Only a few 1500 benchmarks in here , you would have thought everyone would be at 1500 MHz 
https://forums.overclockers.co.uk/showpost.php?p=28112235&postcount=7
https://forums.overclockers.co.uk/showpost.php?p=28112235&postcount=7
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
** The Official Nvidia GeForce 'Pascal' Thread - for general gossip and discussions **
- Thread starter Boomstick777
- Start date
I've settled for 1470/2000.
1531/2103 was firestrike and heaven 4 stable but not game (Squad Greenlight) stable so backed down a bit. No crashes in Doom 4 so far either
1531/2103 was firestrike and heaven 4 stable but not game (Squad Greenlight) stable so backed down a bit. No crashes in Doom 4 so far either
Associate
Yep i am near enough 1500 on stock bios with a G1 i get 1491 but with vrel pref but its stable across games and bench.
With bios mod Core B 1555 8000mem no caps stable in bench but still in testing phase ..as i have not had this G1 for that long...
it starts flaking out at any thing above 1565 requires more juice so the heat and noise is above what i do not like...
So i will more likely stick with 1520 or so to keep fan volts down
With bios mod Core B 1555 8000mem no caps stable in bench but still in testing phase ..as i have not had this G1 for that long...
it starts flaking out at any thing above 1565 requires more juice so the heat and noise is above what i do not like...
So i will more likely stick with 1520 or so to keep fan volts down
Associate
Yep doom is very well optimised as its Open GL even better when you place the triple buffer on
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 29,994
I think even G has seen the light now tbh and is looking at a non-ref card....
No I think the average speed is about 1400-1450, you will just get everyone at 1500 saying about it, probably 1 in every 50 980ti owners and only the cards like lightning, xtreme gaming etc etc. So comparing that to a reference card is not fair, although the review comparing at 1455/8000 is about right, but I don't think most people would even put their memory at 8000 for 24/7, I know I wouldn't anyway. But yes overall the 1080 look good but will entirely depend on how much they want for them.
Last edited:
That would be a 30% overclock which is pretty unlikely. A 980ti with 30% would be 1560mhz...
Last edited:
Could be an NDA in place so any announcements don't stop the early gouge or perhaps they are holding off until Computex next week.
1450 is the most you would get 24/7 on a reference 980Ti, with 1500 being the exception rather than the rule. But it is clear the reference 1080 is getting around 2.0 Ghz for sustained gaming, with anything higher being the exception rather than the rule.
I see far too many posts trying to claim the 1080 is a great overclocker, with numbers such as 23% being flung around. The reality is that the average boost clocks are around 1750 with OC of ~2GHz (with more aggresive custom fan profiles). This is in line with most reviews showing ~5%-15% performance gains on OC 1080 over stock. Obvioulsy you will get some golden samples or even some outlier games that OC better but from what I can see the 1080 FE is not that great at overclocking.
http://www.guru3d.com/articles-pages/nvidia-geforce-gtx-1080-review,29.html
"The Boost clock will now render at roughly 1950~2050 MHz depending on the power and temperature signature. The GPU will continuously be dynamically altered on voltage and clock frequency to match the power and temperature targets versus the increased core clock. In FireStrike we are now hovering at the 2 GHz marker on the Boost frequency for example, but some games jumped to roughly 2.1 GHz one second and dipped below 2 GHz the other."
What I find puzzling in their OC conclusion is the claim that anything from 5% - 20% extra performance can be expected. Yet their graph shows nowhere near this 20% performance increase. The graph clearly shows no more than ~15%.
In fact even HardOCP OC results show only ~2.06GHz average OC and that is with 100% fan speed.
http://www.hardocp.com/article/2016...ers_edition_overclocking_preview#.V0RWTORwY-V
"the average new clock speed is 2062MHz versus 1770MHz average clock speed at default operation"
16% OC with a 100% fan setting, and that's not even an actual performance difference. ooooh awesome.
Better than Fiji I suppose, but it's clearly not as good as reference 980Ti for overclocking potential. Maybe AIB custom cooled versions will be better with the extra power connectors but FE is NOT good overclocker and that's backed up in pretty much most reviews. Even Hardwarecanucks who had the gall to claim "going beyond insanity" shows only a ~13% actual performance gain with their 2.12GHz OC.
