hell with this right now.
Lol. Shunt modding isn’t going to help you hit the max power limit. Shunt modding removes power limits. To hit the 520w limit on the kingpin card you need to increase your voltage using the classified tool.
Shunting moves the power limit up, doesn't remove it. Was referring to shunting pcie only in this instance. Classified tool isn't going to change this.
It doesn’t move it up. It lowers the reading so you can get around limits.
|
edgeofsanity
New Member
- Total Posts : 17
- Reward points : 0
- Joined: 2009/01/13 21:43:37
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 16:52:35
(permalink)
Sajin
edgeofsanity
Sajin
edgeofsanity Can't believe I have shunt mod a "LN2" card to get the advertised power limit. 110% power target is the upper limit with a 460-470w board power draw in timespy extreme GPU test 1. GPU test 2 will see some spikes over 500w but stays well below 99.9% of the time. PCIE power draw typically 55-60w. Strongly considering an RMA, will wait for my ftw3 block first I think. Can't go back to air cooled noise hell with this right now.
Lol. Shunt modding isn’t going to help you hit the max power limit. Shunt modding removes power limits. To hit the 520w limit on the kingpin card you need to increase your voltage using the classified tool.
Shunting moves the power limit up, doesn't remove it. Was referring to shunting pcie only in this instance. Classified tool isn't going to change this.
It doesn’t move it up. It lowers the reading so you can get around limits.
Well aware of what it does.
|
Kylearan
iCX Member
- Total Posts : 288
- Reward points : 0
- Joined: 2013/12/26 04:04:40
- Status: offline
- Ribbons : 2
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 17:03:58
(permalink)
nycx360
Never hit 520. Pwr cap. Timespy extreme
I think we found the problem. ---------------------- You guys may want to look into this a bit more carefully than what you did last time...(with the FTW3 cards). Looks like Kingpin cards are throttling because of WAY too low VRAM (MVDDC) and chip power limits... https://www.overclock.net...page-374#post-28696379 https://docs.google.com/spreadsheets/d/1VWprUJcV6mDwFmg1b6T1cySD1P-sjcG4-bACq-Y3vHs/edit#gid=0 Furthermore, why is eVGA using 92W for VRAM (MVDDC) limit on half of their SKU's--including KINGPIN cards, and 121W for other boards? Why would they use such a low MVDDC limit on a king pin card? Asus seems to be using up to 162W for MVDCC (FVBDD) on a few of their cards... I posted many pages back, when someone did a huge batch of tests and made an excel sheet with the power rails posted, that "MVDDC" might be triggering a power limit. One person seemed to take note but this seemed to have just been completely forgotten and it went back to PCIE Slot again.... But according to the Vbios dump, the PCIE slot limit is 82W, not 75W.... ffrgtm From the 2080ti kp xdevs page: "Make sure to max out power limiters to ensure best performance, even when reported power levels looks same. Now you may ask, how can I get card to show advertised 520W in Precision X1 or OLED? Well, that’s a lot of power and Precision X1 will give you only averaged value. The performance will be capped significantly way before you’d be able to reach averaged 520W power reporting in PX1." Following TiN's word's above: is 520W not a limit, rather than target? Also, I'm afraid I can't lend much credibility to the hardware rev. theory being crafted in this thread: I believe I have one of the first cards shipped (processed before thanksgiving) and my pcie slot draws are well under the spec.
The problem is, changing the power limit slider is ONLY going to change the TDP limit with respect to the 8 pins+Slot power. TDP is specifically 8 pin #1+8 pin #2+8 pin #3+PCIE Slot. The TDP slider doesn't change the internal power rails at all. It looks like "MVDDC" and "Chip Power" are supposed to have a "default" and "max", but they are both the same--which means they don't change at all. Those limits either need to respond to the TDP slider or be set higher by default. (and note: there are shunts onboard for both Chip Power and MVDDC....)
