Conflicting driver versions in Fury X reviews

TaintedSquirrel

TaintedSquirrel

[H]F Junkie
Joined
Aug 5, 2013
Messages
12,689
Aside from Cat 15.5 vs 15.15, there appear to be multiple revs of 15.15 itself under the same name, only differing by date.

http://www.reddit.com/r/buildapc/comments/3b30bt/discussionfury_x_possibly_reviewed_with_incorrect/

http://www.reddit.com/r/pcmasterrac...d_probably_pay_more_attention_to_what/csia5t7

This would at least help explain the discrepancy between various review sites' FX results.

On a personal note, I put the blame squarely on AMD. It's their job to make sure reviewers have access to the proper, up-to-date drivers. So if this ends up being a serious concern then I think it's time for AMD to make an official statement about which driver package people are supposed to be using with the Fury X.

If nothing else then hopefully this "mess" will be squared away by the time Fury launches so we can get proper numbers (assuming that is the case).
 
The day before the launch I double checked the download location AMD sent us for drivers, and it was still the same driver we used, so nothing changed officially from AMD. The driver they told us to use, was still the same driver to use, right up to the launch date. I wanted to ensure we had the latest build, and it never did change, nothing newer came out for us to use.
 
I JUST NOW re-checked the download location for the driver, for press, and the June 15th driver is still up as the driver to use for AMD R9 300 series and Fury X Reviews. Same exact driver we used in our evaluation, still up, still un-changed, no newer builds.
 
And it is still a beta driver :p. Seriously, the last non-beta driver is from last year. That is just way too silly. It isn't like the hardware launch was a surprise, they know their own product schedule. There should have been a stable, non-beta driver available for this for launch.
 
Brent_Justice said:
I JUST NOW re-checked the download location for the driver, for press, and the June 15th driver is still up as the driver to use for AMD R9 300 series and Fury X Reviews. Same exact driver we used in our evaluation, still up, still un-changed, no newer builds.
Click to expand...

It's OK, it'll get spun into you guys are still biased somehow.
 
just checked amd.com has 15.15.1004 and june20.exe instead of june 15

anyone has a fury to check if they done goof ? :p
 
Brent_Justice said:
where are you seeing that? link please

when i go to amd.com and drivers 15.6 beta comes up
Click to expand...

Hi Brent, thanks for doing the review and helping us with some clarifications. Do you think you could get us a screenshot of the Catalyst software page that shows the driver version? I'm just curious, because Catalyst variants are not the same number as the drivers. It's stupidly confusing!

Also, the version you had was Catalyst 15.15 right, not 15.6? Really wish they would change their naming convention.
 
glr123 said:
Hi Brent, thanks for doing the review and helping us with some clarifications. Do you think you could get us a screenshot of the Catalyst software page that shows the driver version? I'm just curious, because Catalyst variants are not the same number as the drivers. It's stupidly confusing!

Also, the version you had was Catalyst 15.15 right, not 15.6? Really wish they would change their naming convention.
Click to expand...

Crazy thing is they actually have driver changes that use the same number driver revisions as well...like they just change the dates and ya have to wonder what if anything was changed....just doesn't help
 
Well if this ever comes to pass, it shows how disorganized AMD really is
 
primetime said:
Crazy thing is they actually have driver changes that use the same number driver revisions as well...like they just change the dates and ya have to wonder what if anything was changed....just doesn't help
Click to expand...

The problem is that the drivers and catalyst all have different versions, but for whatever stupid reason the drivers are now into the 15.x nomenclature. This makes it so incredibly confusing. Catalyst seems to have 3 versions, 15.5, 15.6 beta and 15.15 for the 300 series. 15.5 and 15.6 seem to use the 14.x drivers, whereas 15.15 uses the 15.2xx series drivers which are similar, albeit not the same, as the test drivers for Windows 10.

I can see how this would all get messed up, and I suspect that a lot of review sites were all using different drivers or something. I think we really need a screenshot of the Catalyst software page from multiple reviewers to actually see what driver was being used.
 
Even if this is a driver issue causing the "lack" of performance.

Say they do come out with a statement saying use driver x over driver y for optimum performance.

In the past drivers only increase performance on average of 5% for specific titles!

Correct me if I am wrong, but this seems like the beginning of the end for all of us.
 
OofC_S7v7N said:
Even if this is a driver issue causing the "lack" of performance.

Say they do come out with a statement saying use driver x over driver y for optimum performance.

In the past drivers only increase performance on average of 5% for specific titles!

Correct me if I am wrong, but this seems like the beginning of the end for all of us.
Click to expand...

