AMD Fusion

[jones377]

Originally posted by: Viditor

Originally posted by: jones377
I should add that it has also been reported that Intel is planning to integrate a GPU in some of their Nehalem line of products. If Intel again uses the MCM approach it could be possible that they again beat AMD to market. Fusion is planned for 2009 while Nehalem is coming out in 2H08. It is not entirely impossible that Nehalem+GPU comes out before Fusion but it is too early to say anything for certain.

Well, a few comments...
1. I believe the model AMD has for Fusion is to put several GPUs on each chip, so high-end may be closer than you think...I don't know, but it certainly is possible.

2. I personally don't think Intel is going to introduce graphics OR MCMs to Nehalem in it's first year...it's a radically different architecture compared to Core, and as you said, "before you learn to walk you need to learn to crawl".

3. While Intel is the largest graphics manufacturer (from their on-board graphics), they have never had success with discrete high-end graphics. Their only real stab at it was a dismal failure...and while I have heard of their new upcoming attempt, I can't say I'm very optimistic.
The same can't be said for ATI/AMD however, and while your point on integration is well taken, it's apparent that discussion of this project began well before the buyout occured.

1. It simply isn't possible to get the kind of bandwidth a high-end GPU needs into a single detachable socket processor. Not to mention the power requirement of running both a high-end CPU and GPU under the same cooler. It would have to be soldered right to the motherboard and have a 512-bit memory controller with the RAM soldered to the board as well. That is the only way to get the high frequencies the graphics RAM operates at. GPUs need huge amounts of bandwidth and that won't change until the programming model for graphics changes to something radically different (like raytracing possibly, I believe it is much more friendly to caches)

2. I never said we should expect Nehalem+GPU in 2008. Just that there may be sufficient time between the release of the first Nehalem (probably a straight quadcore with integrated memory controller for the high-end desktop) and the release of Fusion, for Intel to come out with their own integrated CPU-GPU device. Even if it comes out around the same time it could still be considered a win for Intel considering how much AMD has trumpeted it's Fusion project. Intel could come in totally under the radar here (although I suspect we'll hear more about it once it gets closer to release.)

3. We'll see how Larrabee pans out. I have my doubts as well.

 

[firewolfsm]

Fusion focuses on notebooks, because most CPU shipments are from that market and it's the fastest growing market. It will simply move the discrete gpu die from where it is now (it sits on most laptop motherboards like a northbridge chip), to the cpu. This reduces the size of the motherboard and the heatsink and improves efficiency a little. After that, AMD says they will move that gpu die under the L3 cache and have the CPU and GPU share logic. This is where Fusion starts to look better as power consumption will go down and performance won't be affected. I think it will only be dual core + GPU until 32nm (for laptops) but there's no way to be sure.

Bulldozer is a very different project focusing on highend servers and desktops, it's a lot like Terascale only with fewer, more powerful cores. It could have any kind of dedicated processor put on the die as it's very modular, cores, gpus, etc, can all be added under the L3 cache.

 

[Phynaz]

Originally posted by: Viditor

Originally posted by: jones377
I should add that it has also been reported that Intel is planning to integrate a GPU in some of their Nehalem line of products. If Intel again uses the MCM approach it could be possible that they again beat AMD to market. Fusion is planned for 2009 while Nehalem is coming out in 2H08. It is not entirely impossible that Nehalem+GPU comes out before Fusion but it is too early to say anything for certain.

Well, a few comments...
1. I believe the model AMD has for Fusion is to put several GPUs on each chip, so high-end may be closer than you think...I don't know, but it certainly is possible.

2. I personally don't think Intel is going to introduce graphics OR MCMs to Nehalem in it's first year...it's a radically different architecture compared to Core, and as you said, "before you learn to walk you need to learn to crawl".

3. While Intel is the largest graphics manufacturer (from their on-board graphics), they have never had success with discrete high-end graphics. Their only real stab at it was a dismal failure...and while I have heard of their new upcoming attempt, I can't say I'm very optimistic.
The same can't be said for ATI/AMD however, and while your point on integration is well taken, it's apparent that discussion of this project began well before the buyout occured.

"It's not going to replace high-end discrete graphics chips coveted by gamers, and it's not going to deliver the ultimate in CPU performance."

Phil Hester, at the 2007 hot chips conference.

Funsion is low-end, mass market stuff.
BTW, you are aware that fusion is just an idea, and hasn't been designed yet, right?

 

[Nemesis 1]

Originally posted by: Viditor

Originally posted by: jones377
I should add that it has also been reported that Intel is planning to integrate a GPU in some of their Nehalem line of products. If Intel again uses the MCM approach it could be possible that they again beat AMD to market. Fusion is planned for 2009 while Nehalem is coming out in 2H08. It is not entirely impossible that Nehalem+GPU comes out before Fusion but it is too early to say anything for certain.

