AMD: Doing The Work For Everyone

H

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
In response to NVIDIA's statement yesterday that the company doesn't intend to "do the work for everyone," AMD says that is exactly what it intends to do.

In our industry, one of the toughest decisions we continually face is how open we should be with our technology. On the one hand, developing cutting-edge graphics technology requires enormous investments. On the other hand, too much emphasis on keeping technologies proprietary can hinder broad adoption. It’s a dilemma we face practically every day, which is why we decided some time ago that those decisions would be guided by a basic principle: our goal is to support moving the industry forward as a whole, and that we’re proud to take a leadership position to help achieve that goal.
Click to expand...
 
I think I just heard steel nuts clink on NVidia's chin. :D

Marketing, true, but well played. LoL.
 
Way to go AMD. I hope this becomes standardized, do what Intel does and move the industry forward.
 
Using variable VBLANK to implement a variable refresh rate will require a particular way of doing it. It's fine that AMD will give away a particular solution which works with its own GPUs, but that doesn't really help any other manufacturers who could do it in any number of other specific ways (because other hardware controls it differently, there are better ways to do it, etc). This is what you get without industry standards, just another proprietary solution that masquerades as "open".

Dumb, hot-air PR. The rubes should eat it up.
 
{NG}Fidel said:
Way to go AMD. I hope this becomes standardized, do what Intel does and move the industry forward.
Click to expand...

I think you hit the nail on the head ... AMD has been competing with Intel much longer than NVidia and NVidia hasn't had to deal with wide ranging vertical product lines until recently with their expansion in mobile

If you want to be BIG or you want to support the widest range of industry product offerings then you have to be open and drive for industry standards ... if you have a niche that you are willing to live with and think you can protect it by being closed then you can try the NVidia approach ... it really comes down to how big an overall industry player you want to be

The other element is also the famous quote about the easiest way to predict the future is to create it ... Intel has been very good at that approach through their support of standards that benefit them (and opposition to ones that don't) ... the Intel/AMD approach is much safer unless you are protecting a very narrow and comfortable niche
 
bloodhawke83 said:
buuuurn. NVidia got pwned. :D
Click to expand...

No they didn't. Nvidia was right in this regard. If they really wanted an open ended graphics system they would have done it by now. Instead, from a strategic point of view they know that spending money on such a thing would not be a good investment. If someone else wants to throw the money at doing an open-source graphics card or a series of them, then by all means they can. Nvidia is saying they won't because it makes bad business sense and that isn't what they are seeing in the industry. All AMD did was oneupmanship with words. They neither have the money nor the investment to make this happen and they won't and if they do, it will fail. When you try to be everything to everybody, you end up being nothing to nobody..
 
If someone else wants to throw the money at doing an open-source graphics card or a series of them, then by all means they can. Nvidia is saying they won't because it makes bad business sense and that isn't what they are seeing in the industry. All AMD did was oneupmanship with words.
Click to expand...

WUT?
 
Nvidia isn't doing their customers any favors with their technology. Gsync and Physx are just some examples of how Nvidia is trying to lock customers in. It's great if you intend to always buy an Nvidia graphics card, but sucks if you plan to venture to AMD or even Intel.

But AMD can afford to go open, where Nvidia right now is having a shit. Because Nvidia can't make x86 CPUs, they can only make ARM based system on chips. Which is great if you intend to game on tablets and smart phones, but sucks on consoles or PCs. That's why Nvidia is betting on Android is the future gaming platform, even though I find the idea retarded. But in the ARM market, Nvidia isn't the strongest GPU either with Mali, Ardreno, and PowerVR.

Nvidia is feeling the push, and they're pushing back. It won't be long before AMD's APU's dominate the graphics market, along with Intel. Intel BTW is also pretty open about their tech. They do have a functional open source graphics driver that works really well on Linux. Nvidia has the strongest GPU in Linux, and yet they want to fight over the Android market?

