Still, I wanted to get at least 2000 out of it, for no real practical reason. 40mhz probably won't generate even 1 fps once you're this far above normal boost 3.0 clocks. Damn the msi boosted to 1987 out of the box but would do this weird thing where the power limit would drop to 70% and the framerate would drop to the teens. That was without overclocking either. Hindsight I guess. I might try one more side by side with this one and keep the better one.

Category 5
Shoot. My EVGA founders is doing that! I have to reboot to fix it when it happens. Gpu usage won't even go over like 50 and I get completely jerky motion.

Does it do this normally? Or just when extremely overclocked? The msi gaming x did it on stock clocks, and much sooner if I increased the power limit. I thought it was my psu so I went out and got an EVGA G3 1000w...still did it, I like the PSU so I might just keep it.

Valtrius Malleus
If you want the same clock speed with higher voltages you need to do something which seems counter intuitive. You need to lower the offset! but increase the voltage percent. That will allow the card to push more voltage through which will then check the voltage curve and see what frequencies it would be running at with those voltages. If you lower the offset then it will be running at the same speed, but a higher voltage, which might make it more stable.

I'll have to play with this. I did some of this on my 1080 SC before, and was always frustrated with the gimmicky toy-like interface. I'd have been very happy if they just opened a table like a spreadsheet and let me type in the values I want at the various speeds.

Then again, I'm not 100% sure I'll really spend much more time on this. Most of the games I currently play would be pegged at 60fps even with no overclock applied, and those that don't are close enough that the difference between a stable 2076 MHz and a stable 2088 MHz or even 2100 Mhz wouldn't be noticeable. If I do get down in the weeds with the voltage curve it will be only to satisfy the impulses of my e-peen.

I'm happy with this card. My 1080 SC that I upgraded from was very good too. I guess the silicon lottery has done me well so far. Neither has been world class, but they seem to be in like the 80th or 90th percentile just based on comments about what others are getting. That's a guess, but it's definitely good enough for me.

With both cards, the addition of the EKWB helped me get at least another 50-100 Mhz on the top end, and then hold it, while with the air cooler both cards would hit some maximum and then throttle down somewhat as the temperature thresholds were crossed. I'm very satisfied with the EKWBs, and they look great too.

mdcurry
Here is my timespy score with my Titan pascal just for reference. I had a Seahawk 1080 Ti that scored better by a lot but did not run TimeSpy before I returned it. It did not run as cool as I thought it should.
 
http://www.3dmark.com/spy/1608979
 
My graphics score seems higher than the !080 Ti by a little bit.

I would have thought that the Overall Score would be higher then a single GTX 1080 Ti.
http://www.3dmark.com/compare/spy/1608979/spy/1631680
Maybe my AMD Ryzen 7 1800X is better than what we think. Overall Score that is.
Your Intel Core i7-5775C Processer is that what is holding you back not your Graphics Card(s).
But then at the Cost One Titan Pascal Graphics Card you can get two GTX 1080 Ti Graphics Cards and you have no chance.
http://www.3dmark.com/compare/spy/1608979/spy/1631680/spy/1420347 Even on a Slower CPU

mdcurry
Here is my timespy score with my Titan pascal just for reference. I had a Seahawk 1080 Ti that scored better by a lot but did not run TimeSpy before I returned it. It did not run as cool as I thought it should.
 
http://www.3dmark.com/spy/1608979
 
My graphics score seems higher than the !080 Ti by a little bit.

How hot did it run? I'm thinking of converting my founders to a hybrid despite the mediocre OC it doesn't have any coil whine unlike my previous two.

mdcurry
Here is a Orange Room Score with Seahawk 1080 ti
 
http://www.3dmark.com/vrm/19514965

Outstanding! The system comfortably beat the target frame rate for this test. Now try running the more demanding Blue Room benchmark.
Very Few "VRMark Orange Room Benchmark Desktop 1.0" Scores posted to date.
Any Blue Room Benchmark to Post?

bcavnaugh
 
I would have thought that the Overall Score would be higher then a single GTX 1080 Ti.
http://www.3dmark.com/compare/spy/1608979/spy/1631680
Maybe my AMD Ryzen 7 1800X is better than what we think. Overall Score that is.
Your Intel Core i7-5775C Processer is that what is holding you back not your Graphics Card(s).
But then at the Cost One Titan Pascal Graphics Card you can get two GTX 1080 Ti Graphics Cards and you have no chance.
http://www.3dmark.com/compare/spy/1608979/spy/1631680/spy/1420347 Even on a Slower CPU

 
Bcavnaugh, that is an interesting result. I'm really glad that AMD came out with that cpu, since it really puts some pressure on Intel to beat it hard with a future generation. My head wishes that the Ryzen actually beat out the i7 6900, but my e-peen is glad that it just falls short of doing that. 

Don't worry, any tiny twinge of smugness I'm tempted to feel is quashed by the recognition that you've got a 2x1080ti result too.

I am curious though. I see with your Ryzen you've just cracked 4GHz. Is that a max OC for that cpu? I'm curious. My daily use OC on my 6900 is 4.2Ghz, though I've had it up to 4.4GHz before for testing purposes (and it ran nice and stable, I just didn't like the fact that I had to crank more voltage into it, so I lowered it for normal use).

It'll be interesting to see if AMD is able to keep it up, and what Intel's reply is to the advent of 8-core AMD processors with equivalent performance at half the cost. I'm very glad I only paid about what you did for the Ryzen for my 6900 (from a friend). I don't see how Intel can keep charging twice as much for performance that's only modestly better.

