EVGA

Is NV going to allow AIB cards with no power limits?

Author
AnonymousGuy
iCX Member
  • Total Posts : 452
  • Reward points : 0
  • Joined: 2014/09/17 21:15:46
  • Status: offline
  • Ribbons : 2
2020/09/24 16:11:53 (permalink)
(As we're learning) it seems the 3080 and 3090 are both limited to "turn power slider up 15%....that's all you get".  It's so bad that aside from VRAM it seems the 3080 and 3090 OC vs. OC have very similar <10% performance gap, both banging off the same power limits.  Starting several generations ago NV started banning AIB's from unlocked voltage so even top-tier cards like Kingpin were basically the exact same performance as the reference cards: here's a 24 phase 9000W capable PCB costing 3x the reference that you have to hard mod if you want to actually utilize.
 
So is this going to be a generation where the only difference between all the cards is the cooler?

Workstation:  12900KF @ 5.2Ghz || MSI Pro-A Z690 DDR4|| EVGA FTW3 3090 1000W || G.Skill 3866 4x8GB || Corsair AX1200i || whole-house loop.
Mining Box: HP Prodesk G1 (Haswell 4590), 3x Gigabyte 3080Ti, AX1500i @ 240V.
LANRig/GuestGamingBox: 9900nonK || Gigabyte Z390 Master || ASUS TUF 3090 650W shunt || Corsair SF600 || CPU+GPU watercooled 280 rad pull only || whole-house loop.
Server Router (Untangle): 8350K @ 4.7Ghz || ASRock Z370 ITX || 2x8GB 3200 || PicoPSU 250W, running on AX1200i || whole-house loop.
Server Compute/Storage: 10850K @ 5.1Ghz || Gigabyte Z490 Ultra || EVGA 3060 || LSI 9280i-24 port || 4TB Samsung 860 Evo, 5x10TB Seagate Enterprise Raid 6, 4x8TB Seagate Archive Backup ||  whole-house loop.
Laptop: HP Elitebook 840 G8 (Intel 1185G7) + 3060 Thunderbolt Dock, Razer Blade Stealth 13" 2017 (Int
#1

12 Replies Related Threads

    sirien
    New Member
    • Total Posts : 48
    • Reward points : 0
    • Joined: 2007/03/29 20:33:06
    • Status: offline
    • Ribbons : 0
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/24 16:35:21 (permalink)
    Product segmentation and warranty are reasons for the power limits. If an AIB card is already letting you have 400 watts, what if your case is trash, a hotbox, live in a hot climate etc. It’s better for them to limit the card and have no RMA’s than it is to let you have more performance. I’m surprised they even went this high for out of the box performance. Hard mods are the answer to both sides. You get your power limits, they get their asses covered by your voided warranty. You know very well if someone spending 2000$ on a card gets an XOC bios and blows it up, they are going to try to front the bill to evga warranty. The cards are also rediculously easy to hard mod, don’t rock the boat, they could have locked it down if they wanted to on a hardware level but they didn’t. Before you know it whiners will get them to lock down hardmods.
    post edited by sirien - 2020/09/24 16:37:58
    #2
    Sajin
    EVGA Forum Moderator
    • Total Posts : 49165
    • Reward points : 0
    • Joined: 2010/06/07 21:11:51
    • Location: Texas, USA.
    • Status: online
    • Ribbons : 199
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/24 16:41:13 (permalink)
    Uh what? The 2080 ti kingpin allowed like 2000w on the xoc vbios. Will probably be the same this time around too. The kingpin also allows you to up the voltage past nvidia built in limits. Buy kingpin if you want a fully unlocked card.
    #3
    Chamidorix
    New Member
    • Total Posts : 36
    • Reward points : 0
    • Joined: 2020/09/24 05:24:12
    • Status: offline
    • Ribbons : 1
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/24 17:24:47 (permalink)
    We only got the Kingpin XOC vBios from TiN. With him gone, I think it is a tad premature to assume EVGA will freely release it to the public again like they did with Turing. I'm sharpening up my soldering skills, anyways. 
    #4
    Sajin
    EVGA Forum Moderator
    • Total Posts : 49165
    • Reward points : 0
    • Joined: 2010/06/07 21:11:51
    • Location: Texas, USA.
    • Status: online
    • Ribbons : 199
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/24 17:33:40 (permalink)
    Pretty sure we can still get it from Vince.
    #5
    AnonymousGuy
    iCX Member
    • Total Posts : 452
    • Reward points : 0
    • Joined: 2014/09/17 21:15:46
    • Status: offline
    • Ribbons : 2
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/24 18:18:28 (permalink)
    Chamidorix
    We only got the Kingpin XOC vBios from TiN. With him gone, I think it is a tad premature to assume EVGA will freely release it to the public again like they did with Turing. I'm sharpening up my soldering skills, anyways. 


    Are you referring to shunt mods?  Derbauer already tried and didn't get very far...I think the speculation is the VRM controller is monitoring the currents internally as well as with the shunt.

