antandbetty
iCX Member
- Total Posts : 318
- Reward points : 0
- Joined: 2014/12/01 19:13:42
- Status: offline
- Ribbons : 0
This changes EVERYTHING we thought we knew about water cooling our gpu's and there ASIC score..... Here is a quote from Kingpin that explains why the 980 kpe performed like crap under water.....unless EVGA bins the new kpe TI cards, they are not for water coolers any longer and just for LN2. We all knew something was wrong as no matter how many volts we threw at the 980 kpe they did not scale with clocks! I for one am extremely upset at wasting 1,600 dollars on two 980 kpe cards and addition 300 on water blocks for NO REASON. KPE is for LN2 only unless EVGA Bins them with ASIC of 80% or higher. Guess we will see if this is correct in a few days with the release of the TI KPE. "Honestly speaking, I think most end users don't even realize how maxwell gpus are voltage capped at ambient type cooling. I can tell by many of the comments at OC.net, elsewhere, and also here in these card XOC bios threads. Especially compared to kepler. KP 780ti scaled great on voltage with air/water temps. Basically, more voltage = more clocks no matter what temperature.With 980 and later gpus including titanX, the scaling on air/water has all but almost gone. I would say about 95% of all maxwell 980,titan-X, and 980ti gpus NO MATTER what vga brand pcb it is on, DO NOT SCALE with more voltage than 1.25v-1.275v at temps warmer than 25c or so. There is no magical bios that can effectively remove this.This is exactly why almost every moderate-good asic titanX, 980, and yes 980ti clock around 1550mhz MAX AVERAGE at say 45-60C loading temps.If you put 0c and colder on the card, you will see MUCH different behavior than what you see on air (green garbage all over screen when raising volts over 1.23-1.25vv or so)Cards with very good ASIC value (75% and up) will tend to have the most "overclocking", but just like about every other maxwell gpu, they cannot overvolt past 1.23v-1.25v.So highest asic cards like 80% +are almost always going to be the ones that can 1600+ on air/water, and again they do it pretty much WITHOUT overvolting over 1.23v-1.25v. Maxwell gpus with lower asic value like 65% will not be so great at air/water because these low asic gpus need voltage to scale compared to match the overclock of the high asic gpus( USING SAME USABLE VOLTAGE 1.23-1.25v)The bios's I posted basically allow you to set a higher voltage on air/water. Some gpus can scale more, some cant, some actually will NEED more voltage than was previously needed to run same frequency. All different.Have a better understanding now?" Source: http://kingpincooling.com....php?t=3659&page=6 EDIT to add another key post by Vince: "What I do know is that NV is causing a headache or two blocking ambient temp voltage scaling on non-reference card designs after 780ti." So according to the man NVIDIA'S design is blocking the scaling on ambient temps ... even on a unlocked custom card. This may be a characteristic of the Maxwell architecture PLEASE SEE KINGPIN'S Explanation of this issue @ http://forums.evga.com/FindPost/2362290 page 2 post # 47
post edited by antandbetty - 2015/07/21 10:29:01
"Freedom is never more than one generation away from extinction. We didn't pass it to our children in the bloodstream. It must be fought for, protected, and handed on for them to do the same, or one day we will spend our sunset years telling our children and our children's children what it was once like in the United States where men were free" R.R.
|
formula383
New Member
- Total Posts : 42
- Reward points : 0
- Joined: 2008/10/24 05:04:00
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 12:25:32
(permalink)
Thank you for posting this! Otherwise i may have never seen it. And this is why buying a non over clocked gpu is imo the way to go for cost effectiveness. It is really a shame they do not bin these gpu's for that kind of card and it also is damn silly the best card on the market is limited by a normal frame buffer it should have had 8GB.
CoSmos 2 case I7 3930k 1.425v @ 4.9 Gskill 32gig @ 2400 10-12-12-31 Asrock x79 extreme 11 LSI 9260-8i m4 512GB WD 4TB RE SAS x6 Seagate 600GB 15k SAS x2 Thermaltake 1350w 3 EVGA Titan X w EKblocks
|
the_Scarlet_one
formerly Scarlet-tech
- Total Posts : 24581
- Reward points : 0
- Joined: 2013/11/13 02:48:57
- Location: East Coast
- Status: offline
- Ribbons : 79
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 12:31:40
(permalink)
Too many people expect because money is higher that they undoubtedly will get better results.. that is foolish thinking and it has been all along. When you use a product as designed, you get better results. These were designed for extreme overclocking, and mixed with Maxwell, force people to push them properly rather than expecting to empty their wallet for the results of LN2 runs.. It's not logical to waste money on an item that is intended for extreme measures with no plans of ever using anything that would allow it to work properly.
