AMD Demonstrates "FreeSync", Free G-Sync Alternative at CES 2014

Status
Not open for further replies.
Venomous said:
It will still be a hell of a lot cheaper than gsync without it being proprietary is what I'm saying. I doubt it will ever come to market anyhow. Point is, Nvidia is just trying to lock people in.. When that happens they can continue to rape your wallet like they have been. You guys are the ones who will be crying the blues when you will be paying 30% more for your hardware to support the gsync standard which already carried a premium on top of it.

You = Sheep
Click to expand...

Some people are members of focus groups... they get free hardware and on another site they have to mention it in their signatures (after some embarrassing episodes occured)! It is nothing to them to suggest we part with out kidney and the like so they could prove their loyalty to free hardware providers.

I'm a user who pays for hardware with his own money (no free samples etc., though i wish :D), so i hold companies to a higher standard. Even if i were to get free hardware, i would still not, can not bring my self to recommend something i wouldn't myself do. I put my money where my mouth is, and unfortunately as was noted above, some people don't do/ have to.

I hope some remember the debacle that was DX10 for Nvidia... Even now their latest cards only are dx11 compliant, 11.1 is supported through software (or something like that, for lack of a better expression, emulation) but no DX11.2. For one, i wouldn't be surprised if Koduri is right on the money with Nvidia cards not having the basic support for a standard. It is not the first time in last few years that they didn't meet a standard. Say what people will about AMD, but heck even their last generation cards (HD7xxx GCN based) are also DX11.2.
 
wonderfield said:
Do you feel that Direct3D is similarly non-proprietary?
Click to expand...

no, direct3d is proprietary. its locked to MS operating systems.

Mantle is open for everyone to use on any OS. and you are free to bring Mantle to linux or what ever OS you want to try and get it working on AMD will not stop you from implementing it.
 
xoleras said:
It's hilarious that AMD is taking credit for Infineon's (now Qimonda) work on GDDR5. LOL.
Click to expand...

You do realize that AMD was the one who developed GDDR3 on which GDDR5 is based (and submitted to Jedec) right ? And the 1st to implement GDDR4 (X1950), and GDDR5 ..

Edit: If you think adding GDDR5 support is simply "slapping on GDDR5 modules".. I suggest you take a took at this -> HERE or how it took nearly 3 generations for nVidia to get fully operational GGDR5 controllers (GTX260-285 -> GTX580), causing a delay in Fermi's launch
 
Last edited:
cageymaru said:
AMD's response to proprietary hardware versus open standards in the computer industry.

Doing the Work for Everyone

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.

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. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
Click to expand...

Here is my "get back your mojo" plan for AMD:

1) Slash 90% of the marketing staff.
2) Double the engineering staff.

It's surprising to see how many corporate executives don't understand that #1 can't make up for a lack of #2.

As I read that blog entry I couldn't help but think of a parallel with AMD's Linux efforts. Basically, they don't put in the effort (or have the talent) to make a competitive video driver, so they decided to just open up the documentation and support the open source driver and let part of the community do the work for them. That has some benefits of course, but at the end of the day their drivers are abysmal.

I just can't buy into the "we love openness" marketing speak. They eschew open platforms like Linux and SteamOS, and then dance about closed platforms like XB1 and PS4. They have no interest in making competitive OpenGL drivers (hey, an open-standard API) but are quick to support DirectX (a closed, proprietary API). They are completely allergic to the open and ubiquitous Android, but make product announcements for the closed Windows tablets, which is a complete failure of a platform.

TressFX was on closed, proprietary DirectCompute... instead of the open-standards OpenCL.

Hey... it's business. I'm just saying that the marketing never seems to jive with the reality. This FreeSync business looks to me like they got caught with their pants down and are now trying a "me-too" alternative but are using the "open" spin to garner sympathy and make up for their laziness.


To those who insist Mantle is "open" and NVIDIA is free to go ahead and implement it: Say I'm a GPU designer for XYZ Corp. How do I get this "open" API and begin implementing support for it? Where is it? If it's open then every person should be able to go ahead and download the specs, the API, everything. Where is it?
 
Wildace said:
Mantle is open for everyone to use on any OS. and you are free to bring Mantle to linux or what ever OS you want to try and get it working on AMD will not stop you from implementing it.
Click to expand...

