Late 6970?

[T2k]

One thing you guys need to remember it Nvidia is a profitable Company, AMD is not.

AMD is profitable since late last year and ATI part was always profitable.

They still operate under massive debt and in the red.

No, they are not. They are making a profit and some of it they spend on reducing debt.

If anyone is allowed to fail and still survive it would be Nvidia not AMD they have way more cash in hand!

No, it's not true. NV is dead within 24 months without constantly pumping money into R&D while AMD was operating with a loss for decades. You don't understand this, I think - it's a lot more difficult than your home budget, trust me.
BTW they just beat the expectations: http://www.google.com/hostednews/ap...iiS0Gg?docId=81c9a5b5ea7d4bd5964e84fb89418e48

Did I mention their profit margin is rising?

PS: obviously AMD had to hit targets time by time, turned profit time after time, they couldn't operate on a constant loss for 10+ years but I guess it goes without saying...

 

[Makaveli]

Reply to Aristotelian

you are miss reading my post which was to T2k original post.

I never said AMD was going to Die or that I want them to die.

all I said was if anyone was going to screw up Nvidia is in a greater position for recovery than AMD since they don't have any of AMD's debt!

soo umm keep your sorry!

 

[Makaveli]

No, it's not true. NV is dead within 24 months without constantly pumping money into R&D

Do you seriously believe that?

I read your link and yes AMD is making a profit but there competitor still has no debt and is right there with them.

Who do you think is in the better position currently?

 

[T2k]

That's a useless counterfactual. Aside from fanboys/zealots I think we can say with certainty that Nvidia and AMD are going to be around for the next while.

Not at all: Nvidia has no choice but to covert enough companies to CUDA or they will indeed lose their main revenue streams as Fusion-type APUs becoming commonplace in mid/low-end desktops and notebooks.
AMd just shipped its first Fusion APU yesterday, Intel is already in the market with a less integrated but essentially similar package with its i3/i5+Intel HD mobile solutions.
If there are no low/mid sales then there's no money to cover essential R&D costs which are astronomical and growing.

As he correctly said NV will never get an x86 license. They will also never get a chipset license again either and their mobile Tegra is a disaster so far.
They have no choice but bank on CUDA/GPGPU solutions and carve out their own niche market before it's too late, by 2015.

AMD has, last I heard, about 85% of the DX11 market. They are doing something right with graphics cards, at least. And there's nothing your posts can do about that, sorry.

 

[Genx87]

Do you seriously believe that?

The best part is he goes on to talk about how AMD operated for years without turning profit and burning through cash reserves. As if the crap hit the fan Nvidia couldnt do the same. Nvidia is still in a strong position for now with a nice cash reserve and churning a small loss each qtr. It will take years of this kind of performance to have to consider taking on debt to finance the business. By then the economy should turn and so with it, Nvidia's return to profitability.

 

[WelshBloke]

*sigh*

 

[Genx87]

Not at all: Nvidia has no choice but to covert enough companies to CUDA or they will indeed lose their main revenue streams as Fusion-type APUs becoming commonplace in mid/low-end desktops and notebooks.
AMd just shipped its first Fusion APU yesterday, Intel is already in the market with a less integrated but essentially similar package with its i3/i5+Intel HD mobile solutions.
If there are no low/mid sales then there's no money to cover essential R&D costs which are astronomical and growing.

As he correctly said NV will never get an x86 license. They will also never get a chipset license again either and their mobile Tegra is a disaster so far.
They have no choice but bank on CUDA/GPGPU solutions and carve out their own niche market before it's too late, by 2015.

What world do you live within? Nvidia would do just fine if those applications used Directcompute or OpenCL. Cuda is nice because it gives them an edge by being the first in the market with a viable GPGPU. But when the day comes devs and vendors want more hardware freedom because AMD pulled their head from their ass on GPGPU. Nvidia will still be there utilizing other technologies to feed their hardware.

 

[Aristotelian]

Reply to Aristotelian

1)you are miss reading my post which was to T2k original post.

2)I never said AMD was going to Die or that I want them to die.

3)all I said was if anyone was going to screw up Nvidia is in a greater position for recovery than AMD since they don't have any of AMD's debt!

soo umm keep your sorry!

Absolutely not. This:

What does it matter what you produce its either you make money and your business survives or you don't and eventually fail.

Nvidia as we know them now will never produce any X86 cpu's as intel will not grant them a license.

Was a direct response to this, from me:

When Nvidia starts producing CPUs then we can make that comparison regarding profitability. Apples to apples, right?

So (as I labelled your claims in your post):

1) I did not misread your post: it was a direct response to mine. T2K did not even mention CPU manufacturing in his post, I did, and you brought up manufacturing in your response. Please read posts (including your own) more carefully.

2) I never claimed that you said AMD or Nvidia were going to die. Do not straw man what I've argued.

3) You need to look up the definition of counterfactual. It is you who has not properly understood my post.

@T2K: I don't want to get into it here, but let's just say that I see Nvidia surviving for a while, even if the company has to penetrate some other, but related, markets.

 

[OCGuy]

The funny thing is that before the 5XX launch, at least there was some sort of "slyness" about it. Now it seems the cover is blown and it is pedal-to-them-metal. Haha.

 

[badb0y]

The funny thing is that before the 5XX launch, at least there was some sort of "slyness" about it. Now it seems the cover is blown and it is pedal-to-them-metal. Haha.

stop is all I am going to say next time I am going to report you.

 

[WelshBloke]

The funny thing is that before the 5XX launch, at least there was some sort of "slyness" about it. Now it seems the cover is blown and it is pedal-to-them-metal. Haha.

 

[T2k]

Do you seriously believe that?

