works like a charm bro, now the temps are a lot better. never knew such solution 

$650 for R9 Nano. That's close to RM3k

Ah wells.

http://wccftech.com/amd-radeon-r9-nano-unl...faster-gtx-980/

Yeah so the previous article about Nano having a premium price since it's the only awesome mini-ITX card, is right after all

I guess most people doesn't seem to understand the Nano - it's intended for the mini-ITX/microATX crowd, where space is always a premium.  I have a microATX gaming machine, and while it does fit a 980Ti (theoretically) , but it will be a very tight squeeze. 
My 7950 does fit into the case, but it is a huge pain whenever I need to take it out for cleaning - I have to remove the HDD cage and the CPU fan from the case before removing the card, which is tedious. 

It might not appeal to the HTPC crowd, but for those with tiny cases, this card is a godsend to them. 

Don't write off a product just because it does not appeal to you.  This is obviously a product that appeals to a very small crowd, and I happen to belong to that crowd. 

Of course, the low yields of HBM may contribute to the high pricing - the pricing is likely to fall when there is better yields. 

Addendum: If the Nano shares similar specs to the Fury X - then technically we can flash the VBIOS of the Fury X with a VBIOS from a Nano for greater power efficiency? 

Also, this is a very niche product, highlighted by some reddit users:
https://www.reddit.com/r/AdvancedMicroDevic...full_fiji_chip/

A niche product is not going to be cheap... 

390 DARAB 2.

the spacing is quite important, else ppl will think its 390X2 = 390X CF.
since FuryX / Nano is out, will we see dual GPU = FuryX X2?
iinm, i remember AMD slide got shown dual gpu monster

Higher leakage is usually better at overclocking because they tend to be more stable with extra volt and with WC keeping the leakage, it will overclock much better than lower leakage chip.
The Fury-X still lack proper voltage control and the VRM temp is somewhat worrying with extra voltage.
It's a product that cater to a niche audience with specific form factor and use-case scenario. How many people own 970 mini anyway?
Overall the Nano impressed me apart from the steep price. I was expecting 390-ish performance with thermal throttle everywhere.

Well, that's comparison which concluded the Nano is more power efficient than 290X (typical board power: 175W vs. 250 = 30% lower on Nano) but better performance than 290X (37.966 fps vs. 27.207 fps in FC4). I'm sticking with my guns because performance per watts = power efficiency is not accurate IMO.

Anyway, ignoring that, I get your drift (your original previous post). However, since Nano is lower clocks & lower power consumption card of Fury X, then the performance per watts ratio is equal between Nano & Fury X, don't you think?

Why would it not be Power Efficiency?

It's a matter of # of power used to deliver X performance.

No, we're talking at stock.

No need to change PSU, your FSP Hexa 550W is sufficient.

..........You guys do realize 390x has two pcie power connectors. Not to mention the PSU has two +12 rails as well. It is sufficient providing he doesnt overclock.

First PCIE source from PCIE power connector.
Second PCIE source from molex/sata to PCIE power connector.

Well, both from same rail, +12V1. The +12V2 is mainly for CPU.

On this unit, since the PCI-E connector has the 12V2 rail all to itself...

The two +12 V rails are distributed as follows:

    +12V1: All cables but the EPS12V/ATX12V
    +12V2: The EPS12V/ATX12V cable

This is the typical distribution used by power supplies with two +12 V virtual rails.