I admit, I use all of my KPE cards under water.. i also have Ln2 pots for them, and I am too broke to get 50l of Ln2 to test them. When money becomes available, I will be pushing these cards properly. If the 980Ti classified and KPE display the issues the 980 had, I will laugh as people cry about it. I will wait, as others should!
|
ty_ger07
Insert Custom Title Here
- Total Posts : 21169
- Reward points : 0
- Joined: 2008/04/10 23:48:15
- Location: traveler
- Status: offline
- Ribbons : 270
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 12:32:37
(permalink)
Once again, this cements my opinion that Kingpin cards are for world record and benchmark enthusiasts only. I have had these debates in the past where owners said I was wrong to say that their Kingpin cards weren't designed to achieve a better gaming experience.
|
Vlada011
Omnipotent Enthusiast
- Total Posts : 10257
- Reward points : 0
- Joined: 2012/03/25 00:14:05
- Location: Belgrade-Serbia
- Status: offline
- Ribbons : 11
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 12:36:27
(permalink)
ty_ger07
Once again, this cements my opinion that Kingpin cards are for world record and benchmark enthusiasts only. I have had these debates in the past where owners said I was wrong to say that their Kingpin cards weren't designed to achieve a better gaming experience.
What that mean, for gamers best cards are Matrix, Galaxy, ... I hope Galaxy will design one model with black PCB I want to play games on high clock and you chose K|NGP|N for LN2, but maybe again Galaxy win first place same as with GTX780Ti. That card was for gaming and for LN2, and nobody couldn't move from first place 5 months in single and multi GPU testing. She was first long time and after GTX980 is launched.
post edited by Vlada011 - 2015/07/05 12:43:47
|
antandbetty
iCX Member
- Total Posts : 318
- Reward points : 0
- Joined: 2014/12/01 19:13:42
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 12:43:02
(permalink)
Scarlet-Tech
Too many people expect because money is higher that they undoubtedly will get better results.. that is foolish thinking and it has been all along. When you use a product as designed, you get better results. These were designed for extreme overclocking, and mixed with Maxwell, force people to push them properly rather than expecting to empty their wallet for the results of LN2 runs.. It's not logical to waste money on an item that is intended for extreme measures with no plans of ever using anything that would allow it to work properly.
I admit, I use all of my KPE cards under water.. i also have Ln2 pots for them, and I am too broke to get 50l of Ln2 to test them. When money becomes available, I will be pushing these cards properly. If the 980Ti classified and KPE display the issues the 980 had, I will laugh as people cry about it. I will wait, as others should!
When did we ever have a GPU that did not scale to voltage under water? We were always able to get higher clock with addition voltage and water cooling..... We didn't care about ASIC score as it used to be the lower the better for us water coolers. This changes everything we knew
"Freedom is never more than one generation away from extinction. We didn't pass it to our children in the bloodstream. It must be fought for, protected, and handed on for them to do the same, or one day we will spend our sunset years telling our children and our children's children what it was once like in the United States where men were free" R.R.
|
the_Scarlet_one
formerly Scarlet-tech
- Total Posts : 24581
- Reward points : 0
- Joined: 2013/11/13 02:48:57
- Location: East Coast
- Status: offline
- Ribbons : 79
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 12:46:18
(permalink)
My classified scaled when they were cold, using 0 degree ambit air to get them lower, but they never went over 1500, and when it would warm up to 20c ambient, they wouldn't hit 1450 because of thermal limits causing driver crashes. That was forcing 1.35v at 0c and 1.4 when the ambient went up to 20c. That is pretty crap scaling when people were getting much higher with Ln2 and nearly the same voltage. Ask yourself antandbetty, would you buy a top fuel dragster for day to day driving, to go light to light, or would you buy it for the drag strip? Think about the K|ngp|n card the same way and then apply it to this scenario. Maxwell obviously changed something, and people are asking for 1700mhz on the KPE thread for gaming.. Not going to happen on most cards. People need to get over their entitlement and start thinking realistically. As of right now, big numbers plus deep wallets have to mean the best gaming on ambient air/water possible. That is funny thinking right there.
