wouldnt there be nothing for the backplate to dissipate heat from?

most of the time backplate has thermal pads for the RAM modules IINM

AMD Zen will be another hype again...
But, this time, like usual, don't give high-hope to it.

Hopefully Zen will not be a disappointment - if it's a disappointment then Intel will have absolutely no incentive to improve their chips...

I miss the days where AMD and Intel were actively competing each other for better products - still remember where you can get massive improvements in performance for each leap in generation?

If AMD went down under and we get an Intel/nVidia monopoly things are not going to bode well, especially for the PC gaming community.

This time the problem aren't RAM module, but the VRM.
http://www.overclock.net/t/1561860/various...0#post_24083978

The cooler didn't cool the VRMs?

No idea, but to be fair may be isolated case or the particularly guy didn't reassemble them back well enough.
http://www.hardware.fr/articles/937-6/rade...erclocking.html

Also I reckon such high temperature will only occur on stress test with power virus.
http://www.tomshardware.com/reviews/amd-ra...y-x,4196-8.html

Let's wait and see if there's any further story on that.

This post has been edited by Human10: Jun 25 2015, 01:37 AM

yeah but for Fury X, wouldnt the VRAM module be on the die itself? i mean, non HBM cards have their VRAM module soldered on back of PCB, which is why the backplate helps

but theres no back VRAM module from the pictures ive seen, so backplate isnt necessary

OCN server crashed.

Stick some thermal padding at the high heat zone at the back and in contact with backplate (metal?) will help a little than none I think?

Putting a backplate on it without padding might actually entrap a layer of air in between and worsen the cooling of the VRM?
https://cms-images.idgesg.net/images/articl...592012-orig.jpg

So its either put it with thermal pad or else better remove it.

This post has been edited by Human10: Jun 25 2015, 02:06 AM

well they are metal after all, might be heat conductive

I don't think they dare to directly in contact the metal plate with the PCB, later short circuit with the components.

If not in contact, there gonna be air gaps in between, and we all know how worse air is as a heat conductor. Worse it is not perforated.
http://pclab.pl/zdjecia/artykuly/chaostheo...strony_down.jpg

Better remove it or add non-conductive thermal pad to it.

This post has been edited by Human10: Jun 25 2015, 02:13 AM

If the card is cool i dont see the issue of needing thermal pads or anything being stuck to the backplate..

Card hardly goes above 50 degrees. Pretty sure AMD that has the expertise to design such a complex chip would know whether or not to add thermal pads to the back plate

The chip itself is dead cool, the issue now is the VRM aren't (check my post 24th). Kind of remind me on the 290 + kraken g10 saga last time.

But again, we no idea if that is isolated case, user tampering or only occur under extreme stress test with power virus, so let's wait and see how the story develop.

This post has been edited by Human10: Jun 25 2015, 04:45 AM

Zen is not hyped. It's a brand new CPU architecture. The fanboys is strong in this one, this is AMD Radeon discussion not AMD CPU discussion.


@S4PH On the subject, I already told in previous thread that the performance won't be that good, and I was right. Just increasing the stream processor count and not increasing the front end like geometry engine, ACE it not a good idea. AMD doesn't learn from nvidia at all.

even though the VRMs already cool without the thermal pad, but im pretty sure with thermal pads, the temp will be cooler and better.. its cheap and simple, quite disappointing AMD did not put the thermal pads themselves..

I don't think AMD failed to deliver, they actually brave enough to invest on HBM and get it into market first. That alone is a credit. Just like they did with GDDR5. Only this time their involvement in HBM is more direct.

Despite alot of problem in manufacturing the chip, they manage to at least match the 980Ti performance, and the best part their pricing is still lower than 980Ti.

Maybe our expectation is too high on AMD for this card to beat the Titan X. For sure nvidia Pascal is more mature product when it is release, but AMD will have the upper hand in terms of experience and familiarities with HBM technology.

lol what how come its tottaly different from what amd benchmark themselves



I'm so glad I followed your rant about the card and be skeptic about it . My next card going to be 980 ti for sure
To thought that watercooling would actually benefit fury x in terms of overclocking but mere 5% increase

Now its up to the Fury Pro(non-X) to try and deliver(although I don't think they can catch up to the FuryX's level). But it is interesting to know what the AiB partners can conjure up with.

Although the Fury X seems like a miss and hit, I do think as the driver comes into a more mature state, the performance will stabilise. After all, there are too much variance between review sites to be 100% conclusive.

Currently having very bad experience with Sapphire R9 3xx series.

The screen just flickers away, often crash and unable to do anything. NOTE : That's on the same system that's able to run GTX 980 Ti and R9 295X2.

I got fed up, did a fresh installation on different SSD, change mobo, RAM, processor, PSU, SATA cables whatever you can think off. Still the same thing.

Windows is 8.1 of course, driver is 15.6 as on the site. Nope, didn't work. 15.5, 15.4, 14.12 also doesn't work.

Now I'm just waiting for PowerColor units to come in and see if it's the brand issue or the Radeon chip itself.

And so far I have problem with R9 390X and R9 390 Nitro from Sapphire. Fury? Well, I haven't plug it in.

There's another reviewer in youtube that faced even worse problem than your 390x...his review unit can't display anything at all...