[GamersNexus] Building a GTX 1080 Hybrid P 3: Performance Results (Overclocking 1080)

Azix

Golden Member
Apr 18, 2014
1,438
67
91
Gains with OC not so great. Interesting point about needing longer periods of testing to be sure an overclock is actually beneficial beyond benchmarks. Wonder how these 2.5Ghz claims for water-cooled cards will turn out since it wasn't stable with games just under 2200Mhz. Could be down to the single power connector. Another reason to not get a founders edition card as it's going to be gimped even if you put on a water cooling kit

https://www.youtube.com/watch?v=w-3fi1ovAP0

http://www.gamersnexus.net/guides/2441-diy-gtx-1080-hybrid-thermals-100-percent-lower-higher-oc-room

gtx-1080-ghybrid-doom-4k.png


gtx-1080-ghybrid-doom-1440.png


gtx-1080-ghybrid-mordor-4k.png


gtx-1080-ghybrid-gta-4k.png
 
Last edited:

exar333

Diamond Member
Feb 7, 2004
8,518
8
91
The results actually look promising. The scaling between 1733 and 2050 are very solid. If we remove power delivery (more than a single 8-pin) + cooling ability (replace crappy OEM cooler) it should be telling on what the true ceiling is.

Using just the FE, I don't think it provides enough data to say how good (or bad) custom solutions may be. For reference, the 980Ti reference 'only' OC'd 25%. This is around 18%. Historically, that's pretty solid for the standard OEM card to OC to.

Seems to be a lot of conclusions being made, both positve and negative, on a real lack of evidence so far.
 

Det0x

Golden Member
Sep 11, 2014
1,422
4,818
136
. Wonder how these 2.5Ghz claims for water-cooled cards will turn out since it wasn't stable with games just under 2200Mhz

~2050 / ~1783* = ~14% max overclock on stock founders card

~2164 / ~1783* = ~21% max overclock on watercooled founders 1080 card

2500 / ~1783* = needs a 40% max overclock... Unlikely

* = TPU actually shows the average clock of 1080 as 1783 MHz
 
Last edited:

4K_shmoorK

Senior member
Jul 1, 2015
464
43
91
Yeah you are right, these OC results look pretty awful. Throw it in the trash. No async compute? Useless.

Threadcrapping and trolling are not allowed
Markfw900
 
Last edited by a moderator:

railven

Diamond Member
Mar 25, 2010
6,604
561
126
The results actually look promising. The scaling between 1733 and 2050 are very solid. If we remove power delivery (more than a single 8-pin) + cooling ability (replace crappy OEM cooler) it should be telling on what the true ceiling is.

Using just the FE, I don't think it provides enough data to say how good (or bad) custom solutions may be. For reference, the 980Ti reference 'only' OC'd 25%. This is around 18%. Historically, that's pretty solid for the standard OEM card to OC to.

Seems to be a lot of conclusions being made, both positve and negative, on a real lack of evidence so far.

Just makes the wait for custom PCB cards more painful.

Kudos to Gamer Nexus for throwing an EVGA Hybrid on to it. Always good to see out of the box reviews. There was another review site that did an Accelero IV too. Kudos to them too.
 

Azix

Golden Member
Apr 18, 2014
1,438
67
91
Yeah you are right, these OC results look pretty awful. Throw it in the trash. No async compute? Useless.

nobody said the actual OC was bad.

1. question of 2.5Ghz claims
2. so-so scaling. or at least fps gains
 

4K_shmoorK

Senior member
Jul 1, 2015
464
43
91
nobody said the actual OC was bad.

1. question of 2.5Ghz claims
2. so-so scaling. or at least fps gains

Maybe reserve judgement on OC capabilities until a hard launch? It seems there's a new thread every couple hours in the VC&G section discussing pretty much the same thing is all I'm saying.

Just a lurker going back to lurking now. VC&G still as circular in its discourse as ever.
 

MrTeal

Diamond Member
Dec 7, 2003
3,901
2,630
136
The results actually look promising. The scaling between 1733 and 2050 are very solid. If we remove power delivery (more than a single 8-pin) + cooling ability (replace crappy OEM cooler) it should be telling on what the true ceiling is.

