Fury X voltage adjustment now available

[3DVagabond]

EDIT #2: If W1zard works for Sapphire, I'd assume he were bias towards AMD?

If you read TPU you wouldn't think that. There is nothing I see that is biased toward AMD on that site.

 

[dark zero]

I didn't expecting something solid besides the Fury and the 980 Ti. Both archs reached their limits. The next generation will be from new on everything.

 

[Sabrewings]

It is an OC monster. Look at how much power it eats up with that OC.

Nightmares are dreams, right?

 

[antihelten]

EDIT: Derp, you said it right there long as VRAM is boosted too. Ignore the rest


I assumed you included the Memory OC in there? Or are we just talking about core? Because just core, at 1160 it's closer to 53.5 if you ask me.

53.5/50.9 == ~5% gain.

Seems that other half the final ~9% comes from the memory OC.

But man, at those volts for that perf gain, woof.

Those volts? Achieving the 1160 MHz shown in that graph can actually be done whilst undervolting by 24 mV, which lowers power usage by roughly 20W.

 

[tajoh111]

If you read TPU you wouldn't think that. There is nothing I see that is biased toward AMD on that site.

That's good isn't it. Wizzard is pretty neutral when it comes to reviews.

 

[3DVagabond]

That's good isn't it. Wizzard is pretty neutral when it comes to reviews.

I wasn't critiquing W1zzard, just the comment that since he makes Trixx for Sapphire means he's biased towards AMD.

 

[tential]

It is an OC monster. Look at how much power it eats up with that OC.

Maybe this is why amd locked voltage at launch?

Max oc results would look pretty bad with power consumption with voltage unlocked for launch reviews vs now. Is this amd playing the game?

 

[poofyhairguy]

The importance of vcore mod is the undervolt ability, which seems to behave exactly the same as other GCN GPUs as well, with even a possibility to run higher clocks at less vcore to use less power for a slight gain in performance. The efficiency gains are excellent, -50W, +5% performance. If we're assuming its 250W gaming load, it's now 200W, 5% faster, it just got ~30% more perf/w.

Combine that with the Crossfire benefit and AMD might have a winner at the high end for 4K gamers.

It will be interesting to see what happens when people overvolt Fury non-Xs. Can an air cooled rig even handle that much additional wattage/heat?

 

[soccerballtux]

Why would you expect 30% OC?

You didn't pay attention to every other GCN GPUs out there? They peak at 1.25ghz. They don't peak at 1.4ghz. Thats completely unrealistic expectations.

I posted about this on launch, I said with vcore, it should get 1.2 to 1.25ghz and that's what happened. It's GCN, its from TSMC, why would this one be much better than Tahiti or Hawaii?

that's kinda the number we came up with when we were determining the meaning of the word 'dream'

 

[soccerballtux]

Maybe this is why amd locked voltage at launch?

Max oc results would look pretty bad with power consumption with voltage unlocked for launch reviews vs now. Is this amd playing the game?

I'd say so. At least they're finally being the least bit smart....

 

[soccerballtux]

Combine that with the Crossfire benefit and AMD might have a winner at the high end for 4K gamers.

It will be interesting to see what happens when people overvolt Fury non-Xs. Can an air cooled rig even handle that much additional wattage/heat?

I'd rather just have a 390x. In my mind, Fury X was good if and only if it could overclock 30%. Otherwise, it makes more sense to get the 980TI. Fury and Fury Nano are OK, but Fury X is out.

 

[tviceman]

Either way, it's not a massive overclocker like Maxwell 2 BUT it can gain massive perf/w, so it can beat the 980Ti/Titan X on that metric (been awhile, can NV GPUs undervolt?). Funny ain't it?! lol

I think GM200 was a bit of a failure in perf/w with the standard set by GM204. Kepler showed no degradation with GK110 in perf/w vs. GK104, but GM200 can't keep up with GM204 in perf/w. Fury X will never beat GM204 in perf/w. GTX 980 is 50% more efficient at 1080p, 37% more efficient at 1440p, and 26% more efficient at 4k. http://www.techpowerup.com/reviews/AMD/R9_Fury_X/32.html Even still, 980 TI is ~20% more efficient at 4k, and ~25% more efficient at 1440p. Undervolting Fury X will at best tie 980 TI in efficiency at 4k. Yawn.

