Carzy Zarx’s PC Gaming Emporium - Catch Up on All the Latest PC Gaming Related News

[Alby_da_Wolf]

 

 

QUAKECore89 said:

Sigh, fine, i'll get this cute card...

http://videocardz.com/61743/gigabyte-announces-geforce-gtx-1070-mini-itx-oc

Still  hate Nvidia though.

Don't take hurried decisions, AMD usually makes significant improvements refining firmware and drivers, and this time they could be quicker than usual, at least for the most urgent issues.

Alby_da_Wolf said:

JEMC said:

Hopefully. Custom cards like the Sapphire Nitro and PowerColor Devil are replacing the 6-pin power connector for an 8-pin one, and the Asus Strix will use two power connectors (either two 6-pin, or one 8 & one 6-pin), so the card will be able to get the extra power it needs from the connectors, not the PCI-e lane.

 

Or, if you don't care about brands and just performance/money, maybe you should wait for the GTX 1060, that looks to be on par with the GTX 980...

http://videocardz.com/61753/nvidia-geforce-gtx-1060-specifications-leaked-faster-than-rx-480

AMD is investigating the issue, but it looks like it just happened in a few cases  amongst hundreds evaluation cards sent to reviewers. They most probably are just a few faulty cards, not unusual in the first production batches. My additional hypothesis is some quirk in the power management parts of firmware and drivers that's triggered just in some configurations HW and SW, this must have been considered by AMD too, as they are trying to reproduce those conditions in their internal tests (from what I understood, not only for out of PCI-E specs power consumption, but also for cases in which reviewers just measured consumption within those specs, but excessively higher than AMD declared ones).

[...]
The source of the rumours are Tom’s Hardware which found that the RX480 they had received for review drew 86W through the PCIE slot. That’s 11W above the maximum 75W specification required to meet compliance.
[...]
AMD’s Senior VP and Chief Architect Raja Koduri was a bit puzzled when asked about the rumour. Mostly because PCIE compliance is one of the basic tests before a card goes out. He said that the RX480 passed its testing, but he was taking the reviews seriously.

“However we have received feedback from some of the reviewers on high current observed on PCIE in some cases. We are looking into these scenarios as we speak and reproduce these scenarios internally. Our engineering team is fully engaged.”

It is possible that the review card was faulty or something went wrong with the testing.
[...]
However there do seem to be a few GPUs exhibiting anomalous behaviour, and we’ve been in touch with these reviewers for a few days to better understand their test configurations to see how this could be possible, that post says.

“We will have more on this topic soon as we investigate, but it’s worth reminding people that only a very small number of hundreds of RX 480 reviews worldwide encountered this issue.”

So far we have only heard review problems for Tom’s Hardware in fact the card has had reasonable reviews elsewhere so it could be a faulty card.
[...]

http://fudzilla.com/news/graphics/41023-amd-investigating-pci-power-draw-issues

 

Pemalite said:

AMD has released a statement about it's Polaris cards using more power than they are meant to.

http://www.anandtech.com/show/10465/amd-releases-statement-on-radeon-rx-480-power-consumption

"As you know, we continuously tune our GPUs in order to maximize their performance within their given power envelopes and the speed of the memory interface, which in this case is an unprecedented 8Gbps for GDDR5. Recently, we identified select scenarios where the tuning of some RX 480 boards was not optimal. Fortunately, we can adjust the GPU's tuning via software in order to resolve this issue. We are already testing a driver that implements a fix, and we will provide an update to the community on our progress on Tuesday (July 5, 2016)."

In other words, they will fix it with a driver update. There will likely be BIOS updates as well for some cards.

It looks like my guesses weren't too far from reality. Nice to know it isn't a HW design flaw, that would be the worst thing, but just a power management SW problem that can be fixed with driver and firmware updates. Nice to know too that they've been very quick to find the problem and they'll fix it very soon, if all goes well and crossing fingers.