Beginning of the end? lol....So far there selling as fast as they can make and ship them:rolleyes:
 
primetime said:
Beginning of the end? lol....So far there selling as fast as they can make and ship them:rolleyes:
Click to expand...
As long as they sell every unit they produce for the price they're asking, I guess nothing else really matters.
I just wish AMD were making more Furys that way their sales would actually suffer since they deserve it.
 
TaintedSquirrel said:
As long as they sell every unit they produce for the price they're asking, I guess nothing else really matters.
I just wish AMD were making more Furys that way their sales would actually suffer since they deserve it.
Click to expand...

here is what I don't understand. If you are not satisfied with something you have other options. I for once have had mostly nvidia, I would never want amd to suffer. Listen this is a product that is competitive with nvidia top end. So why the hell would you want amd to suffer so Nvidia is the only company selling us cards and you can be satisfied with up the ass prices because you have no other choice? I sometimes don't get the reasoning. Hating a company is one thing but you wanting them to suffer and have no competition in the market is another.
 
Sycraft said:
And it is still a beta driver :p. Seriously, the last non-beta driver is from last year. That is just way too silly. It isn't like the hardware launch was a surprise, they know their own product schedule. There should have been a stable, non-beta driver available for this for launch.
Click to expand...

But AMD fanboys claim there are no more AMD driver issues
 
NKD said:
here is what I don't understand. If you are not satisfied with something you have other options. I for once have had mostly nvidia, I would never want amd to suffer. Listen this is a product that is competitive with nvidia top end. So why the hell would you want amd to suffer so Nvidia is the only company selling us cards and you can be satisfied with up the ass prices because you have no other choice? I sometimes don't get the reasoning. Hating a company is one thing but you wanting them to suffer and have no competition in the market is another.
Click to expand...

Wanting them to suffer actually has nothing to do with them. I'm sure they're nice people, just trying to make a living and keep their jobs. Hating on AMD is about their tard fanboys that reach for every straw they can grasp, always running their mouth. I don't hate AMD, I detest their idiot fans. Just like the Seahawks.

To be clear on your comment about Nvidia reeming us up the ass with prices if AMD wasn't around. That isn't correct. Prices wouldn't change, as prices are dictated by demand. What would go slower is innovation of new product. Nvidia would not be as rushed to bring out new tech as fast. It's like Intel right now with AMD cpu's.

It's a common misconception that a monopoly means high prices. If people don't like the prices, regardless if it's the only company, they won't buy that card, they'll wait or buy a lesser model. If a monopoly has a whole wide product range at different prices, then they can't really price gouge. Demand controls that.
 
glr123 said:
The problem is that the drivers and catalyst all have different versions, but for whatever stupid reason the drivers are now into the 15.x nomenclature. This makes it so incredibly confusing. Catalyst seems to have 3 versions, 15.5, 15.6 beta and 15.15 for the 300 series. 15.5 and 15.6 seem to use the 14.x drivers, whereas 15.15 uses the 15.2xx series drivers which are similar, albeit not the same, as the test drivers for Windows 10.

I can see how this would all get messed up, and I suspect that a lot of review sites were all using different drivers or something. I think we really need a screenshot of the Catalyst software page from multiple reviewers to actually see what driver was being used.
Click to expand...

Like the other guy said, this is the very definition of reach for straws. It didn't destroy, let alone beat the 980ti, nevermind that it was originally aimed to compete with Titan x, but because it didn't beat it, something must be wrong. This just cannot be. HBM was supposed to deliver 500fps, a hot girlfriend, and a winning lottery ticket, endless massages and no more hemorroids, and if it didn't, then it MUST be because a driver is .1 version number off.

Dude, if this version stuff was true, multiple websites would have noticed and more would have said something. We'd also be seeing a bigger range of variation in the benches but every site is showing the same results.......

..Fury x is trounced at 1080 and 1440, and only almost ties the 980ti at 4k. It's a good product, but it still comes in 2nd place, still uses a lot of power, isn't a direct x 12.1 card (12.0 only), doesn't OC nearly as well as Maxwell, NEEDED liquid cooling to compete, and it's still tied to AMD's so so drivers.

It's evident they still have driver issues as that's the entire premise of this thread
 
Saying it is the beginning of the end to me means something different I guess.

Like you said, they can't keep em on the shelf at this price.

Even if a driver update or proper drive usage increases performance.

It will not be a mind boggling increase.

The beginning of the end is referring to the fact that people are eating this up,
when AMD was praising the Fury X as the fastest videocard on the planet.

