• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

AMD confirms Power Issue for RX 480, blames memory bus, will patch

Cause new PCB with new VRMs costs money.

If they already ran a hundred thousand boards, they're not getting tossed.

And I'm sure management set the clock speeds, since they had to match the 970/ 390.
This thing already has a power delivery system that could run a Fury X.
 
Daniel_Chang said:
At $199, the RX 480 is a better value than the formerly $329 R9 390. Yes, the 390 is discounted now as inventory clears. Anyone looking at a 480 should seriously consider a 390 while they are still available. I know that I am.

I've thought of that but it makes less sense in the long run I believe, 390 only seems to approach 480 prices used and for an extra 5 frames performance benefit you gain twice power usage and there is the issue of support/advances going forward
 
we should probably consider this a $240 part until we actually see the 4gb in stock somewhere.
 
It doesn't, though. The 4 GB model is the 480's saving grace since it brings the price down enough to be relevant.
Otherwise the 8GB model is basically a 1:1 successor of the 390. In fact it seems the 390 might actually have better perf:dollar when you look at aib models which will probably be around $270.

A refreshed 3-year old $300 graphics card has better value than its brand new, one-tier-lower (x80) successor. That's insane.

I do not know about that.
The custom AIB Sapphire Nitro OC 8GB is £20 more expensive than 8GB reference model.
Looks nice as well, especially backplate - also has official TDP of 170.
SAPPHIRE RX 480 NITRO NOW AVAILABLE FOR PRE-ORDER AT OVERCLOCKERS UK (WORLD FIRST) !!! - Overclockers UK Forums


GX37BSP_143822_285x255.jpg



GX37BSP_143820_285x255.jpg



Putting aside powergate, pretty good value at £20 more, clocks still to be confirmed but if they also resolve the power issue this looks to be a good card.
Cheers
 
The custom models will surpass the 390, especially when safe overclocking is possible. That said it is still lower end and there's only so much to expect from a cheap card of this class. Also, the 480 is beats the 390 for its new features by a mile, 4k hevc encoder, hdmi2.0b, many dp ports with 1.4 optional, etc etc. the Hawaii serious shit for htpc duty with 4k displays due to its lack of a proper hdmi 2.0 port.


I saw someone post their 480 trifire runs... then I compared to my old runs.

Result

It's insanity to me that people are cross firing low/mid end cards never mind trifire. Shit, crossfire two Fury X's over doing that.

Trifire / triSLI works well in few games. I looked into it extensively when the 970 came out. At least you have a fighting chance with two card multiGPU.

BTW awesome score with those 290x's. I had to double check the 1080s which are around 5-6k for one GPU and you got 9k. Not shabby at all. Those cards had good value and keep chugging along...
 
Last edited:
Sapphire guy did say in the video that they really are not in to slapping another 30 bucks on their card just because its after market. So you won't pay too much premium over the reference cards. So that price is not surprising.
 
Sapphire guy did say in the video that they really are not in to slapping another 30 bucks on their card just because its after market. So you won't pay too much premium over the reference cards. So that price is not surprising.

Gibbo on OCUK wrote a post in which he claims AMD will release a driver fix by the July 5th, so I'm not the only one to have understood that! On the other hand, he claims that it is normal to be over spec and to "voltage variations" if I'm not mistaken, so that doesn't make him sound very reliable :p
 
Gibbo on OCUK wrote a post in which he claims AMD will release a driver fix by the July 5th, so I'm not the only one to have understood that! On the other hand, he claims that it is normal to be over spec and to "voltage variations" if I'm not mistaken, so that doesn't make him sound very reliable :p

My post really had nothing to do what gibbo said. Only thing I believe gibbo on is the prices he promises because he wants to sell the cards and people there really hold him accountable on the pricing he quotes. So he mostly is right about that.

I was talking about the sapphire guy don't know his name. With the big ass beard, saw his video posted when he was talking about the nitro.
 
My post really had nothing to do what gibbo said. Only thing I believe gibbo on is the prices he promises because he wants to sell the cards and people there really hold him accountable on the pricing he quotes. So he mostly is right about that.

I was talking about the sapphire guy don't know his name. With the big ass beard, saw his video posted when he was talking about the nitro.
Oh i wasn't replying to your comment about sapphire, just quoted this post because I saw your name
 
we should probably consider this a $240 part until we actually see the 4gb in stock somewhere.