It is possible that Nano will compete heavily with GTX 980, but Nvidia is smartly using all of their cherry-picked ultra efficient GM204 dies for notebooks.

 

[3DVagabond]

Maybe this is why amd locked voltage at launch?

Max oc results would look pretty bad with power consumption with voltage unlocked for launch reviews vs now. Is this amd playing the game?

What voltage lock? lol Quit making stuff up just to fit an agenda.

 

[tential]

What's my agenda?

 

[3DVagabond]

What's my agenda?

Justifying the bold part that you are simply making up as a motive for something that doesn't exist.

Maybe this is why amd locked voltage at launch?

Max oc results would look pretty bad with power consumption with voltage unlocked for launch reviews vs now. Is this amd playing the game?

Truth is the voltage isn't and never was locked. There just wasn't software available yet. Which no IHV provides. No conspiracy, no ulterior motives.

 

[Silverforce11]

And the same thing happened with all the other GCN parts, OC with +vcore caused major power usage spikes. It's not a new phenomenon. Even Kepler drew insane power with +vcore, I recall seeing some bench-offs from Balla, Face2Face etc where the 780 was pulling over 400W.

Maxwell is very nice for OC enthusiasts since it can reach >1.4ghz without modified +vcore bioses. Without major +vcore, power use doesn't jump up much. But that's the reason NV has locked out manual +vcore, because power usage goes up exponentially and could cause unwanted damage/RMAs.

The only interesting result here is that Fury X can OC + undervolt at the same time to improve its perf/w a lot. This bodes well for Nano and I don't doubt AMD's claims of faster than R290X with 2x perf/w due to my exp with GCN undervolting.

 

[Arachnotronic]

Truth is the voltage isn't and never was locked. There just wasn't software available yet. Which no IHV provides. No conspiracy, no ulterior motives.

You're arguing semantics here and know full well what tential meant.

 

[3DVagabond]

You're arguing semantics here and know full well what tential meant.

Are you his lawyer? :\

He made up something that didn't exist and then used it to put forth a false presumption. AMD locked voltage to conceal power usage when O/V'd. It's completely fabricated.

 

[wege12]

Since the VRMs are hitting temps at and above 95c and causing throttling if not limited, I had an idea. Would it be worthwhile, for over clocking purposes, to open the Fury X up and have a separate fan cooling the Fury X hopefully helping vrm temps and increasing overclocking potential?

 

[geoxile]

Something wrong with the memory. Might be HBM, might be drivers. Hopefully it's not HBM itself.

 

[3DVagabond]

Since the VRMs are hitting temps at and above 95c and causing throttling if not limited, I had an idea. Would it be worthwhile, for over clocking purposes, to open the Fury X up and have a separate fan cooling the Fury X hopefully helping vrm temps and increasing overclocking potential?

95°C for VRM's is nothing. VRM's can, and do, run a lot hotter than that with no issues.

 

[XiandreX]

Very strange how much mileage they are getting out of the supposedly overkill bandwidth of HBM. Id be interested to see if it really is latency related.

AMD should not have played up "overclockers dream." It overclocks to the same spot as Hawaii which needs to be said... a little over 1200. Which is respectable. For some reason though it seems to scale much worse than Hawaii did.

Either this thing really is limited in DX11 or it was too unbalanced of a design for their first max sized die.

That right there is the reason people are irritated. The amount of power required when overclocked is unacceptable especially given the small boost from core clock.

 

[tential]

You're arguing semantics here and know full well what tential meant.

It's the Internet man. He knows what i mean but I have an "agenda" which still hasn't been told to me lol...

 

[96Firebird]

What's next on the Wait and See™ list? DX12, right?

 

[monstercameron]

That right there is the reason people are irritated. The amount of power required when overclocked is unacceptable especially given the small boost from core clock.

yea voltage increase does that besides this is gcn and 1200-1300 is the highest the uarch goes.