AMD Catalyst Mantle Driver Delayed

[btarunr]

As mentioned in our older article from Thursday, AMD discovered a major bug with its Catalyst 14.2 driver at the last minute, which threw a wrench in the works at the company. The driver was originally slated to come out in sync with DICE' update of Battlefield 4 on Thursday. AMD now tells us that their teams have been working overnight to fix the driver, and should have their next status update for us by mid-afternoon EST (New York time), later today. That update doesn't necessarily mean a driver release at that point in time, and so AMD might be forced to label it Catalyst 14.2 beta, keeping up with its calendar-based driver version naming. Catalyst 14.1 beta was expected to ship the first public distribution of Mantle, AMD's ambitious 3D graphics API to rival Direct3D and OpenGL.

AMD also mentioned a 24 hour exclusive period for press to evaluate the driver before public release, so it looks like your download will be at least 36 hours away.



View at TechPowerUp Main Site

 

[dj-electric]

Developers are wayyyy too stressful with these things. Take your time, guys. I don't want more unfinished products 'cause a bunch of douches decided this must come out asap.

Personally, i don't care waiting another month or so for stable drivers and support.

 

[mr2009]

Developers are wayyyy too stressful with these things. Take your time, guys. I don't want more unfinished products 'cause a bunch of douches decided this must come out asap.

Personally, i don't care waiting another month or so for stable drivers and support.

+1 for this comment.

 

[Ja.KooLit]

well atleast they found out first and try fix them. i do hope they will fix it soon. the much awaited mantle has been delayed long time and i do hope it is worth the wait.

 

[HumanSmoke]

Developers are wayyyy too stressful with these things. Take your time, guys. I don't want more unfinished products 'cause a bunch of douches decided this must come out asap.

Beats me why some companies habitually allow their PR departments to drive their software/hardware schedules.
Classic example of the tail wagging the dog

Personally, i don't care waiting another month or so for stable drivers and support.

I don't think the end user factors into the equation. EA have made an art form out of half-assing their way through product launches. AMD, equally, have a record of extreme optimism in calculating software completion (Enduro, frame pacing, and GPU release drivers being relatively high profile examples). You would think that both companies might have learned not to set unrealistic deadlines at the expense of a polished product....short memories, or something worse?

 

[Fiery]

Beats me why some companies habitually allow their PR departments to drive their software/hardware schedules.

Without a pressure from marketing folks, developers would take years to complete something. Not because they're slow or lazy or careless, but because a software or driver is never completed, never done, never being perfect. You just release it at a point and hope that no major bugs are left in there Then you go on fixing existing bugs, and add some more features that will bring more bugs But without a strict release schedule, developers just sink into polishing certain parts of the software for many months, without having anything to roll out.

It's another question when a company keeps promising new stuff, but fails to deliver them on time in almost every case. Such companies mostly do not last as long as AMD

(BTW, I'm a software developer)

 

[NC37]

Whats this...a company keeping the public appraised of bugs and actually acting responsibly and making sure the product is presentable before it is released to market?!

EA execs cry out...BLASPHEMY!!!

 

[tokyoduong]

Beats me why some companies habitually allow their PR departments to drive their software/hardware schedules.
Classic example of the tail wagging the dog

I don't think the end user factors into the equation. EA have made an art form out of half-assing their way through product launches. AMD, equally, have a record of extreme optimism in calculating software completion (Enduro, frame pacing, and GPU release drivers being relatively high profile examples). You would think that both companies might have learned not to set unrealistic deadlines at the expense of a polished product....short memories, or something worse?

I agree that EA is just plain inefficient with their resources. I don't think we can say the same about AMD. Look at what they've tackled and the small amount of resources(both capital and labor) they have to work with. Mantle is a complex project. On top of that, it is not something that's in house. When you have to collaborate with other businesses like EA, your confidence to meet any deadlines drop down to 5% lol.

Without a pressure from marketing folks, developers would take years to complete something.

No, marketing is not where is pressure is. The pressure is from management and project managers. I've never seen seen marketing department set up schedules for any type of project besides their own. Each department(team) manage their own schedule and has to make it fit within the whole organizational goals. Hence the need for the CEO to tie everyone together and point them all in one direction. At least this is what is supposed to happen in an efficiently managed organization.

