• 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 Fury(X) Fiji Owners Thread

So increasing voltage is just basically increasing power to nuclear levels and cooking the VRM's for an insignificant performance improvement, you're best off just overclocking it on stock volts and leaving it at that.

Nevermind, there's still DirectX 12 to cling to. ;)
 
His clockspeed v performance figures looks like he has got something wrong unless the card is throttling due to heat.

It will be interesting when I try it with custom waterblocks on mine.

I think i might have spotted a problem.

We will be using Battlefield 3 at 4K resolution to verify performance gains. All testing was done at stock memory clock of 500 MHz, stock fan speeds, and power limit set to +50% to avoid throttling at higher voltages.

Using the conservative stock fan speeds would mean the fan would be running slowly, the result of this would mean the vrms would overheat, as they run warmer that the core. I'm not surprised this was the case with +150mv. I will be testing it myself and posting the results on this forum when Trixx is released and I'll be using 100% fan speed to keep the vrms cool when adding voltage. :)
 
His clockspeed v performance figures looks like he has got something wrong unless the card is throttling due to heat.

It will be interesting when I try it with custom waterblocks on mine.

VRMs are always the culprit of throttling on GPUs. I'm wondering if it would be worth fitting a heatsink onto the VRM pipe with epoxy and thermal compound... Maybe that would help a bit.
 
I think i might have spotted a problem.

Using the conservative stock fan speeds would mean the fan would be running slowly, the result of this would mean the vrms would overheat, as they run warmer that the core. I'm not surprised this was the case with +150mv. I will be testing it myself and posting the results on this forum when Trixx is released and I'll be using 100% fan speed to keep the vrms cool when adding voltage. :)

In all these tests, GPU temperature barely moves thanks to the watercooling block. Going from 67°C at stock voltage to 71°C at +144 mV isn't worth mentioning.

Not so sure there Matt, you might be right, but just how much difference having the fan spinning faster will make to the VRM's temps, being that they are not directly cooled by the fan. It might make a difference, but probably not much.

It will be interesting to see others results.
 
Relying on overclocking to be the fury's saving grace is a bit like ****ing into the wind. It does surprise me that Joe Macri seems so far to be wrong about it's ocing potential though.

Its still a good card regardless, just didn't live up to the hype, 99% of which was created on internet forums long before anything official was ever mentioned.
 
Not so sure there Matt, you might be right, but just how much difference having the fan spinning faster will make to the VRM's temps, being that they are not directly cooled by the fan. It might make a difference, but probably not much.

It will be interesting to see others results.

The copper tube the water passes through runs over the vrms before it goes to the radiator, so it's pretty important.

Here's a fun fact for you. I can make the vrms throttle at stock clocks if i slow the fan speed down to a halt before the core even gets remotely near 75c. How do i know this? The core and memory clocks jump into 2d mode 300/150. Does that sound familiar from what you just read in the article?
 
VRM throttles at high temps and hes using stock fan profile lmfao, what a pointless test.
Not a huge performance increase for sure, but it wasnt pushed as high as it could be. I think I would OC to 1200 on lowest possible volts personally, and ramp that bloody fan up. AMD's stock profile is far too slow for my liking, I doubt the fan is even audible up to 30% anyway so why have it lower than that...

The power usage goes up by about the same amount as other GPU's when overvolted, but the performance increase is not all that big.

I think if you have 2 fury X at 4k then you have more than enough power to max everything, you might as well go for a slightly lower overclock to use 100w less per card and only lose 5 FPS and 5c imo :)

It's the reason I run my 290's at 1050 :)

Interesting that overclocking memory has (relatively) a much larger performance increase than on GDDR5. That being said, maybe it just seems a lot because the overclocks to the core don't do all that much. What happened to the days of 7950 where a 20% core increase was 15% extra performance. I get the feeling the huge amount of cores on the Fury X are being bottlenecked by some other components on the board, maybe the ROPs... no increase but 50% increase in cores over the 290X, would explain why overclocking doesn't do much :/ I am by no means an electronic engineer though so I could just be talking out my ass.
 
Last edited:
Not so sure there Matt, you might be right, but just how much difference having the fan spinning faster will make to the VRM's temps, being that they are not directly cooled by the fan. It might make a difference, but probably not much.

It will be interesting to see others results.

The coolant in the radiator that passes over the VRMs is cooled by the fan, though.

I'd be interested to see if you could mod the X with a double rad and res :D
 
Relying on overclocking to be the fury's saving grace is a bit like ****ing into the wind. It does surprise me that Joe Macri seems so far to be wrong about it's ocing potential though.

Its still a good card regardless, just didn't live up to the hype, 99% of which was created on internet forums long before anything official was ever mentioned.

That's been the problem from the start, All the Titan killer claims etc. To me it looks like it will do me proud once I can get one, I only want to run 1440p (currently 1080 vsr'd to 1440) but if I go for the 3440 x1440 monitor I'm eyeing up I reckon this is a good choice, I prefer to run my hardware at stock if possible and I haven't even got around to playing with the volts on my 290x yet so it won't effect my choice much but I can see how it will for a lot of people, especially the sorts that frequent sites like this.
 
AMD really seem to have wasted their efforts on HBM. I have yet to see how the extra bandwidth benefits games even at 4K since even the 512bit 300 series seems to be close at 4K and the Maxwell cards aren't lagging much either..

AMD needs to concentrate on the gpu core performance which is severely lacking considering the huge amounts of stream processors they have added with little or no benefit at all. With over 80% more SP than a 290X you would expect a decent jump but somethings gone wrong somewhere.:confused:
 
Last edited:
A FuryX can only get 53fps in BF3?
Then things are even worse than I thought lol

At 4K max settings.
From that DC3 review, this is hardline at maximum settings.

4rmFvW8.png

Glad I stuck with my 290's, £500 to upgrade 2 290's to furies for 10 more fps at 4k when I already (almost) max everything at 60 fps? no thanks :D
I get 60 fps locked in BF4 with FRTC at maximum settings, bar AA which is 2x deferred and post off. Can't tell much difference at all putting AA but I take the FPS hit, not worth it for me.

I think I will stick with this till die shrink.
 
Last edited:
Back
Top Bottom