AMD: Speculation, Rumors, and Discussion (Archive)

Status
Not open for further replies.
After seeing the the rx480 you think that will happen? I would be very skeptical.

I would be very skeptical too, but after so many problems with lawsuits(Bulldozer era, and Phenom versus Core2 era too), how could they were lying on Polaris 11(that shown huge efficiency gain over GTX 950) presentation? And on the AMA they stated that RX470 haves ~2.7 times better efficiency over a 28nm GPU(Ok, from 390x to Nano things vary very much, but 2.7x efficiency gain puts from 10-20% behind GTX1080 to far more efficient than anything on market, all depending on what GPU you are comparing it). I don't believe they missed the target with the whole product line not because of optimism/pessimism, but just contradicts data we're already got from them.
 
So basically what we have here is a very strange behavior indeed, a group of people willingly determined to look the other way on any out of line behavior from AMD. Just because they are the underdog. A card that is out of spec so what? It's only damaging when OC'ed! as if this is suddenly a good thing, instead of talking about how good a card overclocks, now we talk about how it's the user's fault if he damages his hardware! because hey, It's AMD, they need our support, right?

I guess they will forgive driver problems or lack of support or whatever new screwup further down the line. Because hey, they have no resources, we need to support them. By that logic we should have video cards from VIA and S3 as well, doesn't matter if they produce crappy hardware, we will support them regardless!

Well evolution and survival doesn't work that way. Screw ups and half assed architectures are not going to carry your company forward. It hasn't worked in the past with AMD CPUs and it's not going to work now with GPUs. If people are afraid prices will go up if AMD goes under, then guess what, prices will still go up, even with AMD around. Only difference is AMD will be producing more low quality stuff until they become completely irrelevant, then new competitors will take their place. And that's Evolution for you. That's how technology works, and that's how the world advances, not the other way around.
 
We'll see what the perf/W is on RX470.


actually they did say the rx 480 was 2.8 perf.watt more in certain circumstances against a 280x, it was clearly in one of the slides in the footnotes. I don't know why AMD marketing is putting their foot in their mouth so many times when its in black and white........

Seriously this whole mess has not only made AMD look incompetent, but now RTG is following suite. Apple doesn't fall from the tree is true.
 
Last edited:
So basically what we have here is a very strange behavior indeed, a group of people willingly determined to look the other way on any out of line behavior from AMD. Just because they are the underdog. A card that is out of spec so what? It's only damaging when OC'ed! as if this is suddenly a good thing, instead of talking about how good a card overclocks, now we talk about how it's the user's fault if he damages his hardware! because hey, It's AMD, they need our support, right?

I guess they will forgive driver problems or lack of support or whatever new screwup further down the line. Because hey, they have no resources, we need to support them. By that logic we should have video cards from VIA and S3 as well, doesn't matter if they produce crappy hardware, we will support them regardless!

Well evolution and survival doesn't work that way. Screw ups and half assed architectures are not going to carry your company forward. It hasn't worked in the past with AMD CPUs and it's not going to work now with GPUs. If people are afraid prices will go up if AMD goes under, then guess what, prices will still go up, even with AMD around. Only difference is AMD will be producing more low quality stuff until they become completely irrelevant, then new competitors will take their place. And that's Evolution for you. That's how technology works, and that's how the world advances, not the other way around.

Maybe the reason why some of us don't care much about PCIe specs is the fact they were violated many time before and there was 0 outrage? I'm sure this will change now since obviously this is the new norm.
 
actually they did say the rx 480 was 2.8 perf.watt more in certain circumstances against a 280x, it was clearly in one of the slides in the footnotes. I don't know why AMD marketing is putting their foot in their mouth so many times when its in black and white........

Seriously this whole mess has not only made AMD look incompetent, but now RTG is following suite. Apple doesn't fall from the tree is true.

From what I've seen on AMD slides the only 2.7 perf/W was RX470 compared to 270 I think. But I could be wrong so feel free to correct me.
 
Maybe the reason why some of us don't care much about PCIe specs is the fact they were violated many time before and there was 0 outrage? I'm sure this will change now since obviously this is the new norm.
First third of video will explain why ...
 
Only that didn't happen, even the ASUS strix GTX 960 didn't violate the specs on sustainable load.
http://www.pcper.com/reviews/Graphi...s-Radeon-RX-480/Evaluating-ASUS-GTX-960-Strix
Even the dual chip cards that suck huge amount of power did so through the 8-pin connectors and not the PCI-e.

Exactly they did it through PCIe molex connector but violated it non the less. Or how do you explain 295 having only 2x8 pin and consuming up to 600W. 0 outrage. Heck it wouldn't surprise me if every Hawaii GPU is out of spec.

So if the specs are there for a reason as now everyone is saying, they are probably there for the same reason when it comes to power being delivered over the connector.
 
Last edited:
Exactly they did it through PCIe molex connector but violated it non the less. Or how do you explain 295 having only 2x8 pin and consuming up to 600W. 0 outrage. Heck it wouldn't surprise me if every Hawaii GPU is out of spec.

So if the specs are there for a reason as now everyone is saying, they are probably there for the same reason when it comes to power being delivered over the connector.
It's trivial to upsize the power wires to a molex connector, and one can reasonably expect that high power PSUs that are needed for something like a 295X2 do exactly that.

Not true for PCIe motherboard connector pins, where you're basically stuck with one size.
 
Exactly they did it through PCIe molex connector but violated it non the less. Or how do you explain 295 having only 2x8 pin and consuming up to 600W. 0 outrage. Heck it wouldn't surprise me if every Hawaii GPU is out of spec.