The 4GB $199 versions have been in and out of stock at Newegg since launch. I've had an opportunity to buy one multiple times but passed.
 
I doubt one of these with only the 6pin would bother mine in the least;). Still i have no problem waiting for the "Improved" cards to hit shelves.
 
Last edited:
The connector on that card is in the front actually pretty sure its got an 8 pin.
 
For us that have Boards with the additional 8 pin plugged into the motherboard are probably less likely to potentially have issues from the extra current drawn from the pci-e slot?
Nope
 
Hard to tell, but looks like a 6-pin to me.

Go to the website for the description, it details an 8 pin PCIe plug. Also, the picture shows it on the end of the card and not the top. (Not that picture but another one.)
 
I just read this on an investing site (so take with implicit grain of salt), but wow if this is true (emphasis mine):

It has been confirmed that AMD designed the RX 480 card so that half of the Voltage Regulator Modules are wired directly to PCI-E bus power supply pins. Any time the card draws more than the Total Dissipated Power of 150 watts, the card will go out of the PCI spec. Many sites have shown that this out of spec conditions can happen running at stock speeds by as much as 27% and at overclocked, up to 50% out of spec. There is no hardware solution nor is there any BIOS or driver solution other than to reduce the voltage by at least 30% which will greatly affect performance.
There is also the specter of another major problem as the 6 pin power connector sense line is grounded instead of correctly being used for sensing if the connector is indeed installed. If AMD has no other safety feature running, this could mean that the card will boot up using only the PCI-E 75 watt power. That will blow up the motherboard (or the card and the mobo) when a game starts to run.
Even the custom board manufacturers may be in deep trouble as well. If they trusted AMD to design their reference card properly (which they didn't) and followed the same design, all of their cards will be out of spec. Even worse, they will be more out of spec than the reference card due to the increased power they are pumping into the card (which has to be matched by the PCI-E power as half the card cannot be "more powered" than the other).
This may indeed be a PR disaster for AMD. It is going to get very interesting!
BTW my source:
Twitch

If the 6-pin sense pin is indeed grounded, booting up the machine without the 6-pin installed would do...?
 
That report makes it seem as if, rather than AIB cards fixing this issue, they make it worse. I may need to wait for reviews before I buy an AIB 480, or settle for a 390.
 
I just read this on an investing site (so take with implicit grain of salt), but wow if this is true (emphasis mine):



If the 6-pin sense pin is indeed grounded, booting up the machine without the 6-pin installed would do...?
I think the sense pin is a ground like the one on the ATX power connector where you ground the green wire to turn on (have this done on a cheap PSU that only runs fans in my Case).
 
I just read this on an investing site (so take with implicit grain of salt), but wow if this is true (emphasis mine):



If the 6-pin sense pin is indeed grounded, booting up the machine without the 6-pin installed would do...?

lol that sounds like bunch of fud to scare the shit out of people. Has been confirmed by who again? Twitch is a source now? lol
 
lol that sounds like bunch of fud to scare the shit out of people. Has been confirmed by who again? Twitch is a source now? lol

Update: within the first few minutes, streamer states the 6-pin is used solely for Vcore and that the GPU chip does not draw any power from the PCI-E bus. I suppose that explains the grounded sense pin and why the excessive PCI-E power draw is blamed on the VRAM.
 
That report makes it seem as if, rather than AIB cards fixing this issue, they make it worse. I may need to wait for reviews before I buy an AIB 480, or settle for a 390.

I actually really doubt AIBs release cards like that. Why because they actually AIB cards through their paces. I highly doubt they don't know about it when they actually hand test the non reference models and then intentionally release it.
 
so the guy is saying the 6 pin is wired just like the 8 pin and and can suck a whole lot of power through it. And memory vrm is powered by the pci-e slot. So makes sense why AMD is saying its because of memory bus. But he said with the way its wired you can feed power to it just like 8 pin...interesting.
 
so the guy is saying the 6 pin is wired just like the 8 pin and and can suck a whole lot of power through it. And memory vrm is powered by the pci-e slot. So makes sense why AMD is saying its because of memory bus. But he said with the way its wired you can feed power to it just like 8 pin...interesting.