 

[NeoXF]

Well, it was the right thing to do and I can totally get that. If there's anything to hate here it's AMD's slow driver progression (slacky developers and/or low budget dedicated to it) and QA teams for not tracking it earlier.

 

[tokyoduong]

Well, it was the right thing to do and I can totally get that. If there's anything to hate here it's AMD's slow driver progression (slacky developers and/or low budget dedicated to it) and QA teams for not tracking it earlier.

I'm 100% sure AMD's budget for their driver team is a joke. It also sounds like they don't have a large team either. If you also look at their operating expenses, it is extremely low compared to their peers. Everything points to limited resources. This is not good for AMD in the long run. I hope they find their needed sales in their console contracts and upcoming high density low power SeaMicro servers.

As of now, the word is that Rory Read will keep their expenses below $450M for the next year. That is very low for the projects they take on. It's a drop in bucket when you look at Intel's budget allocations.

 

[NeoXF]

I'm 100% sure AMD's budget for their driver team is a joke. It also sounds like they don't have a large team either. If you also look at their operating expenses, it is extremely low compared to their peers. Everything points to limited resources. This is not good for AMD in the long run. I hope they find their needed sales in their console contracts and upcoming high density low power SeaMicro servers.

As of now, the word is that Rory Read will keep their expenses below $450M for the next year. That is very low for the projects they take on. It's a drop in bucket when you look at Intel's budget allocations.

The fact that a company that is pushing so much for a ton of software-specific tie-ins to it's hardware (HSA, Mantle...). that don't make or break their hardware, but if they'd work as intended, will surely bring in sales of several magnitutes higher than now... is allocating so little resources to it's software development/partenerships is quite mind-boggling. Don't they have enough hardware concepts/researches/designs for the moment to skimp on those costs and move the money to software dev/Q&A?

 

[MarcusTaz]

Developers are wayyyy too stressful with these things. Take your time, guys. I don't want more unfinished products 'cause a bunch of douches decided this must come out asap.

Personally, i don't care waiting another month or so for stable drivers and support.

Right on man, totally agree with you... To many kiddy trolls. As it is the game is still an Alpha compared to BF3 with much work to be done on the back-end compared to eye candy. Balance the game out and completely nerf the MAA so both teams can flay again.

 

[tokyoduong]

The fact that a company that is pushing so much for a ton of software-specific tie-ins to it's hardware (HSA, Mantle...). that don't make or break their hardware, but if they'd work as intended, will surely bring in sales of several magnitutes higher than now... is allocating so little resources to it's software development/partenerships is quite mind-boggling. Don't they have enough hardware concepts/researches/designs for the moment to skimp on those costs and move the money to software dev/Q&A?

AMD's plan for the past decades is based on their vision of the future running tiny cores but a ton of them. Mantle is necessary to boost gaming and more specifically mobile gaming for the long term. Their servers movement direction also reflect that. Think about it, I don't see people sitting around with controllers staring at just one TV screen in the future. Every consoles this generation have already incorporated additional screens. It's obvious that future consoles, if they will still call it that, will be more mobile. You will be more connected and your senses will be bombarded with even more information. Mobile gaming will be mainstream because our hardware is getting to that point that it can handle more serious games. With high density low power cloud servers to offload processing, it is a definite future for mobile gaming.

AMD is forced to do spend resources on software because no one else will. Everyone is motivated by money and AMD's grand plan won't work if they don't pay the price 10 years before everyone else. They are basically pushing adoptions and Mantle is meant to change the traditional way of doing things. Have anyone ever wonder how mobile graphics had exponential gains in the past decade? Does anyone remember AMD sold their mobile graphics to the ARM market?

I don't think AMD will ever dominate the market like Intel did because they have historically been horrible with marketing and lack any aggression in pushing their products. As far as business tactics goes, AMD is very tame compared to all their peers.

 

[TheGuruStud]

As far as business tactics goes, AMD is very tame compared to all their peers.

AKA they're not criminals

Pretty sad, really.

 

[ensabrenoir]

AKA they're not criminals

Pretty sad, really.

AKA they're not criminals Yet

Pretty sad, really.