post edited by Scarlet-Tech - 2015/07/05 12:49:01
|
antandbetty
iCX Member
- Total Posts : 318
- Reward points : 0
- Joined: 2014/12/01 19:13:42
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 12:50:57
(permalink)
Scarlet-Tech
Kingpin was built for benchmarks, not as a daily driver.... & everyone should know that. That said, his statement changes everything we knew about overclocking GPU's & ASIC scores under water.
"Freedom is never more than one generation away from extinction. We didn't pass it to our children in the bloodstream. It must be fought for, protected, and handed on for them to do the same, or one day we will spend our sunset years telling our children and our children's children what it was once like in the United States where men were free" R.R.
|
Muezick
iCX Member
- Total Posts : 305
- Reward points : 0
- Joined: 2015/02/10 17:58:59
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 12:53:08
(permalink)
I sure wish I knew what was being talked about in this thread. XD
I am new to Overclocking GPUs.
This all seems very interesting though.
|
DirtySouthWookie
Superclocked Member
- Total Posts : 188
- Reward points : 0
- Joined: 2014/10/19 19:25:36
- Status: offline
- Ribbons : 1
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 13:04:43
(permalink)
This is why I bought reference 980ti cards this time around. I have experience with 16 different 980s from Asus, MSI, gigabyte, and EVGA. None of the custom PCBs mattered, voltage didn't matter, water didn't matter, and neither did custom bios flashes. Maxwell is a different beast and Pascal will be even smaller so expect more of this.
My old 780ti SLI 4770k overvolt custom bios setup still smokes my 980 ti SLI 5820k setup in valley and Haven lol.
|
antandbetty
iCX Member
- Total Posts : 318
- Reward points : 0
- Joined: 2014/12/01 19:13:42
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 13:05:39
(permalink)
Muezick
I sure wish I knew what was being talked about in this thread. XD
I am new to Overclocking GPUs.
This all seems very interesting though.
Hi Muezick, It basically means that the statement in the screenshot below does not apply any longer to those water cooling there gpu's. 
"Freedom is never more than one generation away from extinction. We didn't pass it to our children in the bloodstream. It must be fought for, protected, and handed on for them to do the same, or one day we will spend our sunset years telling our children and our children's children what it was once like in the United States where men were free" R.R.
|
the_Scarlet_one
formerly Scarlet-tech
- Total Posts : 24581
- Reward points : 0
- Joined: 2013/11/13 02:48:57
- Location: East Coast
- Status: offline
- Ribbons : 79
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 13:49:14
(permalink)
antandbetty
Scarlet-Tech
Kingpin was built for benchmarks, not as a daily driver.... & everyone should know that. That said, his statement changes everything we knew about overclocking GPU's & ASIC scores under water.
If people ever believed that putting a card under water would net anywhere near the results of sub ambient temps, then their ideas were skewed slightly. Between Air and Water, we are talking about a 20c-40c difference. Chilled water adds another 15c off the top, for up to 55c difference.. Next stage down in Phase Change, which offers up to 100c difference. There is a huge gap there. I understand Vince's latest analogy shows that there is a gap. This is a new architecture and he even showed that he learnt something from this. I honestly would not, even the slightlest amount, blame EVGA if they were to deny RMA's due to cards not overclocking as well as people want when it is the Classified and K|ngp|n. "How did you test it?" -Evga "Air cooler and water block." "We built them for Extreme Temperatures, so please test it with extreme temperatures before returning it.." "What?! i didn't buy it for that..." "we built it for that and advertised it as such." Could you imagine if they inforced that how few people would jump the gun, and would wait for XOCers to test the card on air and sometimes water before putting LN2 pots or phase change to the cards?
|
SocioPC
iCX Member
- Total Posts : 421
- Reward points : 0
- Joined: 2006/06/03 19:12:05
- Status: offline
- Ribbons : 2
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 13:50:30
(permalink)
Disappointing is an understatement.