You are very much misinformed.
 
xoleras said:
I just told you. Read today's story at PCPer. They just posted it a few hours ago.
"To be clear, just because a monitor would run with DisplayPort 1.3 doesn't guarantee this feature would work. It also requires the controller on the display to understand and be compatible with the variable refresh portions of the spec, which with eDP 1.0 at least, isn't required. AMD is hoping that with the awareness they are building with stories like this display designers will actually increase the speed of DP 1.3 adoption and include support for variable refresh rate with them. That would mean an ecosystem of monitors that could potentially support variable speed refresh on both AMD and NVIDIA cards."
Click to expand...
That says that the controller needs to be compatible with the optional spec not that it needs a separate controller. I also don't understand how you wonder why AMD is demonstrating it at this point after reading that statement, it was obvious to me before reading the quote but that spells it out.

xoleras said:
Also, it should be noted, that THIS VERY WEBSITE complained of this issue as well on 79xx cards.... It is fixed on the 290/X but not on prior cards. CF/eyefinity or DX9 / CF has microstutter on 79xx and older cards. And tearing.

But we'll take your word for it over HardOCPs. Especially since AMD admitted to the issues.
Click to expand...

You do realize that frame pacing was an issue on SLI setups until shortly before Nvidia made a big stink about this right? Stuttering is one of the reasons I swore off multi gpu setups so I'm glad that they worked on it and pushed AMD to do the same but the way they went about it was entirely self serving and a little bit slimy, not that I'm surprised as that's the Nvidia way.
 
jwcalla said:
Here is my "get back your mojo" plan for AMD:

1) Slash 90% of the marketing staff.
2) Double the engineering staff.

It's surprising to see how many corporate executives don't understand that #1 can't make up for a lack of #2.

As I read that blog entry I couldn't help but think of a parallel with AMD's Linux efforts. Basically, they don't put in the effort (or have the talent) to make a competitive video driver, so they decided to just open up the documentation and support the open source driver and let part of the community do the work for them. That has some benefits of course, but at the end of the day their drivers are abysmal.

I just can't buy into the "we love openness" marketing speak. They eschew open platforms like Linux and SteamOS, and then dance about closed platforms like XB1 and PS4. They have no interest in making competitive OpenGL drivers (hey, an open-standard API) but are quick to support DirectX (a closed, proprietary API). They are completely allergic to the open and ubiquitous Android, but make product announcements for the closed Windows tablets, which is a complete failure of a platform.

TressFX was on closed, proprietary DirectCompute... instead of the open-standards OpenCL.

Hey... it's business. I'm just saying that the marketing never seems to jive with the reality. This FreeSync business looks to me like they got caught with their pants down and are now trying a "me-too" alternative but are using the "open" spin to garner sympathy and make up for their laziness.


To those who insist Mantle is "open" and NVIDIA is free to go ahead and implement it: Say I'm a GPU designer for XYZ Corp. How do I get this "open" API and begin implementing support for it? Where is it? If it's open then every person should be able to go ahead and download the specs, the API, everything. Where is it?
Click to expand...

DirectCompute is open. I don't work for AMD nor consult with the writer of the blog. Take your issues up with him as he has a comments section at the bottom of the blog. I saw an interesting article and linked it so that others could read, comprehend, and make their own opinions up about it. Don't blame the messenger. :)
 
Funny how my post got glossed over and everyone just started arguing over proprietary vs. open source standards, and bashing each other like cavemen but over the internet. :p
 
Last edited:
Venomous said:
290x vs 770? I closed that shit the second I saw that.
Click to expand...

It's an article about how Nvidia writes proprietary CUDA to OpenCL wrappers for their sponsored games. AMD isn't allowed to see the coding libraries for those games. Thus they also aren't allowed to optimize drivers for those games. So a GTX 770 is as fast as a R9 290x in those titles.

Now does it make sense? If not, what I'm trying to get you to think about is that if Nvidia is willing to write shit drivers to intentionally make AMD look bad in their sponsored games, then why would I trust them with incorporating GSync tech into every monitor. I'm going to sleep so I'll let you'll debate the implications.