AMD is going to make desktop ARM chips, which would be an interesting idea for Nvidia to do as well. I don't know about Android, but Linux in general is going to be a huge player in the market. Nvidia should forget about the tablet/smartphone market, as it's saturated. Apple uses their own Apple chips with PowerVR. Samsung uses their own Exynos chips or Qualcomm, which use Mali and Andreno. Qualcomm will never use anything else but Andreno, and Apple will always use PowerVR.

Nvidia should find a way to make X86 chips, or go big with desktop ARM chips. Either way they need to make open source drivers for Linux. No matter what, Linux is the future.
 
Who cares?

You want G-Sync? You can go out and buy it.

You want FreeSync? You can read about it on the AMD site.
 
I hope its true and AMD does intend to do the work for everyone. Especially allowing other GPU devs full access to Mantle.

I did chuckle at the disclamer on the bottom:

Disclaimer said:
Jay Lebo is a Product Marketing Manager at AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions.
Click to expand...
 
Methadras said:
No they didn't. Nvidia was right in this regard. If they really wanted an open ended graphics system they would have done it by now. Instead, from a strategic point of view they know that spending money on such a thing would not be a good investment. If someone else wants to throw the money at doing an open-source graphics card or a series of them, then by all means they can. Nvidia is saying they won't because it makes bad business sense and that isn't what they are seeing in the industry. All AMD did was oneupmanship with words. They neither have the money nor the investment to make this happen and they won't and if they do, it will fail. When you try to be everything to everybody, you end up being nothing to nobody..
Click to expand...

What money is needed? Scalar ASICs that support dynamic framerate switching are already available and cost next to nothing.
 
pxc said:
Using variable VBLANK to implement a variable refresh rate will require a particular way of doing it. It's fine that AMD will give away a particular solution which works with its own GPUs, but that doesn't really help any other manufacturers who could do it in any number of other specific ways (because other hardware controls it differently, there are better ways to do it, etc). This is what you get without industry standards, just another proprietary solution that masquerades as "open".

Dumb, hot-air PR. The rubes should eat it up.
Click to expand...

The only hot air statement is yours. You proclaim that no one but AMD will support this and that is not true AMD depends on certain industry standards to come through and then only then "free sync" can work for any vendor.
 
Ashbringer said:
AMD is going to make desktop ARM chips,
Click to expand...
[citation required]

AMD's ARM server chips are based on bog standard licensed Cortex A57 cores, the exact same A57 core that any other licensee has. Performance at target frequencies is equal to around a Yonah running at 1.6GHz at the A57's target frequency. Let's compute like it's 2006! AMD ARM server chips, coming in 2015! lol ARM hype is funny.
 
Ultra Wide said:
Who cares?

You want G-Sync? You can go out and buy it.

You want FreeSync? You can read about it on the AMD site.
Click to expand...

Well reading about it is cheaper then having to buy hardware to make something work most of us handle by either getting faster machines or overclocking them, if that fails turn down some graphics settings (absolutely free).
 
pcjunkie said:
Yes if Nvidia chooses to support it down the road.
Click to expand...
If it's useful to non-AMD architectures and whether AMD actually opens it up some time in the future are bigger concerns than that. ;) Mantle is still very proprietary in the proprietary sense.
 
lol @ people who don't understand the first thing about APIs, what they are, what they do or how they map functionality in a low level way in something like Mantle. Is this THG now?
 
What have they done? Mantle is proprietary. They have no current plans for "freesync". OpenCL was Apple but based off CUDA. Still no good 3D solution or comprehensive physics package. They are not providing any real competition to Intel or NVIDIA so high end chip prices are high.

AMD... Marketing BS for everyone.
 
pxc said:
lol @ people who don't understand the first thing about APIs, what they are, what they do or how they map functionality in a low level way in something like Mantle. Is this THG now?
Click to expand...

Maybe you should read up on the Mantle details yourself. Nobody said they would ever do it or it would be much benefit to other architectures, but yes they could if the want.ed. High horse, get off it...
 