Classy is only $50 more than reference though.. and probably worth that extra cost even if all it does is minimize/remove coil whine.
|
IMPORTEKNIKAICR
Superclocked Member
- Total Posts : 208
- Reward points : 0
- Joined: 2008/03/26 17:34:15
- Status: offline
- Ribbons : 1
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 14:22:00
(permalink)
Scarlet-Tech
My classified scaled when they were cold, using 0 degree ambit air to get them lower, but they never went over 1500, and when it would warm up to 20c ambient, they wouldn't hit 1450 because of thermal limits causing driver crashes. That was forcing 1.35v at 0c and 1.4 when the ambient went up to 20c. That is pretty crap scaling when people were getting much higher with Ln2 and nearly the same voltage.
Ask yourself antandbetty, would you buy a top fuel dragster for day to day driving, to go light to light, or would you buy it for the drag strip? Think about the K|ngp|n card the same way and then apply it to this scenario.
Maxwell obviously changed something, and people are asking for 1700mhz on the KPE thread for gaming.. Not going to happen on most cards. People need to get over their entitlement and start thinking realistically. As of right now, big numbers plus deep wallets have to mean the best gaming on ambient air/water possible. That is funny thinking right there.
Um. Scarlet, does this mean I should hang on to my Titan Black SSCs?
CPU: Intel i7-5930K | Corsair H115i Platinum | MB: Asus Rampage V Extreme/U3.1 X99 | RAM: 128GB Corsair Dominator Platinum GPU: EVGA RTX 3090 K|NGP|N | DISPLAY: 3x LG 4K 32" | CASE: Coolermaster Cosmos C700M | PSU: EVGA SuperNova 1600 T2
|
hallowen
CLASSIFIED ULTRA Member
- Total Posts : 5644
- Reward points : 0
- Joined: 2008/06/18 15:38:00
- Location: In a Galaxy Far, Far Away...
- Status: offline
- Ribbons : 14
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 14:45:47
(permalink)
Scarlet-Tech
antandbetty
Scarlet-Tech
Kingpin was built for benchmarks, not as a daily driver.... & everyone should know that. That said, his statement changes everything we knew about overclocking GPU's & ASIC scores under water.
If people ever believed that putting a card under water would net anywhere near the results of sub ambient temps, then their ideas were skewed slightly. Between Air and Water, we are talking about a 20c-40c difference. Chilled water adds another 15c off the top, for up to 55c difference.. Next stage down in Phase Change, which offers up to 100c difference. There is a huge gap there. I understand Vince's latest analogy shows that there is a gap. This is a new architecture and he even showed that he learnt something from this.
I honestly would not, even the slightlest amount, blame EVGA if they were to deny RMA's due to cards not overclocking as well as people want when it is the Classified and K|ngp|n.
"How did you test it?" -Evga
"Air cooler and water block."
"We built them for Extreme Temperatures, so please test it with extreme temperatures before returning it.."
"What?! i didn't buy it for that..."
"we built it for that and advertised it as such."
Could you imagine if they inforced that how few people would jump the gun, and would wait for XOCers to test the card on air and sometimes water before putting LN2 pots or phase change to the cards?
Looks like I may be Taking a Wait and See attitude before deciding which Card (or even the Brand Name) I will be purchasing Next. Since Vince has made this statement, I have become somewhat delusional about my previous beliefs.
ASUS: Rampage VI Extreme | i9-7940X | 2X RTX 2080 Ti Kingpin SLI | 32GB DDR4 3200MHz Memory - SAGER: NP9870-G | i7-6700K | GTX 980M 8GB | 64GB DDR4 | 950 PRO M.2 512GB | 17.3 QHD 120Hz Matte G-Sync | Prema bios - EVOC Premamod: P870TM1 | i9-9900K-LM | RTX 2080N 8GB | Modded Vapor Chamber | 32GB 3000MHz Ripjaws | 960 EVO M.2 1TB | 17.3 3K QHD 120Hz Matte G-Sync | Intel 8265 -
|
Halo_003
Omnipotent Enthusiast
- Total Posts : 12859
- Reward points : 0
- Joined: 2009/03/20 18:18:10
- Status: offline
- Ribbons : 51
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 14:59:24
(permalink)
Scarlet-Tech
antandbetty
Scarlet-Tech
Kingpin was built for benchmarks, not as a daily driver.... & everyone should know that. That said, his statement changes everything we knew about overclocking GPU's & ASIC scores under water.
