Look at this $1900 monster...

Why not watercool the VRMs too? Those things make some heat.
Before you say "it's because of LN2!!", is there a reason you cannot have a pot/whatever to LN2 cool the VRMs too? LN2 cool everything?
 
Be advised you can still get a dud chip with one of these as I did (reference the JayzTwoCents and GN OC'ing video where Vince says one of the kingpin cards isn't so good) Costly return for sure with the restocking fee, but no way I was hanging onto a $1900 card that couldn't even get past 2130 stable.

My 2080 ti XC Ultra can hit 2160 with the underequipped 120mm rad hybrid kit from EVGA and costs a heck of a lot less. Haven't flashed the GALAX bios either.
It wasnt a dud card it just wasnt a golden card. Those were 3 personal samples and still made it through the binning process and can clock what they are guaranteed to clock.


Why not watercool the VRMs too? Those things make some heat.
Before you say "it's because of LN2!!", is there a reason you cannot have a pot/whatever to LN2 cool the VRMs too? LN2 cool everything?
With waterblocks they are cooling the VRMs. Just not the stock asetek cooler because its not needed out of the box.
because those components dont produce enough heat to need to be cooled via LN2 and when you take them down to -125F will end up causing problems. For example. The memory isnt cooled on LN2 because it doesnt like cold. They actually run a heater on the back of the card to prevent moisture as well as prevent the memory from getting to cold.
 
If spending this much money, why wouldn't you just build a custom loop? Special binning maybe?

My highly overclocked Pascal Titan X can sit at 24C full load if I crank the fans up on a cool day. More typically I have the fans on auto to keep load temp under 40C so it doesn't get too loud.

Thing is, at $1900, you aren't that far off from an RTX Titan...

This seems like the stuff bad decisions are made of.

Kingpin cards, to my knowledge, are the very top bin of GPUs EVGA receives. The custom VRM on them is also designed to deliver a mammoth and clean amount of power for sub-ambient overclocking, with a good amount of extra probe points easily marked and accessible for accurate voltage readings.

these are definitely niche cards, they're meant for competitive LN2 overclocking. So the funny thing is the intended use is for you to just strip that stock cooler off (even though KPE cards usually have heavy copper heatsinks and perform very well) and install your ln2 pot.

Definitely not worth $1900 if you just want to keep it running ambient, a water cooled founders edition will likely hit the same or even better clocks depending on luck.
 


"Look at this $1900 monster..." = Barnum & Baily's
egress.jpg


 

Attachments

  • egress.jpg
    egress.jpg
    33.8 KB · Views: 17
It wasnt a dud card it just wasnt a golden card. Those were 3 personal samples and still made it through the binning process and can clock what they are guaranteed to clock.

Their extra binning process should of made them almost golden on a $1900 card, so I don't agree with you at all. They definitely toned down the binning language in advertising materials this time as opposed to older KP releases, and the only real confirmation came from TiN himself.

The average OC I was able to glean from EVGAs forums seemed to be 2160-2175 (which I would have been fine with), with a very lucky few getting 2190 and 2205. 2130 was definitely at the lower end and should have been relegated to FTW3 cards.
 
So 2100 to 2200Mhz, that would put these cards about 10% higher performance than the 2080Ti's with nice aftermarket coolers right? Those get what 1900 to 2000mhz, with factory boost clocks of 1755, 1770 or 1815 (The Zotac, but they OC their chips a bit too hard imho), which then self boost another 100 to 200 Mhz?

Someone was saying earlier in the thread that these would get beaten in less than a year by the next card that comes out, but with the way the increases have slowed down lately (silicon speed improvements), and estimating these as being 10% faster, I wouldn't be surprised if the next gen cards get a 10% gain over the 2080Ti, which would make them equal.

(Obviously some rough generalizations about performance increases).

Now for me that doesn't justify spending the money now, cause that really is up there. But I would expect someone to easily get 2 years, maybe 3 out of one of these.
 
Their extra binning process should of made them almost golden on a $1900 card, so I don't agree with you at all. They definitely toned down the binning language in advertising materials this time as opposed to older KP releases, and the only real confirmation came from TiN himself.

The average OC I was able to glean from EVGAs forums seemed to be 2160-2175 (which I would have been fine with), with a very lucky few getting 2190 and 2205. 2130 was definitely at the lower end and should have been relegated to FTW3 cards.
the point i was trying to make was those werent cards pulled off the line. Those were ES cards that vince had. Two were golden and one was average. I dont really care if you agree with what happened in that video or not.
 
the point i was trying to make was those werent cards pulled off the line. Those were ES cards that vince had. Two were golden and one was average. I dont really care if you agree with what happened in that video or not.

I don't really care what you think either. I was just saying dud cards can happen, and it's not worth rolling the dice at $1900. I don't really care if they were ES samples or not, it happens in both scenarios, so your point is?
 