post edited by Kylearan - 2020/12/18 17:07:29
|
Magrecite
New Member
- Total Posts : 35
- Reward points : 0
- Joined: 2012/12/19 08:31:28
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 17:42:31
(permalink)
Kylearan
nycx360
Never hit 520. Pwr cap. Timespy extreme
I think we found the problem. ---------------------- You guys may want to look into this a bit more carefully than what you did last time...(with the FTW3 cards). Looks like Kingpin cards are throttling because of WAY too low VRAM (MVDDC) and chip power limits... https://www.overclock.net...page-374#post-28696379 https://docs.google.com/spreadsheets/d/1VWprUJcV6mDwFmg1b6T1cySD1P-sjcG4-bACq-Y3vHs/edit#gid=0 Furthermore, why is eVGA using 92W for VRAM (MVDDC) limit on half of their SKU's--including KINGPIN cards, and 121W for other boards? Why would they use such a low MVDDC limit on a king pin card? Asus seems to be using up to 162W for MVDCC (FVBDD) on a few of their cards... I posted many pages back, when someone did a huge batch of tests and made an excel sheet with the power rails posted, that "MVDDC" might be triggering a power limit. One person seemed to take note but this seemed to have just been completely forgotten and it went back to PCIE Slot again.... But according to the Vbios dump, the PCIE slot limit is 82W, not 75W....
ffrgtm From the 2080ti kp xdevs page: "Make sure to max out power limiters to ensure best performance, even when reported power levels looks same. Now you may ask, how can I get card to show advertised 520W in Precision X1 or OLED? Well, that’s a lot of power and Precision X1 will give you only averaged value. The performance will be capped significantly way before you’d be able to reach averaged 520W power reporting in PX1." Following TiN's word's above: is 520W not a limit, rather than target? Also, I'm afraid I can't lend much credibility to the hardware rev. theory being crafted in this thread: I believe I have one of the first cards shipped (processed before thanksgiving) and my pcie slot draws are well under the spec.
The problem is, changing the power limit slider is ONLY going to change the TDP limit with respect to the 8 pins+Slot power. TDP is specifically 8 pin #1+8 pin #2+8 pin #3+PCIE Slot. The TDP slider doesn't change the internal power rails at all. It looks like "MVDDC" and "Chip Power" are supposed to have a "default" and "max", but they are both the same--which means they don't change at all. Those limits either need to respond to the TDP slider or be set higher by default. (and note: there are shunts onboard for both Chip Power and MVDDC....)
So does this essentially mean it would still need an RMA or is that something that could be completely tweaked within the Classified tool itself? I tried playing around with the voltages a bit more (even though the card is being RMA'd for the obnoxious buzzing) and the only result was more throttling.
|
edgeofsanity
New Member
- Total Posts : 17
- Reward points : 0
- Joined: 2009/01/13 21:43:37
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 17:57:37
(permalink)
Magrecite
Kylearan
nycx360
Never hit 520. Pwr cap. Timespy extreme
I think we found the problem. ---------------------- You guys may want to look into this a bit more carefully than what you did last time...(with the FTW3 cards). Looks like Kingpin cards are throttling because of WAY too low VRAM (MVDDC) and chip power limits...
Furthermore, why is eVGA using 92W for VRAM (MVDDC) limit on half of their SKU's--including KINGPIN cards, and 121W for other boards? Why would they use such a low MVDDC limit on a king pin card? Asus seems to be using up to 162W for MVDCC (FVBDD) on a few of their cards... I posted many pages back, when someone did a huge batch of tests and made an excel sheet with the power rails posted, that "MVDDC" might be triggering a power limit. One person seemed to take note but this seemed to have just been completely forgotten and it went back to PCIE Slot again.... But according to the Vbios dump, the PCIE slot limit is 82W, not 75W....
ffrgtm From the 2080ti kp xdevs page: "Make sure to max out power limiters to ensure best performance, even when reported power levels looks same. Now you may ask, how can I get card to show advertised 520W in Precision X1 or OLED? Well, that’s a lot of power and Precision X1 will give you only averaged value. The performance will be capped significantly way before you’d be able to reach averaged 520W power reporting in PX1." Following TiN's word's above: is 520W not a limit, rather than target? Also, I'm afraid I can't lend much credibility to the hardware rev. theory being crafted in this thread: I believe I have one of the first cards shipped (processed before thanksgiving) and my pcie slot draws are well under the spec.