If people ever believed that putting a card under water would net anywhere near the results of sub ambient temps, then their ideas were skewed slightly. Between Air and Water, we are talking about a 20c-40c difference. Chilled water adds another 15c off the top, for up to 55c difference.. Next stage down in Phase Change, which offers up to 100c difference. There is a huge gap there. I understand Vince's latest analogy shows that there is a gap. This is a new architecture and he even showed that he learnt something from this.
This is not surprising though. Even back on X58 and the GTX 200 series we were seeing (in the Get It Colder threads & Kingpincooling) that you really had to be subzero to get high clockspeeds at the same voltage. For example, my 285 Classified worked the same way, caps out on air at about 800MHz, water at 825MHz, but dry ice takes it to 1000MHz at the same voltages, just 100C colder than water. Not the same tech I know, but we see generally the same effect. All this really means (at least how I interpret it) is that the colder you go the more the top end clockspeed is opened up at any given voltage, and that the difference between ambient and subzero is directly related to your ASIC score, and is limited by the 1.25V cap before the voltage becomes less of a factor.
i7 7700K - ASUS Maximus IX Apex - 16GB G.Skill TridentZ RGB 3466MHz C16 - 5700 XT 50th Anniversary
|
the_Scarlet_one
formerly Scarlet-tech
- Total Posts : 24581
- Reward points : 0
- Joined: 2013/11/13 02:48:57
- Location: East Coast
- Status: offline
- Ribbons : 79
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 15:10:08
(permalink)
IMPORTEKNIKAICR
Um. Scarlet, does this mean I should hang on to my Titan Black SSCs?
I am a hoarder. I would keep them simply because of the DP capability. It is really a choice for you though, as it is your money that you have to consider. You may be able to get nearly the same amount you would buy the 980ti KPE for, but it is all in preference in your case. If you are going to bench mark, even occasionally, I would suggest the KPE. If not, then I don't see the point of spending the money on them when a regular 980ti would do you well, and you could even net a few extra dollars back from selling the Titan's Everyone needs to realize, I am not trying to stop anyone from buy these cards... I am just trying to get them to buy them for the right reason with the right amount of knowledge as to why they are buying them.
|
antandbetty
iCX Member
- Total Posts : 318
- Reward points : 0
- Joined: 2014/12/01 19:13:42
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 17:14:09
(permalink)
I don't think some are getting the point.... Voltage means NOTHING on maxwell unless the gpu is frozen. This is not "normal" in the sense that addition voltage produces additional heat which requires better cooling. that was always the basis for overclocking. My 2 kpe's have never seen temps above 40c yet they can not get out of there own way even at 1.4 volts. That is the issue. There is no voltage cap on the KPE I don't think anyone here was expecting 1700-1800 on water BUT without a doubt we should have at least been able to hit 1600 on water. Even TIN stated that these cards are effectively "overheating" at way lower temps then he ever seen before. This was Kingpins response to a question i asked him.... "What I do know is that NV is causing a headache or two blocking ambient temp voltage scaling on non-reference card designs after 780ti." So according to the man NVIDIA is blocking the scaling even on a unlocked card.
post edited by antandbetty - 2015/07/06 13:21:19
"Freedom is never more than one generation away from extinction. We didn't pass it to our children in the bloodstream. It must be fought for, protected, and handed on for them to do the same, or one day we will spend our sunset years telling our children and our children's children what it was once like in the United States where men were free" R.R.
|
ty_ger07
Insert Custom Title Here
- Total Posts : 21169
- Reward points : 0
- Joined: 2008/04/10 23:48:15
- Location: traveler
- Status: offline
- Ribbons : 270
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 17:39:00
(permalink)
Blocking. Intentionally. LOL. Yeah right...
[sarcasm]I am sure NVIDIA built a hardware device to monitor temperature and cap frequency based on temperature and then provide a fake crash response if exceeded.[/sarcasm]
I am sure it is a hardware limitation based on design or manufacturing size; not a conspiracy. TiN is from Taiwan. I am sure something was lost in the translation.
|
antandbetty
iCX Member
- Total Posts : 318
- Reward points : 0
- Joined: 2014/12/01 19:13:42
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 17:56:02
(permalink)
The response from my question was from Kingpin not Tin.