Yet it is having a hard time beating a card that came out a few months or so before it!
 
Ultranifty said:
Like the other guy said, this is the very definition of reach for straws. It didn't destroy, let alone beat the 980ti, nevermind that it was originally aimed to compete with Titan x, but because it didn't beat it, something must be wrong. This just cannot be. HBM was supposed to deliver 500fps, a hot girlfriend, and a winning lottery ticket, endless massages and no more hemorroids, and if it didn't, then it MUST be because a driver is .1 version number off.

Dude, if this version stuff was true, multiple websites would have noticed and more would have said something. We'd also be seeing a bigger range of variation in the benches but every site is showing the same results.......

..Fury x is trounced at 1080 and 1440, and only almost ties the 980ti at 4k. It's a good product, but it still comes in 2nd place, still uses a lot of power, isn't a direct x 12.1 card (12.0 only), doesn't OC nearly as well as Maxwell, NEEDED liquid cooling to compete, and it's still tied to AMD's so so drivers.

It's evident they still have driver issues as that's the entire premise of this thread
Click to expand...

I agree with a lot but how many times does it have to be repeated that it doesn't need water cooling and it's not required! God that is the Most ignorant statement that I read all the time. Be subjective don't just throw fud! There are going to be air cooled versions of fury coming.
 
NKD said:
I agree with a lot but how many times does it have to be repeated that it doesn't need water cooling and it's not required! God that is the Most ignorant statement that I read all the time. Be subjective don't just throw fud! There are going to be air cooled versions of fury coming.
Click to expand...

no, isn't needed in a typical card, but they were forced to use water cooling to be able to cool the card in that small form factor.. also I could see people crying if they instead were choosing to use a regular air cooler of being hot and loud.. water isn't needed but is necessary in this case because also the VRM are cooled with a heatpipe attached to the AIO waterblock. also, it reduced the overall product cost, as the R&D cost to make a new air cooler are bigger than just take a AIO and just attach it, is just cheaper, which is good.
 
Araxie said:
no, isn't needed in a typical card, but they were forced to use water cooling to be able to cool the card in that small form factor.. also I could see people crying if they instead were choosing to use a regular air cooler of being hot and loud.. water isn't needed but is necessary in this case because also the VRM are cooled with a heatpipe attached to the AIO waterblock. also, it reduced the overall product cost, as the R&D cost to make a new air cooler are bigger than just take a AIO and just attach it, is just cheaper, which is good.
Click to expand...

The cooler is custom made and costs ~$85 according to the video posted in the other thread. They are also releasing a version that is air cooled.
 
Keep spinning lies AMD, that's what u're capable of.

I am sick of excuses. If it was a half decent launch we would have seen a quiet AMD.
 
OofC_S7v7N said:
Saying it is the beginning of the end to me means something different I guess.

Like you said, they can't keep em on the shelf at this price.

Even if a driver update or proper drive usage increases performance.

It will not be a mind boggling increase.

The beginning of the end is referring to the fact that people are eating this up,
when AMD was praising the Fury X as the fastest videocard on the planet.

Yet it is having a hard time beating a card that came out a few months or so before it!
Click to expand...

I'm willing to bet Fury's are being sold to WoW players, who else would buy this over a 980Ti....
 
This guy got a 10% improvement in Firestrike with a different driver (modded W10 driver): http://forums.guru3d.com/showpost.php?p=5108529&postcount=642

I do not think that it will beat 980Ti by 20%, but it will boost performance really soon.
I did another experiment:
Firestrike-Driver-Improvements
See graphics scores:
1st release driver: 15664
2nd 15.200.1040.0 - Tonga mode: 16320
3rd 15.200.1040.0 - Hawaii mode: 17296

Hilbert's release driver score: 16081
Any guesses where this can go with proper code path?
Click to expand...

That said, he didn't test any games. Why do these people not TEST ANY GAMES. All they have to do is boot up Witcher 3 or GTA5 or fucking whatever and check.
 
NKD said:
I agree with a lot but how many times does it have to be repeated that it doesn't need water cooling and it's not required! God that is the Most ignorant statement that I read all the time. Be subjective don't just throw fud! There are going to be air cooled versions of fury coming.
Click to expand...

If they keep repeating it eventually people get tired of refuting it and they win. nVidia does this with all their marketing. Ever wonder why you see 15 reviews a day of slightly different versions of the same model card? Saturation. They use he same exact technique with their stealth marketing. Then their fanboys regurgitate it.
 
You must log in or register to reply here.
Back
Top