I don't really care what you think either. I was just saying dud cards can happen, and it's not worth rolling the dice at $1900. I don't really care if they were ES samples or not, it happens in both scenarios, so your point is?
lol k

Anyway Steponz is streaming live again on LN2 for his 2080ti KP

 
If spending this much money, why wouldn't you just build a custom loop? Special binning maybe?

My highly overclocked Pascal Titan X can sit at 24C full load if I crank the fans up on a cool day. More typically I have the fans on auto to keep load temp under 40C so it doesn't get too loud.

Thing is, at $1900, you aren't that far off from an RTX Titan...

This seems like the stuff bad decisions are made of.

You are absolutely right with your view.. But for me, I want a water cooled solution on a card that is highly over clocked, and I don't want to put the work into making it happen. This kind of product is the kind of product I look for. All the guess work is done, you just plug it in, adjust some basic tweaks out the box, and it just works for the next 4 years. I'm fine with paying a premium.
 
This is a boutique product for a specific crowd, a really small crowd at that. I can appreciate the serious nature of the card even though it is not my cup of tea these days.
 
Awesome construction on this card for sure and I really enjoy watching Kingpin's liquid nitrogen shenanigans, but unless that's what you are doing,
I can't really see the point outside of massive E-Peen.

Paying close to 100% premium over a standard 2080Ti for a 15% improvement in frame rate, at best, doesn't seem like money well spent.
 
Awesome construction on this card for sure and I really enjoy watching Kingpin's liquid nitrogen shenanigans, but unless that's what you are doing,
I can't really see the point outside of massive E-Peen.

Paying close to 100% premium over a standard 2080Ti for a 15% improvement in frame rate, at best, doesn't seem like money well spent.

And considering you aren't going to get anywhere near 15% more performance over overclocking a "Regular " 2080Ti. Especially with the lottery. You can put all the ambient cooling and power section/voltage you want into a card, but if the silicon doesn't care it won't matter.
 
...Paying close to 100% premium over a standard 2080Ti for a 15% improvement in frame rate, at best, doesn't seem like money well spent.

Watched the video where TechJesus was at EVGA with Kingpin, and they went thru overclocking. The max OC was 2220. Then they overclocked a high end 2080Ti (1770 boost clock, several from different aib's fall in this range), and were comparing game performance. They had goofed on the memory oc, but even still were getting 7.5% to 10% higher performance than the top end 'standard' 2080ti. This go for abour $1249 to $1399 (Or $1499 direct from EVGA... wth with evgas' prices on their own website anyway?).

So yeah its pricey, but not double. 36% over evga's prices, 42% ish from the next tier down, or about 50% if you find a good deal. That's not even close to 100% price premium.

And considering you aren't going to get anywhere near 15% more performance over overclocking a "Regular " 2080Ti. Especially with the lottery. You can put all the ambient cooling and power section/voltage you want into a card, but if the silicon doesn't care it won't matter.

Well, 7 to 10% isn't nothing, and that's oc to oc comparison. But yeah, this card isn't going to be worth it to most.

I wonder how many of these have even been made? Someone will probably win the silicon lottery and get one that does 2300Mhz, if the number manufactured gets high enough.
 
My 2175 MHz RTX Titan+EK block+shunt mod would rape this thing.
To be fair, that card with the water cooling parts costs less than twice as much as the K|NGP|N. But, damn, shunt mod on a $3k video card. You are a mad man.
 
2625Mhz!! Holy shit! That guy uses a kingpin card pretty sure. Wish there was a way to tell different models of 2080ti's on that page, I wonder how many of those are Kingpins...
 
I don't do synthetic benchmarks anymore. Tweaking drivers/benchmarks/BIOS's/hooking up air cons to radiators etc to get an irrelevant number just a little bit higher is a waste of time. It's too contrived. Just like LN2 runs, has zero affect on any actual usage of a PC.

All I care about is if under ambient temps my 2175 MHz RTX Titan and 5.4 GHz 9900K can run PUBG and EvE Online at 4K/144 FPS min 24/7... which it does.
 
Watched the video where TechJesus was at EVGA with Kingpin, and they went thru overclocking. The max OC was 2220. Then they overclocked a high end 2080Ti (1770 boost clock, several from different aib's fall in this range), and were comparing game performance. They had goofed on the memory oc, but even still were getting 7.5% to 10% higher performance than the top end 'standard' 2080ti. This go for abour $1249 to $1399 (Or $1499 direct from EVGA... wth with evgas' prices on their own website anyway?).

So yeah its pricey, but not double. 36% over evga's prices, 42% ish from the next tier down, or about 50% if you find a good deal. That's not even close to 100% price premium.



Well, 7 to 10% isn't nothing, and that's oc to oc comparison. But yeah, this card isn't going to be worth it to most.