So if the specs are there for a reason as now everyone is saying, they are probably there for the same reason when it comes to power being delivered over the connector.
But as I keep saying, there is a much lower value and only 1 specification standard with the pcie x16 slot compared to the Molex-ATXV12 spec and auxiliary connectors.
This is why as I showed AMD with the 295x2 pushed the out of spec current through the auxiliary Molex connectors, as I mentioned each of the 8-pin connectors actually supports 288W (caveat as before down to a good PSU and associated cables)...
And this shows something went wrong with the 480, because I doubt they would intentionally put such pressure on the mainboard PCIe slot, especially when they are looking to promote mGPU and 2x480.

BTW That 288W is also just the standard rating and not even the HCS spec introduced in ATX12V v2.2
And none of that is applicable to the PCIe slot (apart from the indirectly associated ATX 24-pin connector that supports all slots and some other devices without separate power - here the ATX 24-pin in standard is 6A per 12V contact but it also has a spec for HCS).
Cheers
 
Last edited:
they can't have some bits that are turned from 0s to 1s if it gets written too from wattman or over drive or something ? I would think they could just code that in so when you check okay to overclocking it shows up in the bios
Well it* hasn't happened before with OverDrive, why would it happen now with WattMan?
*it being warranty denied because OC'ing via software
 
Exactly they did it through PCIe molex connector but violated it non the less. Or how do you explain 295 having only 2x8 pin and consuming up to 600W. 0 outrage. Heck it wouldn't surprise me if every Hawaii GPU is out of spec.
Having 2x8 pin was already pretty up-front about being out of spec, as that isn't a valid option.
If every Hawaii that physically appears to be in spec truly is out of spec, then AMD would deserve some criticism.

edit: At the time I wasn't outraged, but it did give me pause.
 
Last edited:
PcPer's re-evaluation of the Asus GTX 960
One of the first responses from the AMD community when we posted our story about the power draw concerns on the new Radeon RX 480 was to point out that apparently the ASUS GeForce GTX 960 Strix card had a similar issue. Keep in mind that the GTX 960 reference specifications put the TDP of this product at 120 watts, 30 watts lower than the RX 480, though it still utilizes a single 6-pin power connection in addition to the motherboard supplied power.
...
With our ability to measure voltage and current, we return back to Metro: Last Light at 4K to find that amperage draw over the motherboard's +12V line stays right at 2.5A. That is a drastic difference compared to the RX 480 hitting more than 7A over the same line, especially considering the 5.5A limit from the PCI Express specification.
So there you have it - while I cannot say for certain that NO previous graphics card in recent memory hasn't behaved in the same fashion that the new AMD Radeon RX 480 does, I can categorically discount the notion that the ASUS GTX 960 Strix is somehow equivalent in its power delivery.
http://www.pcper.com/reviews/Graphi...s-Radeon-RX-480/Evaluating-ASUS-GTX-960-Strix
 
It's starting to be called "powergate."
For once, I can live with a controversy having the -gate moniker! It's just too perfect. ;)
It starts to be a PR nightmare with all these articles, videos, forum talks and reported dead motherboards. Very soon, even average Joe will know and conversations will be like:
A.Joe1: "I fried my mobo" :no:
A.Joe2: "did you buy a RX480?" :runaway:

A.Joe1: "I want a new vid card" :idea:
A.Joe2: "Dont' get a RX480, it will fry your mobo" :runaway:

Its very obvious AMD pushed the clocks/voltage at last minute (after PCIe certification) to be more competitive but they had no time to redesign the boards (already in production). And now they pay the consequence...
 
Last edited:
It starts to be a PR nightmare with all these articles, videos, forum talks and reported dead motherboards. Very soon, even average Joe will know and conversations will be like:

Its very obvious AMD pushed the clocks/voltage at last minute (after PCIe certification) to be more competitive but they had no time to redesign the boards (already in production). And now they pay the consequence...


Well its going to be the OEM's first, they can't use this card as is. Then its going to come down to the local system builders, they don't want to warranty a system that has that card in it.

The average joe will not know anything about it but hear it through the grape vine its a problem card and go else where.
 
Maybe the reason why some of us don't care much about PCIe specs is the fact they were violated many time before and there was 0 outrage? I'm sure this will change now since obviously this is the new norm.

This line of argument has little merit. It is also much overused.

The fact that something similar may or may not have occurred in the past and author can not recall a specific individual or a group of people reacting in the same way has no bearing on anything, pretty much ever.

Other tangents about how AMD deserves credit for trying to do the best it can with limited resources, how competition is good for all, and that marketing people are scum are also as true as they are irrelevant.

There's an issue here, and it happens at stock speed. That is something potential buyers should be aware of. Do you agree?
 
This line of argument has little merit. It is also much overused.

The fact that something similar may or may not have occurred in the past and author can not recall a specific individual or a group of people reacting in the same way has no bearing on anything, pretty much ever.

Other tangents about how AMD deserves credit for trying to do the best it can with limited resources, how competition is good for all, and that marketing people are scum are also as true as they are irrelevant.

There's an issue here, and it happens at stock speed. That is something potential buyers should be aware of. Do you agree?

I was not talking about specific individuals reacting one way or another I was talking about reviews. I can't remember PCIe specs ever being a front page news. Clearly this has changed now and is the new norm. And I also gave an example where specs were clearly violated and this was not made a huge issue or a 'gate'.

As for AMD deserving credit because it's somehow financially handicapped, I don't think I ever said that. Frankly, I couldn't care less since I'm currently just a GPU enthusiast and a sucker for perf/$, not an investor.
 
Status
Not open for further replies.
Back
Top