And Nvidia admits to blocking AMD from optimizing in the article near the top.
 
octoberasian said:
Funny how my post got glossed over and everyone just started arguing over proprietary vs. open source standards, and bashing each other like cavemen but over the internet. :p
Click to expand...

I think your post disappeared man. I looked through the past 3 pages multiple times and it doesn't exist in my timeline. :)
 
cageymaru said:
It's an article about how Nvidia writes proprietary CUDA to OpenCL wrappers for their sponsored games. AMD isn't allowed to see the coding libraries for those games. Thus they also aren't allowed to optimize drivers for those games. So a GTX 770 is as fast as a R9 290x in those titles.

Now does it make sense? If not, what I'm trying to get you to think about is that if Nvidia is willing to write shit drivers to intentionally make AMD look bad in their sponsored games, then why would I trust them with incorporating GSync tech into every monitor. I'm going to sleep so I'll let you'll debate the implications.

And Nvidia admits to blocking AMD from optimizing in the article near the top.
Click to expand...

Yeah I figured that out and made a mention of it in my reply. Still would of liked to seen their top tier cards tested to see if there was a much wider gap on the high end.
 
Wildace said:
no, direct3d is proprietary. its locked to MS operating systems.
Click to expand...
Whether it runs on multiple operating systems or not doesn't change that it's a Microsoft property, and one Microsoft retains sole ownership of. That is what a proprietary technology is.

AMD is not giving away intellectual property, and they aren't submitting Mantle to national or international standards bodies. They're merely allowing, allegedly, other vendors to provide conforming implementations.
 
wonderfield said:
Whether it runs on multiple operating systems or not doesn't change that it's a Microsoft property, and one Microsoft retains sole ownership of. That is what a proprietary technology is.

AMD is not giving away intellectual property, and they aren't submitting Mantle to national or international standards bodies. They're merely allowing, allegedly, other vendors to provide conforming implementations.
Click to expand...
So mantel is property, so is freesync and so is android so is etc...
 
One company that does a lot of free work that others reap benefit out of is Intel. Look at what they did with USB 2.0, 3.0, Firewire and other open standards.

At the same time they can be nasty towards miniatures like AMD and NVidia. But credit should be given where it's due.
 
madgun said:
One company that does a lot of free work that others reap benefit out of is Intel. Look at what they did with USB 2.0, 3.0, Firewire and other open standards.

At the same time they can be nasty towards miniatures like AMD and NVidia. But credit should be given where it's due.
Click to expand...

When Intel allows open licensing of x86 then we can give them credit.

I'm not criticizing their business decision, but they're no different than anyone else in the industry.
 
Unknown-One said:
Right... so what's your point?

AMD, Nvidia, Intel, Matrox, and S3 all have GPUs that support DirectCompute, so it works everywhere.
Click to expand...

And by everywhere you mean Windows, right?

DirectX isn't an open standard, its MS's standard. OpenGL/OpenCL are open standards.
 
Unknown-One said:
Right... so what's your point?

AMD, Nvidia, Intel, Matrox, and S3 all have GPUs that support DirectCompute, so it works everywhere.
Click to expand...

CUDA works on Linux, OSX and Windows, so it works everywhere. Truly an open standard.
 
PRIME1 said:
CUDA works on Linux, OSX and Windows, so it works everywhere. Truly an open standard.
Click to expand...

CUDA does not work on OSX, NVIDIA broke their driver and never fixed it.

Typical NVIDIA, releasing broken unstable drivers and never fixing them.
 
Spazturtle said:
CUDA does not work on OSX, NVIDIA broke their driver and never fixed it.

Typical NVIDIA, releasing broken unstable drivers and never fixing them.
Click to expand...

Not to throw one in your face, but NVIDIA does not write or release drivers for OSX. Apple does. This is required by Apple so they can be in control of their ecosystem.

It's also possible Apple did it intentionally or doesn't care because they prefer OpenCL anyway so they can vendor balance between AMD and NVIDIA (After all, they created OpenCL in the image of CUDA for this purpose).
 
bexamous said:
DirectX isn't an open standard
Click to expand...
We were talking about DirectCompute, not DirectX.

The first version of DirectCompute shipped with the DirectX 11 suite, but it has since been decoupled. Any hardware that supports Shader Model 4.0 (or higher) can run DirectCompute, regardless of DirectX.