Fixed that for ya watch what success does to ya

 

[eachus]

It's another question when a company keeps promising new stuff, but fails to deliver them on time in almost every case. Such companies mostly do not last as long as AMD

(BTW, I'm a software developer)

Then you must be familiar with the issue of version skew. AMD can test their (beta) drivers in house all they want, but game developers also have code waiting to test with the new (AMD) drivers. AMD can't do that testing in house, they don't have access to the unreleased code. (Yes, in this case the alpha testing almost certainly included the Frostbite developers, which is why they were able to release Mantle benchmarks.)

A company puts a beta out, first as a "restricted" beta. no big deal, just limiting the flood of reports if there is an easily found problem missed by in house testing.* Then comes an unrestricted beta, then a release candidate and so on. If there were no problems found (right!) all this would go smoothly. With more normal lluck beta and candidate release dates will bounce around like water on a hot skillet.

Is it possible to avoid this? Sure. Either don't have beta releases, or ignore all the bug reports--at least those that can't be fixed in the released version before the release date. We've all dealt with companies like that, and wish we hadn't. :-(

* It won't be missed next time, that's what regression test suites are for.

 

[Fiery]

Then you must be familiar with the issue of version skew. AMD can test their (beta) drivers in house all they want, but game developers also have code waiting to test with the new (AMD) drivers. AMD can't do that testing in house, they don't have access to the unreleased code. (Yes, in this case the alpha testing almost certainly included the Frostbite developers, which is why they were able to release Mantle benchmarks.)

A company puts a beta out, first as a "restricted" beta. no big deal, just limiting the flood of reports if there is an easily found problem missed by in house testing.* Then comes an unrestricted beta, then a release candidate and so on. If there were no problems found (right!) all this would go smoothly. With more normal lluck beta and candidate release dates will bounce around like water on a hot skillet.

Is it possible to avoid this? Sure. Either don't have beta releases, or ignore all the bug reports--at least those that can't be fixed in the released version before the release date. We've all dealt with companies like that, and wish we hadn't. :-(

* It won't be missed next time, that's what regression test suites are for.

Of course I'm familiar with the process. The problem with AMD is not the alpha and beta stages (those are fine and useful), but that they keep promising specific release schedules that they tend to keep missing. First they promise October, then December, then January. Then in January they keep saying it won't slip to February, when in fact it will be. Not smart to keep promising: it will make you look like an unprofessional company -- which AMD is not. After the first slip they should have stated "When it's done" or Q1 2014 or H1 2014...

 

[MarcusTaz]

With the Game bundled into the new higher end cards to push sales, they of course wanted to hype Mantle to help sales. For Dice I feel it was more hype to push a game that was released in an Alpha stage, more of a win win for both, but in the end it does nothing for the end user who struggles with a game that is completely unbalanced in many ways.

BF3 is still better by far in comparison.

 

[MxPhenom 216]

I guess its good they are delaying it so AMD users don't get a botched driver release.

 

[HumanSmoke]

AKA they're not criminals
Pretty sad, really.

Given the opportunity, they'll gladly take money out our wallets just like any other large corporation.


(And before anyone jumps in with the ATI isn't AMD excuse, bear in mind that 1. ATI's management remained largely intact after the AMD acquisition, and 2. the lawsuit covered the product line a full year under the AMD corporate ownership)

 

[fullinfusion]

I guess its good they are delaying it so AMD users don't get a botched driver release.

Yeah we sure dont want a snap crackel poP driver like what happened to the 590 lol

And W1zzard hows about a PM with the driver? I wont tell

 

[HumanSmoke]

Yeah we sure dont want a snap crackel poP driver like what happened to the 690 lol

I think the SKU you're attempting to refer to is the GTX 590....the one from nearly three years ago.

 

[fullinfusion]

I think the SKU you're attempting to refer to is the GTX 590....the one from nearly three years ago.

Was that the one W1zzard blew up?

 

[tokyoduong]

Given the opportunity, they'll gladly take money out our wallets just like any other large corporation.

That's the point of any business large or small. The real question is "are you getting your money's worth"

That's subjective and entirely your call.

 

[Crap Daddy]

The idea behind mantle is purely a business decision and from what I understood from these controlled benchmarks, it is a proprietary software tool aimed at increasing performance on their hardware, particularly on the CPU side which lags heavily behind Intel. Of course we are talking only about gaming and one game in particular. So, in essence, it is aimed to be a practical marketing tool to raise their tiny market share in the niche segment of CPUs used mainly for gaming. It will never ever threaten directx as an industry standard and it will survive just as long as devs will see a benefit in working overtime.