Look, they don't have an CPU business to rely on... you have to understand their whole industry is turning around - when they lost their chipset business it was 1/3rd of their entire revenue which is pretty shocking, don't you think?

I know they are a totally unscrupulous bunch of people so I'm sure they quick to turn but it's CUDA or die right now.

I read your link and yes AMD is making a profit but there competitor still has no debt and is right there with them.

Right but that's a different scenario than having a competitor making its own CPUs (with GPU direct connected to the CPU) versus you are trying to come up with Optimus-like hackjobs to stay in the market where your bread & butter is (mobile/low/mid, that is), don't you agree?

Who do you think is in the better position currently?

On the long run? AMD, it's not a question for me especially watching how popular their 8-12 core Opterons are now thanks to the missing Intel levy on 4-way chips...
But stuff changes fast, I obviously know - say AMD screws up the APU development then NV will live on forever, pushing CUDA more and more mainstream. Of course, Intel will step in then but that's usually not good for AMD either...:biggrin:
If NV can convert enough people to CUDA then it will have a chance. If they cannot they will probably have to do some drastic changes eg open up or port their stack or I dont' know... one thing is sure they cannot finance their crazy expensive R&D without raking in lots of cash on zillions of low/mid cards. Just to give you an example when I order 20-30 T7500 from Dell I have to pick a VGA even though I will move my existing Quadros or buy new ones somewhere else - so which one I pick? The cheapest which is a sub-$200 Quadro NVS. Bang, 20-30 cards sold, some money made. But if they will come with CPUs sporting GPUs these cards are not sold anymore.

 

[OCGuy]

stop is all I am going to say next time I am going to report you.

Huh?

 

[WelshBloke]

Huh?

Indeed, mind you your post was a bit cryptic, its easy to read things into it that you probably didn't mean.

 

[Genx87]

On the long run? AMD, it's not a question for me especially watching how popular their 8-12 core Opterons are now thanks to the missing Intel levy on 4-way chips...
But stuff changes fast, I obviously know - say AMD screws up the APU development then NV will live on forever, pushing CUDA more and more mainstream. Of course, Intel will step in then but that's usually not good for AMD either...:biggrin:
If NV can convert enough people to CUDA then it will have a chance. If they cannot they will probably have to do some drastic changes eg open up or port their stack or I dont' know... one thing is sure they cannot finance their crazy expensive R&D without raking in lots of cash on zillions of low/mid cards. Just to give you an example when I order 20-30 T7500 from Dell I have to pick a VGA even though I will move my existing Quadros or buy new ones somewhere else - so which one I pick? The cheapest which is a sub-$200 Quadro NVS. Bang, 20-30 cards sold, some money made. But if they will come with CPUs sporting GPUs these cards are not sold anymore.

I just dont understand your fascination with CUDA. As if applications went to openCL or DirectCompute Nvidia would cease to exist. Why? What gives you the impression applications using DirectCompute or OpenCl wont work on Nvidia hardware?

 

[badb0y]

Huh?

Don't play stupid, you know damn well what you were inferring and I am not going stand by and make this your personal witch-hunt.

 

[chewietobbacca]

I just dont understand your fascination with CUDA. As if applications went to openCL or DirectCompute Nvidia would cease to exist. Why? What gives you the impression applications using DirectCompute or OpenCl wont work on Nvidia hardware?

CUDA's been out how long? And it still isn't the industry standard

 

[96Firebird]

The software guys where I work use CUDA, I hear them talking about it all the time. I have no idea how to even begin commenting on CUDA, but what is the industry standard right now?

 

[Genx87]

CUDA's been out how long? And it still isn't the industry standard

Ok?

 

[T2k]

CUDA's been out how long? And it still isn't the industry standard

Right on. We use CUDA because there wasn't anything to choose from when we decided to start developing our proprietary renderers etc for GPUs.

BTW if OpenCL or DC ever takes off that's even worse for NV because they run on *any* hardware... but it's a small market anyway, unless they can turn it into a mainstream one.

 

[Genx87]

Right on. We use CUDA because there wasn't anything to choose from when we decided to start developing our proprietary renderers etc for GPUs.

BTW if OpenCL or DC ever takes off that's even worse for NV because they run on *any* hardware...

Why? Nvidia will still have the performance crown until AMD decides to take GPGPU seriously.

 

[T2k]

The software guys where I work use CUDA, I hear them talking about it all the time. I have no idea how to even begin commenting on CUDA, but what is the industry standard right now?

There's no single "industry standard", there are at least 3 different one but CUDA is waaay ahead of the other two so it's far the most popular, obviously.
It's never about the actual standard but about the whole shebang: the tools, the support, the community etc - "developers, developers, developers!" :twisted:

 

[AnandThenMan]

I just dont understand your fascination with CUDA. As if applications went to openCL or DirectCompute Nvidia would cease to exist. Why? What gives you the impression applications using DirectCompute or OpenCl wont work on Nvidia hardware?

Why is Nvidia pushing CUDA so hard if they can just use DC or OpenCL instead? What you should have posted is, "I don't understand Nvidia's fascination with CUDA".

 

[busydude]

Why is Nvidia pushing CUDA so hard if they can just use DC or OpenCL instead?

To make some money?

 

[Genx87]

Why is Nvidia pushing CUDA so hard if they can just use DC or OpenCL instead? What you should have posted is, "I don't understand Nvidia's fascination with CUDA".

Do I really need to reply with an answer to this? Think about it for a second and get back to us.

That said, OpenCL and DirectCompute will be supported by Nvidia. So when\if Devs move to an OpenCL or DC Nvidia wont implode like T2K wants us to believe.