Using just the FE, I don't think it provides enough data to say how good (or bad) custom solutions may be. For reference, the 980Ti reference 'only' OC'd 25%. This is around 18%. Historically, that's pretty solid for the standard OEM card to OC to.

Seems to be a lot of conclusions being made, both positve and negative, on a real lack of evidence so far.

This overclocking did remove thermals. If you look at the review, the highest GPU delta T was under 20C, so the GPU temps would have stayed in the mid 40s. Hopefully better PCBs expose more overclocking headroom, and it's not just a limit of the chip itself.
 

boozzer

Golden Member
Jan 12, 2012
1,549
18
81
damn it, why the hell can't they include a 980 ti lightning in there? would give a much, much clearer comparison.
 

96Firebird

Diamond Member
Nov 8, 2010
5,735
329
126
Compared to the "overclocker's dream", this must be the overclocker's utopia!


Threadcrapping and trolling are not allowed
Markfw900
 
Last edited by a moderator:

Azix

Golden Member
Apr 18, 2014
1,438
67
91
Maybe reserve judgement on OC capabilities until a hard launch? It seems there's a new thread every couple hours in the VC&G section discussing pretty much the same thing is all I'm saying.

Just a lurker going back to lurking now. VC&G still as circular in its discourse as ever.

For the founders edition I think its solid. this was under water. Probably a power limit. They apparently are launching other boards on the 27th so those might do better getting to that 2500Mhz under water claim I keep seeing around.
 

4K_shmoorK

Senior member
Jul 1, 2015
464
43
91
For the founders edition I think its solid. this was under water. Probably a power limit. They apparently are launching other boards on the 27th so those might do better getting to that 2500Mhz under water claim I keep seeing around.

It would be interesting to know if there is a hard limit to OC clocks (even under water) because of the 8-pin. Even more interested to know if there will be 8 + 6-pin custom versions or will enthusiast cards (1080 Ti/Titan) be the only green pascal cards that see 8+6 configurations.

I think I, like most people in this forum, am more interested in what AMD will bring to the table.
 
Last edited:

Face2Face

Diamond Member
Jun 6, 2001
4,100
215
106

maddie

Diamond Member
Jul 18, 2010
5,086
5,413
136
Just makes the wait for custom PCB cards more painful.

Kudos to Gamer Nexus for throwing an EVGA Hybrid on to it. Always good to see out of the box reviews. There was another review site that did an Accelero IV too. Kudos to them too.
Should save you and others quite a lot by stopping most from watercooling the FE model. 2% gains in games for all that trouble.

A solid board with at least 1 6 pin PCIE extra needed for watercooling.

With the power/mm^2 generated however, I guess its uncharted territory. Interested in seeing how it works out.
 

Sweepr

Diamond Member
May 12, 2006
5,148
1,143
136
nobody said the actual OC was bad.

1. question of 2.5Ghz claims

The 2.5 GHz rumour mentions custom models with 2-3x 8-pin connectors, with watercooling.

Still, partners such as ASUS, Colorful, EVGA, Galax, GigaByte, MSI are preparing custom boards with 2-3 8-pin connectors. According to our sources, reaching 2.5 GHz using a liquid cooling setup such as Corsair H115i or EK Waterblocks should not be too much of a hassle.

Thread should be closed, there's no way to predict how custom boards with extra power will do based on single 8-pin Founder's Edition.
 
Last edited:

Grooveriding

Diamond Member
Dec 25, 2008
9,147
1,329
126

Complete trainwreck. The reps are floundering to justify the new premium on the reference card. They can't even stop calling it a reference card themselves. They were ill prepared as the marketers to the press to represent the product and justify the inflated price point D:

Founders is a trainwreck of an idea and a real blunder in what is usually savvy nvidia marketing. This is one of those bad ideas you can only really save by reverting it and saying we dun goofed. Even if there was some credence to the idea that there is an actual demand for reference cards (maybe by system builders, I doubt end-users) after launch, then sell reference at launch without the founders vernacular for MSRP. Once the rush is over, rebrand it as founders and add the $100 premium and let that demand get filled.

The whole premise they put forward there was weak. It's more than likely a cash grab and an attempt to attach a premium to something that has generally been regarded as the bottom of the barrel lowest tier model.
 
Last edited: