AMD always leave little headroom for overclocking, I can’t see Navi being any different.
-
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 AMD Navi Thread **
- Thread starter subbytna
- Start date
AMD always leave little headroom for overclocking, I can’t see Navi being any different.
Soldato
They needed to do that though as they are playing catch up. It isn't really by choice, as they -upto now- have been put of the efficiency sweet spot of their power envelope in order for their chips to compete with Nvidias. As you say I expect them to be pushing the boundaries again, but I am hoping since Vega II showed some good overhead that Navi will leave a little in the tank if prepared to mess around more.
Soldato
Just as you think AMD finally slipped one in there, nvidia release the 2060 2070 and 2080 super, at 299 399 499 with faster gddr6 ram.... damn it green guys give us a break for once you ball touchers
Soldato
- Joined
- 1 Jun 2014
- Posts
- 5,066
AsRock had some concept GPUs on show at Computex. Highly likely they're designs for Navi (skip to 4:48).
Edit: Sorry - I see they've already been talked about.
Edit: Sorry - I see they've already been talked about.
Soldato
And as Navi hasn't been released yet, and the prices were a Sapphire guy talking, AMD have time to adjust their price points to remain competitive. Don't end the world until AMD officially do it on launch.
Tbf assumption is almost always bad... I am hoping to see a 5800 in a weeks time that delivers north of 2070 performance... But only based on scaling up polaris performance. However the last few GPU releases have been luke warm so not holding my breath.
Soldato
Okay while we are being hopeful, i wish for a 2080ti performance card suprise !!
Called the rx5900 or rx 5790 or even the radeon 7 XT with matured 7nm die and higher clocks and all shader cores unlocked on the new rdna arc. BOOM!!
Soldato
HAHAHAHA
Nvidia has the shortest headroom of them all since Pascal, where barely can OC them for 70Mhz while all Pascal & Turing start throttling from 32C. (many current Nvidia cards cannot even push more than 60Mhz OC)
Making completely useless any attempt for watercooling to gain better clocks. At least with AMD, especially the Vega 64, going under water or with good cooler (Nitro) and downvolting, gives you plenty of headroom for overclocking of around 20% until you hit 80C. And thats 300Mhz, 5 times more OC than the Nvidia cards allow which doesn't fluctuate until you hit 80C, not 32C
FYI even the RX580 allows for almost 300Mhz overlock (20%), the only card that doesn't OC that well by only 100Mhz is the 590.
Compare this to ~3% can do (if that given the heavy throttling) from Pascal & Turing.
How AMD "has little headroom" for overlocking at 20% when Nvidia has 3%?
Soldato
Radeon 7 stock 1000mhz HBM 1800mhz coreHAHAHAHA
Nvidia has the shortest headroom of them all since Pascal, where barely can OC them for 70Mhz while all Pascal & Turing start throttling from 32C. (many current Nvidia cards cannot even push more than 60Mhz OC)
Making completely useless any attempt for watercooling to gain better clocks. At least with AMD, especially the Vega 64, going under water or with good cooler (Nitro) and downvolting, gives you plenty of headroom for overclocking of around 20% until you hit 80C. And thats 300Mhz, 5 times more OC than the Nvidia cards allow which doesn't fluctuate until you hit 80C, not 32C
FYI even the RX580 allows for almost 300Mhz overlock (20%), the only card that doesn't OC that well only 100Mhz is the 590.
Compare this to ~3% can do (if that given the heavy throttling) from Pascal & Turing.
How AMD "has little headroom" for overlocking at 20% when Nvidia has 3%?
Me - Radeon 7 1275mhz hbm max 2235mzh core max - i believe thats a 24% overclock no ?
@Chrisc
Last edited:
Soldato
You have blocked Chrisc? Because my reply was for his post.
Caporegime
How much performance uplift from that OC?
Soldato
I was backing you up bro !
Soldato
How do you want that quantified? In fps ? Which game? In benchmark points? Which one?
I'm not being sarcastic, let me know and I can tell ya..
Caporegime
Was just thinking on average when playing games. Yes percentage increase in fps
Soldato
Ahh
sorry.Yeah, AMD allows 20%+ OC (except on the 590), Nvidia allows 3% at best (if you can avoid the 32C+ throttling) yet you see posts "AMD doesn't have OC headroom".
I am here for many years yet never before the hypocrisy and ignorance hit that levels.
And we are supposed to be an Overclocking forum, not a stock clocking forum.
We did overclocking for almost 20 years, to get more performance out of a cheaper parts. Not buy the most expensive parts, and start bragging about FPS following a brand like religious fanatics.
Last edited:
Soldato
Depends on the game, and the resolution but its about 10% ish some games its more, like resident evil 2 remake goes from 50fps at 4k to 60+fpsWas just thinking on average when playing games. Yes percentage increase in fps
Soldato
Idk about Radeon 7 but Vega 64, I can see the FPS performance on all games across the board. Even when the card runs at stock 1632 clocks over the "air cooled" 1453 (thermal throttled)
Last edited:
Soldato
What was that pci-ex 4.0 3dmark test they ran at computex for navi, why was that important?
@Panos I love how when you prove a point wrong, suddenly the person dissapears as if the evidence never was given, maybe for them to return later as if none of it ever happened.... lame
@Panos I love how when you prove a point wrong, suddenly the person dissapears as if the evidence never was given, maybe for them to return later as if none of it ever happened.... lame
Funnily enough he does the same
Caporegime
The only thing I can think of is that that test was there to display the clear performance difference between pcie4 and 3. However I am also a bit confused because GPUs don't max out pcie3, let alone 4...So whilst it is impressive and perhaps a glimpse of what they could do, it seems pointless.
Unless I am missing something entirely?
- I don't know enough about it really.