"Freedom is never more than one generation away from extinction. We didn't pass it to our children in the bloodstream. It must be fought for, protected, and handed on for them to do the same, or one day we will spend our sunset years telling our children and our children's children what it was once like in the United States where men were free" R.R.
|
ty_ger07
Insert Custom Title Here
- Total Posts : 21169
- Reward points : 0
- Joined: 2008/04/10 23:48:15
- Location: traveler
- Status: offline
- Ribbons : 270
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 18:06:59
(permalink)
antandbetty
The response from my question was from Kingpin not Tin.
The limit cannot be overcome by BIOS as stated. So, you are telling me that there is a hardware-based device which monitors voltage, temperature, and frequency and creates a fake crash if exceeded? We know that there is a system called GPU Boost 2.0 which does monitor those things.... but do you really believe they programmed it to create fake crashes in order to purposely limit the product and purposely create an elite club? Seems very far fetched. Seems more like an artifact or limitation of the design instead of a planned limitation. Wouldn't it be easier to just cause GPU Boost 2.0 to cap the frequency instead of create fake crashes? I agree that the product requires extreme cooling to reach extreme numbers, but I find it hard to believe that NVIDIA is in the business of creating fake crashes. Every time manufacturing size is reduced, we get closer and closer to this observation. It seems more like a reality than a conspiracy. When was the last time you exceeded 3.5 volts on the core like could be done with ancient 90 nm GPU cores? As manufacturing size decreases, ability to handle voltage and temperature also decreases. This has been a long-standing trend.
post edited by ty_ger07 - 2015/07/05 18:14:49
|
Vlada011
Omnipotent Enthusiast
- Total Posts : 10257
- Reward points : 0
- Joined: 2012/03/25 00:14:05
- Location: Belgrade-Serbia
- Status: offline
- Ribbons : 11
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 18:11:44
(permalink)
I read this and I can say one thing only... I'm sad because I didn't bought TITAN Black. I can say that was biggest mistake in life and cost me lot of nerves but someone to explain me that GK110 is now better option than GM200 because some scaling with more voltage and I understand nothing... And you will see many models much much faster than GTX980Ti reference, you will see results, scores, much higher fps in games and everything else what is most important for 99% of people. I will be very sad if EVGA not offer GTX980 Classified ULTRA or Classified with Hybrid cooler with at least 50MHz faster than 980Ti Classified. Because even if they launch K|NGP|N with 1250MHz no one want to card not work him and someone to talk that such cards are not for games. What is solution if NVIDIA try to sabotage our fun and when only their earning is important not our fun What is solution...? I mean if really card not work good, but I can't see that on paper... But what than... R9-380X 8GB CF? Fury X... they have own theories that card is better than TITAN X...  I can't listen that... They even show that Fury X have much better picture than TITAN X and details in games on far distance are more visible than on NVIDIA cards. In that I believe, I told that to people constantly but when someone fall in love he is blind. But they try to explain that NVIDIA sabotage and graphic details from first day of their work company only to offer better performance than competition. They sacrifice about 10-15% in sharpness and picture is more blurry because better performance.
|
antandbetty
iCX Member
- Total Posts : 318
- Reward points : 0
- Joined: 2014/12/01 19:13:42
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 18:23:43
(permalink)
ty_ger07
antandbetty
The response from my question was from Kingpin not Tin.
The limit cannot be overcome by BIOS as stated. So, you are telling me that there is a hardware-based device which monitors voltage, temperature, and frequency and creates a fake crash if exceeded? We know that there is a system called GPU Boost 2.0 which does monitor those things.... but do you really believe they programmed it to create fake crashes in order to purposely limit the product and purposely create an elite club? Seems very far fetched. Seems more like an artifact or limitation of the design instead of a planned limitation. Wouldn't it be easier to just cause GPU Boost 2.0 to cap the frequency instead of create fake crashes?
I agree that the product requires extreme cooling to reach extreme numbers, but I find it hard to believe that NVIDIA is in the business of creating fake crashes.
Every time manufacturing size is reduced, we get closer and closer to this observation. It seems more like a reality than a conspiracy. When was the last time you exceeded 3.5 volts on the core like could be done with ancient 90 nm GPU cores? As manufacturing size decreases, ability to handle voltage and temperature also decreases. This has been a long-standing trend.