Advil said:
^^^^^^ THIS +1,000,000 ^^^^^^

Mantle kind of flies in the face of what they just said.
Click to expand...

nVidia is welcome to use the technology. Of course we all know that nVidia will block it and screw their "customers" as has always been nVidia policy. We also all know that nVidia has fans, not customers, just like Apple. I'm still not fully sold on either of them yet, AMD's software has been so bad so many times, but at least a statement has been made to throw down the gauntlet, competition is an awesome thing.
 
Advil said:
^^^^^^ THIS +1,000,000 ^^^^^^

Mantle kind of flies in the face of what they just said.
Click to expand...

No it doesn't. AMD have already stated that they intend mantle to work on other graphics systems, minus the purpose built performance enhancements built directly into their chips and software to use it. They're opening it up, and if Nvidia wants to play ball, they can do their own software and hardware changes to take the most advantage of it like AMD is.
 
AMD's Mantle is currently not open-platform: mostly because it is still in-development. AMD has spared no effort in letting the world know that as soon as it's done, the standard will be open, and freely licensed, and the hardware optimizations that take advantage of the API can be implemented into any architecture. The main thing with GCN is that it is the 'test bed': the first graphics architecture to fully conform to Mantle standards. let me put it in a way that others cannot ignore.

GCN is conforming to Mantle standards. Mantle is not conforming to GCN standards. Mantle is not developed for GCN: GCN is simply the first platform to support the standard.

Does that clear it up? Nvidia can license Mantle, and hell, they can even implement hardware optimizations that give their architecture a 'boost' like GCN cards do.

But they won't. Not AMD's Fault.
 
I more willing to believe Nvidia will resurrect Glide than licensing Mantle.
 
I see pxc and PRIME1 have posted, this thread is officially worthless and should probably be closed.
 
It seems like very article I read with Nvidia in it makes me hate those greedy bastards more and more.
 
Not even relevant to Mantle. Driver support and monitor side scalar ASIC is all this is, is it not?
 
That really was about the best possible way to respond. Now let's hope they do more to back up those words with actions.
 
KazeoHin said:
GCN is conforming to Mantle standards. Mantle is not conforming to GCN standards. Mantle is not developed for GCN: GCN is simply the first platform to support the standard.

Does that clear it up?
Click to expand...
If your goal was to misinform, then you succeeded. :D

From TechReport's excellent coverage:

In Mantle's case, according to Riguer, AMD has lowered the abstraction level in some areas but "not across the board." DICE's Johan Andersson described the traditional approach as "middle-ground abstraction," where a compromise is struck between performance and usability. Mantle, by comparison, offers "thin low-level abstraction" that exposes how the underlying hardware works. Riguer boiled it down further by comparing Mantle to driving a car with a manual transmission—more responsibility, but also more fun.

Also, while Graphics Core Next is the "hardware foundation" for Mantle, AMD's Guennadi Riguer and some of the other Mantle luminaries at APU13 made it clear that the API is by no means tied down to GCN hardware. Some of Mantle's features are targeted at GCN, but others are generic. "We don't want to paint ourselves in a corner," Riguer explained. "What we would like to do with Mantle is to have [the] ability to innovate on future graphics architectures for years to come, and possibly even enable our competitors to run Mantle." Jurjen Katsman of Nixxes was even bolder in his assessment, stating, "There's nothing that I can see from my perspective that stops [Mantle] from running on pretty much any hardware out there that is somewhat recent."
Click to expand...
Proprietary and tied to GCN, at least according to people who know what they're talking about.

Does that clear it up?
 
"What we would like to do with Mantle is to have [the] ability to innovate on future graphics architectures for years to come, and possibly even enable our competitors to run Mantle."

"There's nothing that I can see from my perspective that stops [Mantle] from running on pretty much any hardware out there that is somewhat recent."
Click to expand...

Sounds pretty promising to me.
 
KazeoHin said:
GCN is conforming to Mantle standards. Mantle is not conforming to GCN standards. Mantle is not developed for GCN: GCN is simply the first platform to support the standard.
Click to expand...
That would make sense if Mantle came before GCN, which it didn't.
 
You must log in or register to reply here.
Back
Top