I wonder how many of these have even been made? Someone will probably win the silicon lottery and get one that does 2300Mhz, if the number manufactured gets high enough.


My logic is often fatally flawed but AFAIK cheapest 2080ti is around $999 if you can get one which puts the premium at 90% more for this card.
 
My logic is often fatally flawed but AFAIK cheapest 2080ti is around $999 if you can get one which puts the premium at 90% more for this card.

True, but that is a different card. This has a different PCB, vrm layout, binned GPU, watercooling, et.al.

Never said they weren't expensive, never said there was not a cheaper option.

It should have been clear in the first post of the thread from the words "this is a halo product for sure", that most will not be in the market for one of these.
As sold it's almost 30% faster clocks for 90% more price, never said otherwise.
What I disputed was "100% price increase for 15% performance" which isn't accurate. Implying it is barely better is also not accurate.

I'm not going to buy one, but it is still a cool as shit card, and fun to watch it go!
 
I don't do synthetic benchmarks anymore. Tweaking drivers/benchmarks/BIOS's/hooking up air cons to radiators etc to get an irrelevant number just a little bit higher is a waste of time. It's too contrived. Just like LN2 runs, has zero affect on any actual usage of a PC.

All I care about is if under ambient temps my 2175 MHz RTX Titan and 5.4 GHz 9900K can run PUBG and EvE Online at 4K/144 FPS min 24/7... which it does.

I agree with the above. There absolutely nothing wrong with tweaking and tuning your setup on a regular basis if that is your hobby and it makes you happy. That kind of stuff just isn't for me. I don't have an RTX Titan, but after doing computer all freaking day in my day job.. When I get home, all I want to do is move my mouse around, unlock my machine, and play my games with zero fumbling around to make them work.
 
The difference is he came in and said his card would rape the 2080ti KP. The best way to test that is via synthetic benchmarks. But when that statement was challenged he came up with some lame ass excuse and said all he cares about is running PUBG and EvE Online at 4k/144 lol. Thats why i attached a screenshot of the benchmark showing a RTX Titan running the same clock he is and its obviously NOT raping this card at all. :)

I bench quite a bit to measure overclock performance gains/losses to see what is worth it and what is not. I dont tweak drivers, the benchmarks, nor the BIOSs or hook up to air conditioners.
 
Actually the best way to test that is to run actual game benchmarks. That is what matters, not synthetic. In my head to head competition of my water-cooled RTX Titan versus my water-cooled 2150 MHz 2080Ti in a suite of GAMES, the RTX Titan was on average 9.5% faster. No way in hell a Kingpin card which is barely getting any higher clocks than my old reference 2080Ti with EK block is coming anywhere close to closing that delta. Especially with some crappy AIO.

Real world use is where the rubber meets the road.

And it's funny to post that #12 qrigs 2175 MHz RTX Titan run. Every card faster than that is running LN2 and 300-500 or more MHz core.

You basically proved my point and didn't even realize it. My card on ambient water 24/7 clocks is keeping up with LN2 2080 Ti runs.
 
Last edited:
You just realized a slower card under LN2 can be faster than a faster card not under LN2. You win a cookie.
 
You going into a thread showing off that your top tier card can beat a lower tier card. You win nothing as you just wanted to flex your epeen.
 
Why not watercool the VRMs too? Those things make some heat.
Before you say "it's because of LN2!!", is there a reason you cannot have a pot/whatever to LN2 cool the VRMs too? LN2 cool everything?

It is because of LN2. VRMs do not operate well frozen. You want them cooled not frozen.
 
Vega doesn't flex his Epeen. Everyone here knows Vegas Epeen is huge. He regularly ruins monitors to see if he can get rid of the matte coatings on them. Just for that 1% more of image quality.
 
2625Mhz!! Holy shit! That guy uses a kingpin card pretty sure. Wish there was a way to tell different models of 2080ti's on that page, I wonder how many of those are Kingpins...

They are using nitrogen. I watched Kingpin making the card run at that speed. Was on nitro.

I run my 2080ti at 2100mhz and a 1000mhz ram OC under liquid. Evga card.

There is NOTHING I would gain bu buying a Kingpin card. NOTHING.

My 2080ti can damn near max out my 240hz panel even if its 1080p. My 3440x1440 gsync is maxed out all the time too. I dont even own 4k because 3440x1440 is so many pixels already 4k is a waste of money for me right now.

Shit even my 5700xt, waiting on corsair h2o block availability, can max out 3440 in many cases.

So a 1900 gpu is 1000% electronic dick girth

But it amazing and exciting to see barriers pushed and these kingpin cards are so dsmn beautiful looking I'd buy one just to put in a shadow box on the wall.
 
Last edited:
  • Like
Reactions: N4CR
like this
Back
Top