The problem is, changing the power limit slider is ONLY going to change the TDP limit with respect to the 8 pins+Slot power. TDP is specifically 8 pin #1+8 pin #2+8 pin #3+PCIE Slot. The TDP slider doesn't change the internal power rails at all. It looks like "MVDDC" and "Chip Power" are supposed to have a "default" and "max", but they are both the same--which means they don't change at all. Those limits either need to respond to the TDP slider or be set higher by default. (and note: there are shunts onboard for both Chip Power and MVDDC....)
So does this essentially mean it would still need an RMA or is that something that could be completely tweaked within the Classified tool itself? I tried playing around with the voltages a bit more (even though the card is being RMA'd for the obnoxious buzzing) and the only result was more throttling.
Sounds like you would have to undervolt.
|
ffrgtm
New Member
- Total Posts : 15
- Reward points : 0
- Joined: 2016/07/13 23:50:46
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 18:00:18
(permalink)
Magrecite So does this essentially mean it would still need an RMA or is that something that could be completely tweaked within the Classified tool itself? I tried playing around with the voltages a bit more (even though the card is being RMA'd for the obnoxious buzzing) and the only result was more throttling.
The classified tool lets you disable OCP for NVVDD, MSVDD, but not FBVDD. FBVDD is the one we would want to be able to disable if I'm understanding Kyleran correctly.
|
808sting
iCX Member
- Total Posts : 462
- Reward points : 0
- Joined: 2012/04/13 15:04:51
- Location: US
- Status: offline
- Ribbons : 6
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 18:11:56
(permalink)
Has anyone used the new classified tool on the 3090 KP with LN2 bios? Then, emailed Vince for the true XOC bios? Tool added msvdd(?) tweaks. My problem with my 2080ti KP had been cooling. With my lazy efforts, I could only cool the GPU to 14C idle and high 30s at load using 1.17v. GPU maxed at mid 2250s. It follows Tin's boost/temp tables for his results. I would think the 3000 series is similar based on Steve's LN2 baselines boost tables he did early on the 3090.
i9 13900KS, Asus ROG Hero Z690 MB, 32GB G.Skill DDR5-6400Gigabyte 4090 Gaming OCExternal open-loop w/Velocity2 WBCorsair 900D CaseFirecuda SSD 1TB & 2TB, EVO 970P 1TB, Black 8TBAsus & LG BD-RECorsair AX1600i PSLG 38GL950B-G
|
LVNeptune
iCX Member
- Total Posts : 287
- Reward points : 0
- Joined: 2011/03/16 23:41:53
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 18:20:05
(permalink)
wwxww
a·pol·o·gist
/əˈpäləjəst/
noun plural noun: apologists
- a person who offers an argument in defense of something controversial.
"critics said he was an apologist for colonialism"
h Similar:
defender supporter
upholder
advocate
proponent
exponent
propagandist
apostle
champion
backer
promoter
campaigner
spokesman
spokeswoman
spokesperson
Hmmmmm.....
Spam.
|
LVNeptune
iCX Member
- Total Posts : 287
- Reward points : 0
- Joined: 2011/03/16 23:41:53
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 18:25:08
(permalink)
808sting Has anyone used the new classified tool on the 3090 KP with LN2 bios? Then, emailed Vince for the true XOC bios? Tool added msvdd(?) tweaks. My problem with my 2080ti KP had been cooling. With my lazy efforts, I could only cool the GPU to 14C idle and high 30s at load using 1.17v. GPU maxed at mid 2250s. It follows Tin's boost/temp tables for his results. I would think the 3000 series is similar based on Steve's LN2 baselines boost tables he did early on the 3090.
No one is going to be stupid enough to invalidate their warranty at this point. I think they were hoping people would request the no limit bios so they wouldn't have to RMA cards.
|
Sajin
EVGA Forum Moderator
- Total Posts : 49137
- Reward points : 0
- Joined: 2010/06/07 21:11:51
- Location: Texas, USA.