No, nothing to do with crashes... according to them just limiting the scaling of voltage at ambient temps as he stated which in turn would produce the poor under performance we all saw in the 980 kpe. many Vanilla 980's were achieving 1500 on stock volts yet a beast of card like the kpe could not hit 1600 on water regardless of applied volts.
"Freedom is never more than one generation away from extinction. We didn't pass it to our children in the bloodstream. It must be fought for, protected, and handed on for them to do the same, or one day we will spend our sunset years telling our children and our children's children what it was once like in the United States where men were free" R.R.
|
Vlada011
Omnipotent Enthusiast
- Total Posts : 10257
- Reward points : 0
- Joined: 2012/03/25 00:14:05
- Location: Belgrade-Serbia
- Status: offline
- Ribbons : 11
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 18:45:11
(permalink)
Every person who finish school will know that materials behave different on deep minus and on ambient temps.
It's nothing new if you can pull few MHz more on 1.5V on -150 than on ambient temp with 1.5V even if you keep card on 20-30-40C constantly, not more.
Last 10 years people could OC graphic cards 300-400MHz on ambient temps, nothing more, even if temps is 20C....
But on -150, -200 things are completely different and that influence and on steel not on silicon.
For 90% of people is most important that Maxwell behave as others cards in last 10 years, that mean people could pull 200-500MHz depend from architecture and 10 other factors.
|
ty_ger07
Insert Custom Title Here
- Total Posts : 21169
- Reward points : 0
- Joined: 2008/04/10 23:48:15
- Location: traveler
- Status: offline
- Ribbons : 270
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 18:48:57
(permalink)
antandbetty
ty_ger07
antandbetty
The response from my question was from Kingpin not Tin.
The limit cannot be overcome by BIOS as stated. So, you are telling me that there is a hardware-based device which monitors voltage, temperature, and frequency and creates a fake crash if exceeded? We know that there is a system called GPU Boost 2.0 which does monitor those things.... but do you really believe they programmed it to create fake crashes in order to purposely limit the product and purposely create an elite club? Seems very far fetched. Seems more like an artifact or limitation of the design instead of a planned limitation. Wouldn't it be easier to just cause GPU Boost 2.0 to cap the frequency instead of create fake crashes?
I agree that the product requires extreme cooling to reach extreme numbers, but I find it hard to believe that NVIDIA is in the business of creating fake crashes.
Every time manufacturing size is reduced, we get closer and closer to this observation. It seems more like a reality than a conspiracy. When was the last time you exceeded 3.5 volts on the core like could be done with ancient 90 nm GPU cores? As manufacturing size decreases, ability to handle voltage and temperature also decreases. This has been a long-standing trend.
No, nothing to do with crashes... according to them just limiting the scaling of voltage at ambient temps as he stated which in turn would produce the poor under performance we all saw in the 980 kpe. many Vanilla 980's were achieving 1500 on stock volts yet a beast of card like the kpe could not hit 1600 on water regardless of applied volts.
Let me quote what you quoted: ... If you put 0c and colder on the card, you will see MUCH different behavior than what you see on air (green garbage all over screen when raising volts over 1.23-1.25vv or so) ...
... What I do know is that NV is causing a headache or two blocking ambient temp voltage scaling on non-reference card designs after 780ti...
... There is no magical bios that can effectively remove this. ...
... So according to the man NVIDIA is blocking the scaling intentionally even on a unlocked card... What you said or at least repeated is that NVIDIA is intentionally blocking overclocking by purposely placing green garbage all over the screen if you try to overclock too high without subzero cooling. How are you not implying that NVIDIA is purposely creating crashes? In my opinion, it is an artifact of our progression towards smaller and smaller manufacturing size which is less and less able to cope with temperature and voltage the way we are used to. If this is not the case, in the very least, it would be much easier to believe (although there is no hard evidence for this theory either) that the ring bus is independently and internally voltage controlled and therefore does not "benefit" from core voltage increases and therefore equally limits all cards to the same general level unless kept cool in order to be stabilized at higher frequencies. Or, even less far-fetched (but also not proven), perhaps the PLL voltage makes quite a difference and since it is not adjustable by software, getting it colder is one solution.
