Radeon RX 480 Performance Benchmarks with 16.7.1 driver

It's probably the only way to drop frequency/voltage to meet the pci-e spec for amps of motherboards. They don't want to decrease the RX480's performance.....

 

The 750Ti's were around 60W continuous draw but would get spikes to the 120's all that was within spec.

Honestly I didn't like that too much and most AIB partners didn't either as they added 6pin plugs to most 750Ti's.

Careful that's fanboy speak to vase's ears. You may have to give your self an infraction.

 

I have seen no people in this thread even remotely admitting they were wrong on calling downgrades moment AMD made official statement about this.

But I see here some of them anyway, making more uneducated comments. Or pretending that they do not know better.

 

You didn't see the third post in this thread then when I said it's good to see there was no drop in performance.

Did it hurt your feelings.

The concerns about a performance drop were reasonable. Stop defending a company that just uses you for money and demand better products.

 

They changed way current is drawn through VRMs. RX-480 has so overpowered VRMs that it could have used same board with 8-pin and just feed it through 3 VRMs connected to it (Ignoring PCIe).

There is test with original RX-480 configuration. It had OC + power limit increased to point card ate 260W.
No problem at all. Board can shut you down if you are too hungry. And on top of that we still have not seen any real confirm-able damage this caused.

Yes, AMD used higher than necessary voltages. But their main mistake was in thinking that there would not be group of people with interest in harming them any way possible.
From what I have seen, at least 3 fake accounts were used to support false claim. One including our site.
Dozens got benefit of doubt. But none really could prove actual damage caused by card.

Because if specification tells you that CPU can draw 95W, manufacturer builds design which does not burn at 125W. Or they lock CPUs with higher power consumption of of socket or limit them.

And as being said: "Where is The Will, there is The Way."
And this case shown that people did more damage than actual cards did.

 

I'll be honest with you and maybe you'll learn something.
There is day and night difference between admitting mistake/saying as little as "sorry" and stating that "Actual reality is good" after it hits you.

No, sorry, you are of no consequence and as such I have no emotional attachment to anything you can say. At worst I just wonder where did it came from.

And for rest of your comment. AMD did what I said they will because I spent time actually looking for board analysis and way it works. So, yes, expecting downgrade is reasonable till you know better.

Ending your comment with jokes? Or are you into self inflicted wounds now. Or are you projecting your mental state to others?
I do not really know, but it is laughable.

 

I don't take sides in this. I want both AMD and Nvida to be strong. AMD has **** the bed too many times as of late and Nvida has capitalized on it by screwing every customer over.

YOU can continue to stick your head in the ground and sing' "la la la la nothing to see here" all you want but the fact remains AMD is tripping up left and right. This problem should have never HAD to be "fixed" after release.

You look at my critique of AMD and all you see is an Nvidia fanboy and that helps you cope with it when all you are showing is your own AMD bias.

I want AMD to stick it to Nvidia but sadly at the moment I don't think they have smart enough people running the company to do that. And their mindless, AMD is my savior all hail AMD, drones out there don't help them any.

 

Fox2232: This AMD problem hasn't helped us the consumers at all. What's actually happened is it's justifying pascal prices which is basically f****** us over.

I fully believe in part due to RX480 power problem that it's caused some cards like mine to cost much more as there's no competition.

My GTX1070 was £420. It's now £470. Brexit can't account for a £50 price hike. I completely put this down to no competition from AMD causing this. We're getting rinsed for every penny basically.

Looking at the current RX480 situation; how can people like me recommend it? I can't. I'd like to, but, I don't want people coming back to me if things do go wrong. Real life recommendations are accountable, unlike the internet.

 

Well in the US, GTX 1070 prices have been the same before and after RX 480 launch. Also, I just bought an item I was looking for that came from England, and with the rate conversion I saved money because of the better dollar to pound.

You act like AMD causes cancer....

 

You just don't get it, period.

 

I'll paraphrase you:
"I want AMD to force nVidia prices down, so I can buy nVidia cheaper."

That mentality is what allows nVidia to increase price. Because they know that you'll buy their card anyway. Be it upon launch or 3 weeks after at same or higher price.

You are one who's processed and placing guilt on someone else.

Only guilty party in business is always customer who pays for something he does not consider a good deal.

Now you have few years to grow a pair and next time you can take a stand. Alone as you are not going to persuade rest of nVidia's purse holders.
- - - -
So once more. "If you complain that you made a bad deal, then you are only guilty party in that deal."

 

Don't get me wrong, I got a good deal especially looking at it from todays perspective. I just feel bad for those who come after me, especially those expecting prices to come down a month after launch.

Pascal release itself, especially after GTX1070, did drive down prices of Nvidia's cards anyway, especially noticeable with GTX980ti which is something.

A lack of competition from AMD side isn't consumer's fault tbh.

 

How is this on topic?

 

This is a non-issue at this point. There is still the "Compatibility" tab on offer, if people's equipment is that bad. Although if your motherboard can't work with a card that draws max 70W from the PCIe slot, then you most likely have other problems.

 

What's there to discuss? Performance with new driver is fine.

 

After this update the compatibility mode isn't needed for those boards as all it does is throttle the card to 150W. The driver itself addresses the PCIe over current at all times.

 

You do not need competition.Monopoly is not optimal state, but only because people will pay whatever monopoly asks for product.

If everyone had enough guts to say: "Stop! No More!" price would move down even with one company having monopoly.

It is similar to stock market crashes. That entire system has nearly no safeguards and stands on belief that people are rational and calculative and that they buy for good price. But in reality people do not know much about what they are doing, especially when things get crazy.

This can easily happen to PC hardware prices. But it takes company to realize that they are losing customers to their own high prices.
- - - -
As for GTX 1070 price. Before GTX 980Ti dropped down it looked very impressive. But with mentioned price drop GTX 980Ti is more attractive for OC crowd and GTX 1070 is more attractive for casual crowd.
And it looks like I did read your wording wrongly. It made impression that you are mad at price you had to pay. My bad.
- - - -
Anyway, if this not so great (from PR perspective) launch of RX-480 inflated nVidia's prices. You can thank partly to AMD, but mostly to those who blew it out of proportions.
But I personally think that it is mainly due to RX-480 having bit higher price than expected in Europe.

 

Allright, with 16.7.1 driver we're having a driver fix for the RX 480 and it works something like this: "Starting with the 16.7.1 driver, AMD will automatically reprogram the power controller on the RX 480 to better divide the power draw from the 6 available phases on the reference cards. This is a process that occurs at each and every boot, it is not a permanent change to the VBIOS. It’s possible, and likely, that future cards and partner cards may have this change integrated at a lower level, negating the need for the driver to recognize and update the controller logic. (credits to pcper for this explanation)".
One more thing that we need to see now is if and how RX 480 will overclock after this 16.7.1 driver fix, plese HH can you give us an update on this for the reference model? Thanks.