- Status: online
- Ribbons : 199
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 18:33:22
(permalink)
LVNeptune
808sting Has anyone used the new classified tool on the 3090 KP with LN2 bios? Then, emailed Vince for the true XOC bios? Tool added msvdd(?) tweaks. My problem with my 2080ti KP had been cooling. With my lazy efforts, I could only cool the GPU to 14C idle and high 30s at load using 1.17v. GPU maxed at mid 2250s. It follows Tin's boost/temp tables for his results. I would think the 3000 series is similar based on Steve's LN2 baselines boost tables he did early on the 3090.
No one is going to be stupid enough to invalidate their warranty at this point. I think they were hoping people would request the no limit bios so they wouldn't have to RMA cards.
I ran the true xoc vbios on my card, and pushed 625w through it.
|
USMC1419
Superclocked Member
- Total Posts : 179
- Reward points : 0
- Joined: 2008/11/05 14:36:05
- Status: offline
- Ribbons : 0
Re: RTX 3090 K|NGP|N Power Draw Issues
2020/12/18 18:35:08
(permalink)
Chamidorix It is quite simple, they revised and fixed the problem on newer manufactured pcbs. As stated, the load balancing chips take an input voltage to determine the ratio so they simply adjusted this input voltage, or similar. It's already appearing more and more with newer FTW3s (an increasing number of reports that do 500W with the XOC bios), the very simple way to check is if you pci-e slot is drawing ~75+W or 50-60W under time spy extreme load. PCBs are manufactured in batches and there certainly are multiple batches of KPE PCBs that have been created, so it follows that you can reasonably assume the newer ones have the correct balance while the older ones made before the whole FTW3 XOC debacle (or whenever EVGA became aware) still have the problem. My KPE edition finally arrived today, so I will be testing the PCI-E draw and asking for an RMA if I have this issue. I'd recommend all other KPE owners do the same (run your card under TSE on 520W bios and observe PCI-E slot power draw in GPU-Z or HWInfo), to hopefully pressure EVGA into providing more support or clarifying if there is a misunderstanding somewhere.
What were your results? Also see below... Kylearan I need you guys to do some testing and REPORT your TDP Normalized vs TDP % vs "TDP Slider % set". I am Falkentyne on overclock.net, by the way.
Can you explain the difference and what it means related to power draw? Kylearan "Looks like Kingpin cards are throttling because of WAY too low VRAM (MVDDC) and chip power limits..."
This goes directly against what was said at the top of my post as a reason why these cards wont use the power they should. The top post states it is a hardware, PCB issue design flaw, and this statement suggests its BIOS/Firmware/etc (software/coding) based problems that apparently can't? be modified even with the classified tool. I'm trying to follow along here but there seems to be some conflicting opinions at play. Also, those who started receiving this second drop of Kingpin cards, you are all noticing the same issue? And why are people like Sajin not having this issue if he got one of the first cards/PCB designs? LVNeptune No one is going to be stupid enough to invalidate their warranty at this point. I think they were hoping people would request the no limit bios so they wouldn't have to RMA cards.
So you're saying that this is intentional by EVGA and the only way to get the full power out of these cards even on AIO (not even true LN2) is to void the warranty with some Leeroy Jenkins BIOS that forces more power? Sajin I ran the true xoc vbios on my card, and pushed 625w through it.
Can you be more specific? What XOC version? Where is this version? Not the one on the card? Is it a Kingpin specific XOC BIOS? Can we all agree to please be more specific when mentioning all of these variables? We are not all pro's here and spending the extra 10 seconds to be more definitive especially when you drop a "I pulled 625w" bombshell would be appreciated by all, especially those who are new to the game and trying to learn.
post edited by USMC1419 - 2020/12/18 18:39:56
12900k @ 5.2 W/Optimus Sig V2 64GB DDR5 @ 5000 RTX 3090 K|NGP|N W/ Optimus WB EVGA 1200w Supernova G2 Sammy 980 Pro 2Tb Predator 27" 1440p G-Sync IPS 165 hz Resident "wise guy"
|