post edited by ty_ger07 - 2015/07/05 19:00:59
|
clo007
CLASSIFIED ULTRA Member
- Total Posts : 5274
- Reward points : 0
- Joined: 2006/02/09 11:20:25
- Status: offline
- Ribbons : 11
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 19:06:57
(permalink)
Here is an idea...go buy a KPE 980/ti. Then buy a phase-change unit. Strap it to your KPE and game on.
|
Vlada011
Omnipotent Enthusiast
- Total Posts : 10257
- Reward points : 0
- Joined: 2012/03/25 00:14:05
- Location: Belgrade-Serbia
- Status: offline
- Ribbons : 11
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 19:08:39
(permalink)
I can say as person who want to buy cards with highest clock on market I will not buy EVGA K|NGP|N Graphic Cards any more. It's too much dust about that and weird explanation while other brands produce cards overclocked or same or higher speed and I didn't notice such statements as here. Nobody have so much money to risk with something what will not work as graphic cards from other brands and because high clock is still important for me I will continue with fabric overclocked cards as much possible... To be honest and I and market is sick of mistakes no matter who make them and to throwing ball in other yard. NVIDIA is guilty, EVGA is guilty, ASUS is guilty, nobody is guilty and customers at the end pay that with own money and read non sense only to time pass and to calm them. At the end no matter who is guilty end customer will pay that.
post edited by Vlada011 - 2015/07/05 19:14:08
|
ty_ger07
Insert Custom Title Here
- Total Posts : 21169
- Reward points : 0
- Joined: 2008/04/10 23:48:15
- Location: traveler
- Status: offline
- Ribbons : 270
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 19:19:08
(permalink)
Vlada011
I can say as person who want to buy cards with highest clock on market I will not buy EVGA K|NGP|N Graphic Cards any more.
It's too much dust about that and weird explanation while other brands produce cards overclocked or same or higher speed and I didn't notice such statements as here. The Kingpin cards are designed for subzero cooling for benchmark enthusiasts. They always have been. They are not designed for gamers to have the fastest gaming card. Never have been. Kingpin cards are designed to be received, have the stock cooler removed immediately, have a LN2 pot strapped to them, and be tortured and crashed over and over for hours until a world-record run is finally obtained. Kingpin cards will not run for a gamer -- in general -- any better than a normal card when air cooled. Never have. Most of the money EVGA puts into the Kingpin cards is spent on the voltage regulation section. If you can't use high voltage because you don't want to put a LN2 pot on it, there is no reason to buy a Kingpin card. I know I got into a big argument with an unhappy air-cooled Kingpin owner in the past... They simply did not want to believe that they made a poor (overpriced) purchasing decision when choosing an air-cooled card to game with. There just isn't anything very special about the GPU core on the Kingpin versus a regular card. They try to bin them a bit, but binning doesn't account for much. The extra money for the Kingpin cards is invested in the magic in the voltage regulation design and subzero controls.
post edited by ty_ger07 - 2015/07/05 19:57:57
|
clo007
CLASSIFIED ULTRA Member
- Total Posts : 5274
- Reward points : 0
- Joined: 2006/02/09 11:20:25
- Status: offline
- Ribbons : 11
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 19:52:27
(permalink)
ty_ger07 nailed it.
And eVGA even markets them accordingly.
|
Heini2
iCX Member
- Total Posts : 498
- Reward points : 0
- Joined: 2014/08/26 17:39:23
- Status: offline
- Ribbons : 0
Re: STATEMENT BY KINGPIN THAT PROVES 900 SERIES DOES NOT SCALE WITH VOLTAGE NO MATTER BRAN
2015/07/05 19:59:49
(permalink)
Vlada011...They sacrifice about 10-15% in sharpness and picture is more blurry because better performance... Vlada, I would like to thank you for pointing to this. I've seen only three people make this observation, me included, though I'm sure others have. OC'ing, mhz, fps, etc. are well enough but what's the point when one disregards quality for quantity? It reminds me of dogs that chase their tails. Also: Thanks to ty_ger07 for being the voice of reason.
post edited by Heini2 - 2015/07/05 20:02:57
Z170 Sabertooth Mk1 / i7-6700K + NH-D15 / Trident Z (2x16GB)@3000 / 2080Ti XC Ultra / Essence STX II / 950 Pro 512GB / G3 1000W / MG279Q / HAF XB EVO / Win10 X64 Pro
|