This wouldn't make sense. AMD has confirmed that the chip alone is sucking 110W. If that was all via the 6-pin, then PCIE would be drawing 40-50W, not 80+.
 
so the guy is saying the 6 pin is wired just like the 8 pin and and can suck a whole lot of power through it. And memory vrm is powered by the pci-e slot. So makes sense why AMD is saying its because of memory bus. But he said with the way its wired you can feed power to it just like 8 pin...interesting.

Hopefully it an easy fix for AMD. Does make me glad I decided to wait for the AIB cards to come out. If I can get 980 performance or better, then that might just do me until Vega or the Ti are released. This 7950 is really aging atm.
 
This wouldn't make sense. AMD has confirmed that the chip alone is sucking 110W. If that was all via the 6-pin, then PCIE would be drawing 40-50W, not 80+.

Yea its confusing for sure. But looks like he knows what he is talking about the 6 pin connector atleast that it is wired just like the 8 pin and can pull alot more power. So confusing part is if memory vrm powered by the pci-e why its consuming so much power.
 
Yea its confusing for sure. But looks like he knows what he is talking about the 6 pin connector atleast that it is wired just like the 8 pin and can pull alot more power. So confusing part is if memory vrm powered by the pci-e why its consuming so much power.

I'm sorry, I can't agree. "Knows what he's talking about" and "conflicts with reality" gives me alarm bells. I'll take it with a grain of salt.
 
I just read this on an investing site (so take with implicit grain of salt), but wow if this is true (emphasis mine):
you mean a whole fucking salt mine right?

wtf?! its some 18 yr old with TWO twitch streams that we're supposed to trust?! what "investing" site are you talking aboot?
 
hopefully this is just a driver fix, or a bios update for some folks.
 
you mean a whole fucking salt mine right?

wtf?! its some 18 yr old with TWO twitch streams that we're supposed to trust?! what "investing" site are you talking aboot?

SeekingAlpha. The fanboism there is palpable.
 
fuck fanboism, how about stock manipulation?!
wtf are people doing going to an stock investment site for tech info?! I have never heard of it until all shit started with the 480. looking into it, the site could be used to manipulate stocks based on their "crowd sourced info". get enough hype/hate going and watch the stocks soar/tank... now that I know what it is ill be sure to tell anyone quoting it for tech news/info to FOFF/GTFO.
 
Last edited:
fuck fanboism, how stock manipulation?!
wtf are people doing going to an stock investment site for tech info?! I have never heard of it until all shit started with the 480. looking into it, the site could be used to manipulate stocks based on their "crowd sourced info". get enough hype/hate going and watch the stocks soar/tank... now that I know what it is ill be sure to tell anyone quoting it for tech news/info to FOFF/GTFO.

It's a blog-ish site with content written by professional investors for professional investors. You know, those nervous wrecks you see yelling on the street in suits? :ROFLMAO:
 
fuck fanboism, how about stock manipulation?!
wtf are people doing going to an stock investment site for tech info?! I have never heard of it until all shit started with the 480. looking into it, the site could be used to manipulate stocks based on their "crowd sourced info". get enough hype/hate going and watch the stocks soar/tank... now that I know what it is ill be sure to tell anyone quoting it for tech news/info to FOFF/GTFO.

You're not wrong - AMD up over 100% in the last year and I'm sure much of that is due to rabid fanboyism.

The power issue began to be discussed and the drones kept dismissing it outright with no technical info or reviews to back their position. It is AMD mania over there. One author wrote an article less than a day after the power issue was discovered claiming "Idle power exceeding 75W is not a problem" then posted AnandTech total system power consumption charts, stating "See? At 300W in some instances! Disregard this issue, nothing to worry about". He got pummeled by commenters who knew what they were talking about. Article was deleted but comments remained. Author published comments stating he mistakenly used the wrong charts but there's still nothing to worry about, look at the 960! Lesson for all of us: don't listen to someone else's investment advice, do your own research.


I started reading that site for insights into Micron which has, for the most part, been very good. Particularly, read Stephen Breezy's work. His investment advice and speculation hasn't been good (i.e. he lost all of his money on LEAPS, the poor sap) but his technical info is quite enlightening, particularly around HMC and 3DXP. Kind of the exact opposite of what you'd expect from an investment site.
 
Back
Top