    Workstation:  12900KF @ 5.2Ghz || MSI Pro-A Z690 DDR4|| EVGA FTW3 3090 1000W || G.Skill 3866 4x8GB || Corsair AX1200i || whole-house loop.
    Mining Box: HP Prodesk G1 (Haswell 4590), 3x Gigabyte 3080Ti, AX1500i @ 240V.
    LANRig/GuestGamingBox: 9900nonK || Gigabyte Z390 Master || ASUS TUF 3090 650W shunt || Corsair SF600 || CPU+GPU watercooled 280 rad pull only || whole-house loop.
    Server Router (Untangle): 8350K @ 4.7Ghz || ASRock Z370 ITX || 2x8GB 3200 || PicoPSU 250W, running on AX1200i || whole-house loop.
    Server Compute/Storage: 10850K @ 5.1Ghz || Gigabyte Z490 Ultra || EVGA 3060 || LSI 9280i-24 port || 4TB Samsung 860 Evo, 5x10TB Seagate Enterprise Raid 6, 4x8TB Seagate Archive Backup ||  whole-house loop.
    Laptop: HP Elitebook 840 G8 (Intel 1185G7) + 3060 Thunderbolt Dock, Razer Blade Stealth 13" 2017 (Int
    #6
    ty_ger07
    Insert Custom Title Here
    • Total Posts : 21171
    • Reward points : 0
    • Joined: 2008/04/10 23:48:15
    • Location: traveler
    • Status: offline
    • Ribbons : 270
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/24 21:44:27 (permalink)
    AnonymousGuy
    Chamidorix
    We only got the Kingpin XOC vBios from TiN. With him gone, I think it is a tad premature to assume EVGA will freely release it to the public again like they did with Turing. I'm sharpening up my soldering skills, anyways. 


    Are you referring to shunt mods?  Derbauer already tried and didn't get very far...I think the speculation is the VRM controller is monitoring the currents internally as well as with the shunt.

    I wholeheartedly disagree. Der8auer shunt modded only 1 input as a proof of concept to show how doable it is, but it was 11 PM and he had to go to bed and work on 3090 reviews the next morning, so he said that he would pick it up again later. WAIT until he picks it up later and publishes his results and opinions before drawing conclusions about his preliminary testing.
    The reason why he didn't "get very far" was because he only modded 1 input at that point. He needed to mod the other 2 inputs before the card would actually respond well to the mod, because otherwise power balancing would be messed up.
    I don't know how you drew the conclusion about internal VRM current measurement, because he never mentioned it and never mentioned that he had any concerns or disappointment about how the card was responding to the mod. What he showed is that it is clearly doable and it does respond positively to the mod.

    ASRock Z77 • Intel Core i7 3770K • EVGA GTX 1080 • Samsung 850 Pro • Seasonic PRIME 600W Titanium
    My EVGA Score: 1546 • Zero Associates Points • I don't shill

    #7
    stroker425
    Superclocked Member
    • Total Posts : 101
    • Reward points : 0
    • Joined: 2020/09/23 13:41:58
    • Location: NYC
    • Status: offline
    • Ribbons : 0
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/25 08:40:36 (permalink)
    Not anytime soon. Also, no source, just speculation.
    #8
    mchang1984
    Superclocked Member
    • Total Posts : 114
    • Reward points : 0
    • Joined: 2018/06/07 21:35:19
    • Status: offline
    • Ribbons : 0
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/25 12:41:54 (permalink)
    Last generation there were only 2 cards without power limits.
     
    One was EVGA's 2080 Ti Kingpin
    Second was Galax's Hall of Fame RTX 2080 Ti Watercooled OC Lab Edition out of which 100 were made in a limited production run.
     
    It was really easy to get the XOC VBios for the Kingpin.
    For the Galax Hall of Fame, customers that purchased the card had to LN2 Overlock with a GPU pot and post their results to HWBOT. The highest scores were invited to the Galax Overclocking Carnival in Vietnam where competitors were given the unlimited power VBios.
     
    So comparing the difficult between the two, the EVGA Kingpin was easier because no need to compete, you just download it and you are good to go. 
    #9
    Chamidorix
    New Member
    • Total Posts : 36
    • Reward points : 0
    • Joined: 2020/09/24 05:24:12
    • Status: offline
    • Ribbons : 1
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/25 15:51:00 (permalink)
    It is really silly the misinformation here.
     
    The only place to ever publicly get the unlocked power limit vBios for the 2080ti KPE edition has been on TiN's Kingpin overclocking article at xdevs: xdevs.com /guide /2080ti_kpe /#cbios
     
    TiN himself has stated he will no longer be taking the money and resources to create these articles at xdev: 
    "
    Sadly I wouldn't be able to continue guides and teardowns of PC hardware on xDevs.com, which I enjoyed a great deal, as there is no budget for this. Like some famous people say "Want to learn something well - try to explain and teach it"... But its one thing to use some ES lab samples to make a content, and completely different to spend thousands for one article.
    "
     
    We do not have a definitive answer how it will be handled with ampere. But as I said it seems very premature to assume EVGA will cater to the extreme niche enthusiast, outside of overclocking events (AKA marketing), without TiN's self described effort in providing resources for it. 
     