Of course ~13% is free performance and not to be sniffed at but please stop the 1080 is an awesome overclocker BS. I think people are seeing 2.0-2.1GHz and thinking "wow, that's a big number" but failing to realise as an actual percentage over stock it's not that great.
FE 1080 - not a great overclocker
Custom cooled AIB versions with better power deliver = wait and see.
I see far too many posts trying to claim the 1080 is a great overclocker, with numbers such as 23% being flung around. The reality is that the average boost clocks are around 1750 with OC of ~2GHz (with more aggresive custom fan profiles). This is in line with most reviews showing ~5%-15% performance gains on OC 1080 over stock. Obvioulsy you will get some golden samples or even some outlier games that OC better but from what I can see the 1080 FE is not that great at overclocking.
http://www.guru3d.com/articles-pages/nvidia-geforce-gtx-1080-review,29.html
"The Boost clock will now render at roughly 1950~2050 MHz depending on the power and temperature signature. The GPU will continuously be dynamically altered on voltage and clock frequency to match the power and temperature targets versus the increased core clock. In FireStrike we are now hovering at the 2 GHz marker on the Boost frequency for example, but some games jumped to roughly 2.1 GHz one second and dipped below 2 GHz the other."
What I find puzzling in their OC conclusion is the claim that anything from 5% - 20% extra performance can be expected. Yet their graph shows nowhere near this 20% performance increase. The graph clearly shows no more than ~15%.
In fact even HardOCP OC results show only ~2.06GHz average OC and that is with 100% fan speed.
http://www.hardocp.com/article/2016...ers_edition_overclocking_preview#.V0RWTORwY-V
"the average new clock speed is 2062MHz versus 1770MHz average clock speed at default operation"
16% OC with a 100% fan setting, and that's not even an actual performance difference. ooooh awesome.
Better than Fiji I suppose, but it's clearly not as good as reference 980Ti for overclocking potential. Maybe AIB custom cooled versions will be better with the extra power connectors but FE is NOT good overclocker and that's backed up in pretty much most reviews. Even Hardwarecanucks who had the gall to claim "going beyond insanity" shows only a ~13% actual performance gain with their 2.12GHz OC.
Of course ~13% is free performance and not to be sniffed at but please stop the 1080 is an awesome overclocker BS. I think people are seeing 2.0-2.1GHz and thinking "wow, that's a big number" but failing to realise as an actual percentage over stock it's not that great.
FE 1080 - not a great overclocker
Custom cooled AIB versions with better power deliver = wait and see.
Last edited:
Soldato
Videocardz managed to get some news about custom cards.
http://videocardz.com/60269/custom-geforce-gtx-1080-1070-cards-what-we-know
Nice to see EVGA cards will have ACX 3.0 cooler as I predicted but shame no news about MSI with Frozer VI cooler yet.
Seen that before, still not much info. I think nvidia have an NDA on anything non FE until the release date... well it certainly looks like that.
Associate
Simple: Upgrade less often/be smarter with your upgrades. It depends on what exactly you want out of a card, what resolution and what framerate. Hopping from top-end to not-quite-top-end is not as advantageous. From what I've seen, the best way to upgrade is to stick to one particular part of the stack. For best results, one would go with the mid-range x70. But if one has the x80ti, it's best to wait for the next x80ti before upgrading. Patience is a virtue. If one is too impatient, then SLI is an option for much better performance. Even in games with poor SLI scaling, 2 980ti will always outperform a single 1080.
1080 (non-founder's edition) is best upgraded from a 980 or below.
The only exception to this is if one is swimming in money, in which case buy whatever you want, a few hundred quid is probably worth nothing to such a person.
And the funny thing is that many companies like Nvidia have turned impatience into a business. It's like pre-ordering a video game. Buy before you try and enjoy the day 1 issues! Thankfully it's not so bad with GPUs.
The claim that FE 1080 is not a good overclocker (on existing evidence) is not invalidated because you keep claiming "neither is reference 980Ti".
Last edited:
Associate
- Joined
- 16 Mar 2011
- Posts
- 17
I don't think that is standard. I have reference 980ti's in SLI in an FT05 case and with an 180mhz overclock they both reach 1342 boost using between 50 and 55% fan speed, depending on the game.