Here's an Orange Room score:

Orange Room 1.0:   11,826  http://www.3dmark.com/vrpor/106555
 
I'll look at what I have to do to run the Blue Room and then post that too.

I couldn't find any 1080ti results to compare by. Did I really just upload the first 1080ti Blue Room results?

I don't feel so bad now. I crushed a whole bunch of 1080 SLI results (all but four, and the top 1080 SLI score was only 6 points higher). 
 

Category 5

With my card still crashing with no overclock, I am wondering id it's just the card and not the overclock that's causing the instability.  Will run everything non-OC for a few days and see what kind of experience I have.  If I'm still getting crashes then I guess this card goes back and I play the lottery again.

Man, that's terrible. It sounds like a bad card. It shouldn't crash under stock settings. Not being able to clock high is one thing, but instability at the factory settings is a deal breaker. I would (and did) exchange it if I had another dud like the last gaming x was. 
 

TRClark911

sethleigh
That's kind of scary that they say an overclocked Core i7 6900 with an overclocked 1080ti can't be expected to perform in a complex VR situation. ****

Sounds as if this 3DMark VR test is a crock... but I just remembered I bought it when I bought the other tests... I've just never ran it.  What I can tell you is the 7700k and 1080 Ti knock VR out of the park (so did the 1080 before I stepped up) so I'm not exactly sure what the frame rate requirement is on the blue room test but I'll run them when I get home.
 
You're more than good for VR.  I VR just fine with max settings on the Rift.

Yeah, they've got the Orange Room test and the Blue Room test. The Orange Room tests machines like this blow away, getting massively more than the 109fps that they set up as the target framerate. Then there's the Blue Room test, which bring these machines to their knees. I guess they're laying out a standard for future content that's just incredibly rich, whether there's any current content like it or not. 

I don't know how it's implemented, but I get the feeling the Blue Room test isn't that well optimized for SLI, since my single 1080ti on my 8-core cpu dominated literally all but 4 of the 1080 SLI results, and the ones that beat mine were so close (within 6 points out of over 3400) that they're within the margin of error. After posting that I couldn't find any 1080ti results, I did find some 1080ti SLI results.

sethleigh
Here's an Orange Room score:

Orange Room 1.0:   11,826  http://www.3dmark.com/vrpor/106555
 
Interesting. My machine didn't achieve their standard framerate in the Blue Room test.

Blue Room:  3410  http://www.3dmark.com/vrm/19653703?
 
That's kind of scary that they say an overclocked Core i7 6900 with an overclocked 1080ti can't be expected to perform in a complex VR situation. ****?

Just got home and here it is...
 
Orange: 12,021 http://www.3dmark.com/vrpor/106746
Blue: 3,410  http://www.3dmark.com/vrpbr/19098
 
7700k vs 6900k?  I think it's safe to say our machines are pretty much on par with each other and for the record my PC has no problem with anything currently available in VR.  I'm not sure exactly what this blue room is supposed to represent... VR in 5 years maybe? 
 
 

sethleigh
I don't know how it's implemented, but I get the feeling the Blue Room test isn't that well optimized for SLI, since my single 1080ti on my 8-core cpu dominated literally all but 4 of the 1080 SLI results, and the ones that beat mine were so close (within 6 points out of over 3400) that they're within the margin of error. After posting that I couldn't find any 1080ti results, I did find some 1080ti SLI results.

This blue room test is way beyond anything currently available in VR.  I wouldn't lose sleep over whatever this score is supposed to represent. 
 

sethleigh
Having browsed through more results I'd go so far as to say that no gaming machines anywhere reach that target. Not to say that a machine couldn't be hand-crafted just to do so, but one's wallet would have to be both heavy and open, and even then I'm not 100% it could be done with today's cards. I think this test is there to give us a challenge going into the generations of cards that will come out in the 2-4 year horizon.

Like I said... maybe the blue test represents VR in 5 years... and we tied!  LOL

sethleigh
 
Yeah, I haven't followed the Intel processors that came out after I overhauled my machine with the 6900, but I like the fact that what it lacks in single-threaded performance vs. cpus like yours, it makes up for in massive amounts of threading. Even if most games don't take much advantage of it, it still benefits me when I can leave my browsers, Skype, SolidWorks, email program, chat programs, VirtualBox running a Linux or Solaris VM, etc. all run while I game, and not notice the difference. Also the 6900 has a massive cache, which helps.

I'm willing to bet that your Orange room score comes down to the cpu clockspeed, because our framerates are so high in that test that it's like cranking resolution down to 1080p in a game, where at 4K it's gpu-bound, but at 1080p it's generating so many frames it becomes cpu-bound instead. The Blue Room seems so gpu-bound it's not even funny, which is why statistically your score and mine are identical, despite your cpu clockspeed advantage.

Btw, your talking VR stuff with me has me interested in checking it out, and I'll have to find a chance to drop in a Best Buy or wherever they have a demo setup and try it out.

It's worth checking out... when done right VR is an amazing experience.  It's not something you dive into with minimum system requirements.  The coolest thing is it's only gonna get better.
 
As for the CPUs... I went with the 7700k because everything I read said it is currently the best gaming CPU.  I also do other things such as CAD and video encoding and it handles those tasks nicely as well.  I didn't see the need to invest in a 6 or 8 core CPU for that.
 
Generally I score slightly behind said 6 and 8 core CPUs... and I'm still very happy with the 7700k.
 

Sli 1080ti on air stable 2025 also have K boost on