Just like OpenCL (the compute library) has nothing to do with OpenGL (the graphics library).
 
madgun said:
One company that does a lot of free work that others reap benefit out of is Intel. Look at what they did with USB 2.0, 3.0, Firewire and other open standards.

At the same time they can be nasty towards miniatures like AMD and NVidia. But credit should be given where it's due.
Click to expand...

How many companies produced chipsets for socket 775 ?
How many companies produce chipsets for 1150 or 2011 ?
 
Unknown-One said:
We were talking about DirectCompute, not DirectX.

The first version of DirectCompute shipped with DirectX 11, but it has since been decoupled. Any hardware that supports Shader Model 4.0 (or higher) can run DirectCompute, regardless of DirectX.
Click to expand...

You're forgetting to tell me how its a open standard, the current topic.
 
That is not an open standard.

Here is someone from MS defining open standard:
Let's look at what an open standard means: 'open' refers to it being royalty-free, while 'standard' means a technology approved by formalized committees that are open to participation by all interested parties and operate on a consensus basis. An open standard is publicly available, and developed, approved and maintained via a collaborative and consensus driven process.
 
bexamous said:
That is not an open standard.

Here is someone from MS defining open standard:
Let's look at what an open standard means: 'open' refers to it being royalty-free, while 'standard' means a technology approved by formalized committees that are open to participation by all interested parties and operate on a consensus basis. An open standard is publicly available, and developed, approved and maintained via a collaborative and consensus driven process.
Click to expand...

Implementing it yourself can be done freely.

What standard's committee is open to all interested parties? I can't think of any.
 
Even if Mantle qualified as open(Mantle is a proprietary attempt to derive a competitive performance advantage from getting developers to program for AMD hardware specifically. Doesn't get any less open than that.) it wouldn't matter. What's the point in having an "open" low-level hardware API for a proprietary hardware architecture? It's not like anyone can just go ahead and reimplement that on whatever hardware. That's not how low-level hardware programming works.
 
jimmyb said:
What standard's committee is open to all interested parties? I can't think of any.
Click to expand...
ISO/IEC JTC1/SC22/WG21, to name one. No membership is required to submit proposals for acceptance in future standards.
 
jimmyb said:
Implementing it yourself can be done freely.

What standard's committee is open to all interested parties? I can't think of any.
Click to expand...

Why are you still trying to argue this?

DirectCompute is MS's standard, therei s no group or commetee behind it. MS can do whatever they want. That is not an open standard. This is way too stupid to argue.

/me done
 
bexamous said:
Why are you still trying to argue this?

DirectCompute is MS's standard, therei s no group or commetee behind it. MS can do whatever they want. That is not an open standard. This is way too stupid to argue.

/me done
Click to expand...
That's his target, things like mantel etc are not open standards by that detention as you pointed out you said if only one person control it, it's not open standard.
 
Michaelius said:
How many companies produced chipsets for socket 775 ?
How many companies produce chipsets for 1150 or 2011 ?
Click to expand...

Umm but everyone including AMD use USB 3.0 in their motherboard.

Just check the USB 3.0 spec sheet and see which company contributed the most to the standard.
 
Marcdaddy said:
Will FREESYNC be out before MANTLE, I'm so confused!
Click to expand...

Well if you listen to some here both are never coming and when they do they'll be horrible.

The most relevant points of the discussion are:
A. Whether something like this can be beneficial(I think there is at least a consensus here)
B. Which is the better approach from a performance standpoint(lag, repeated frames, etc.)
C. Which is most cost effective to implement, making it more likely to be used in a larger number of monitors.

Unfortunately until more is known about both technologies(especially freesync) points b and c are based more off of speculation than actual information.

This whole discussion about open standards vs proprietary has gone way past where it needed to. One company is pushing trying to help develop a universal standard that everyone can use and the other is trying to make something that requires their gpu and special hardware in the monitor that is made or licensed by them in order to function, you can't argue that the former is at least more of an open standard than the latter. It's also easy to tell which route both of those companies tend to take if you have been paying attention the last 5-10 years, personally I think that industry standards that allow for compatibility across brands is a big part of what makes PCs great.
 
Status
Not open for further replies.
Back
Top