    Annendum, we only have access to the 2080 ti Galax HOF unlocked bios via a fortunate leak by a friend of a participant of the HOF overclocking event. A writer at techpowerup managed to get in contact with this friend and at some point the XOC bios was uploaded first surreptitiously to the unverified VGA TPU vbios archive, and then later to the verified section. It is completely possible that a chanced leak like this will not happen with Ampere. 
    post edited by Chamidorix - 2020/09/25 15:54:41
    #10
    Sajin
    EVGA Forum Moderator
    • Total Posts : 49165
    • Reward points : 0
    • Joined: 2010/06/07 21:11:51
    • Location: Texas, USA.
    • Status: online
    • Ribbons : 199
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/25 17:01:23 (permalink)
    I was able to get an xoc vbios for my 1080 ti kingpin by e-mailing Vince. Others were also able to do the same. 😁
    #11
    ty_ger07
    Insert Custom Title Here
    • Total Posts : 21171
    • Reward points : 0
    • Joined: 2008/04/10 23:48:15
    • Location: traveler
    • Status: offline
    • Ribbons : 270
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/27 08:45:16 (permalink)
    ty_ger07
    AnonymousGuy
    Chamidorix
    We only got the Kingpin XOC vBios from TiN. With him gone, I think it is a tad premature to assume EVGA will freely release it to the public again like they did with Turing. I'm sharpening up my soldering skills, anyways. 


    Are you referring to shunt mods?  Derbauer already tried and didn't get very far...I think the speculation is the VRM controller is monitoring the currents internally as well as with the shunt.

    I wholeheartedly disagree. Der8auer shunt modded only 1 input as a proof of concept to show how doable it is, but it was 11 PM and he had to go to bed and work on 3090 reviews the next morning, so he said that he would pick it up again later. WAIT until he picks it up later and publishes his results and opinions before drawing conclusions about his preliminary testing.
    The reason why he didn't "get very far" was because he only modded 1 input at that point. He needed to mod the other 2 inputs before the card would actually respond well to the mod, because otherwise power balancing would be messed up.
    I don't know how you drew the conclusion about internal VRM current measurement, because he never mentioned it and never mentioned that he had any concerns or disappointment about how the card was responding to the mod. What he showed is that it is clearly doable and it does respond positively to the mod.


    Der8auer recently released this follow up video.



    The shunt mod is definitely an option and definitely works. What you stated and presumed is far off in left field and is not anything that he ever said.

    ASRock Z77 • Intel Core i7 3770K • EVGA GTX 1080 • Samsung 850 Pro • Seasonic PRIME 600W Titanium
    My EVGA Score: 1546 • Zero Associates Points • I don't shill

    #12
    kevinc313
    CLASSIFIED ULTRA Member
    • Total Posts : 5004
    • Reward points : 0
    • Joined: 2019/02/28 09:27:55
    • Status: offline
    • Ribbons : 22
    Re: Is NV going to allow AIB cards with no power limits? 2020/09/27 13:02:22 (permalink)
    ty_ger07
    ty_ger07
    AnonymousGuy
    Chamidorix
    We only got the Kingpin XOC vBios from TiN. With him gone, I think it is a tad premature to assume EVGA will freely release it to the public again like they did with Turing. I'm sharpening up my soldering skills, anyways. 


    Are you referring to shunt mods?  Derbauer already tried and didn't get very far...I think the speculation is the VRM controller is monitoring the currents internally as well as with the shunt.

    I wholeheartedly disagree. Der8auer shunt modded only 1 input as a proof of concept to show how doable it is, but it was 11 PM and he had to go to bed and work on 3090 reviews the next morning, so he said that he would pick it up again later. WAIT until he picks it up later and publishes his results and opinions before drawing conclusions about his preliminary testing.
    The reason why he didn't "get very far" was because he only modded 1 input at that point. He needed to mod the other 2 inputs before the card would actually respond well to the mod, because otherwise power balancing would be messed up.
    I don't know how you drew the conclusion about internal VRM current measurement, because he never mentioned it and never mentioned that he had any concerns or disappointment about how the card was responding to the mod. What he showed is that it is clearly doable and it does respond positively to the mod.


    Der8auer recently released this follow up video.



    The shunt mod is definitely an option and definitely works. What you stated and presumed is far off in left field and is not anything that he ever said.



    https://youtu.be/5AA7GlV-xis?t=702
     
    So with the shunt and external vcore control, he's putting 503w into the card and 260w into the chip, to get clocks (sustained?) at 2,100mhz on an air cooler.   To get a 10% improvement in TSE?  I guess that's pretty good for no other changes? 
     
    The 3090 Strix has a 480w bios.
     
    #13
    Jump to:
  • Back to Mobile