Well, a few comments...
1. I believe the model AMD has for Fusion is to put several GPUs on each chip, so high-end may be closer than you think...I don't know, but it certainly is possible.

2. I personally don't think Intel is going to introduce graphics OR MCMs to Nehalem in it's first year...it's a radically different architecture compared to Core, and as you said, "before you learn to walk you need to learn to crawl".

3. While Intel is the largest graphics manufacturer (from their on-board graphics), they have never had success with discrete high-end graphics. Their only real stab at it was a dismal failure...and while I have heard of their new upcoming attempt, I can't say I'm very optimistic.
The same can't be said for ATI/AMD however, and while your point on integration is well taken, it's apparent that discussion of this project began well before the buyout occured.

Really good stuff here. Since I don't know alot about AMD I need to ask a few questions.

1) Viditor on the Z ram thing . Z-ram works only with SOI is that true. So your saying that AMD on 45nm will be SOI.
2) Back on subject . aren't ATI GPU on Bulk and AMD is on soi . So how will the 2 play nice together.

 

[Phynaz]

Originally posted by: Nemesis 1

Originally posted by: Viditor

Originally posted by: jones377
I should add that it has also been reported that Intel is planning to integrate a GPU in some of their Nehalem line of products. If Intel again uses the MCM approach it could be possible that they again beat AMD to market. Fusion is planned for 2009 while Nehalem is coming out in 2H08. It is not entirely impossible that Nehalem+GPU comes out before Fusion but it is too early to say anything for certain.

Well, a few comments...
1. I believe the model AMD has for Fusion is to put several GPUs on each chip, so high-end may be closer than you think...I don't know, but it certainly is possible.

2. I personally don't think Intel is going to introduce graphics OR MCMs to Nehalem in it's first year...it's a radically different architecture compared to Core, and as you said, "before you learn to walk you need to learn to crawl".

3. While Intel is the largest graphics manufacturer (from their on-board graphics), they have never had success with discrete high-end graphics. Their only real stab at it was a dismal failure...and while I have heard of their new upcoming attempt, I can't say I'm very optimistic.
The same can't be said for ATI/AMD however, and while your point on integration is well taken, it's apparent that discussion of this project began well before the buyout occured.

Really good stuff here. Since I don't know alot about AMD I need to ask a few questions.

1) Viditor on the Z ram thing . Z-ram works only with SOI is that true. So your saying that AMD on 45nm will be SOI.
2) Back on subject . aren't ATI GPU on Bulk and AMD is on soi . So how will the 2 play nice together.

Let me jump in with the info that there is a report that AMD is studying moving to bulk and eliminating SOI.

 

[Gikaseixas]

nice topic/tread

a friend told me that the cpu/gpu combo will enable AMD system to run much faster than the competition while saving more energy.

 

[Inspector Jihad]

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: jones377
I should add that it has also been reported that Intel is planning to integrate a GPU in some of their Nehalem line of products. If Intel again uses the MCM approach it could be possible that they again beat AMD to market. Fusion is planned for 2009 while Nehalem is coming out in 2H08. It is not entirely impossible that Nehalem+GPU comes out before Fusion but it is too early to say anything for certain.

Well, a few comments...
1. I believe the model AMD has for Fusion is to put several GPUs on each chip, so high-end may be closer than you think...I don't know, but it certainly is possible.

2. I personally don't think Intel is going to introduce graphics OR MCMs to Nehalem in it's first year...it's a radically different architecture compared to Core, and as you said, "before you learn to walk you need to learn to crawl".

3. While Intel is the largest graphics manufacturer (from their on-board graphics), they have never had success with discrete high-end graphics. Their only real stab at it was a dismal failure...and while I have heard of their new upcoming attempt, I can't say I'm very optimistic.
The same can't be said for ATI/AMD however, and while your point on integration is well taken, it's apparent that discussion of this project began well before the buyout occured.

"It's not going to replace high-end discrete graphics chips coveted by gamers, and it's not going to deliver the ultimate in CPU performance."

Phil Hester, at the 2007 hot chips conference.

Funsion is low-end, mass market stuff.
BTW, you are aware that fusion is just an idea, and hasn't been designed yet, right?

this makes the most sense. i can see how its very attractive in the mobile computer scene.

 

[Viditor]

Originally posted by: jones377

Originally posted by: Viditor

Originally posted by: jones377
I should add that it has also been reported that Intel is planning to integrate a GPU in some of their Nehalem line of products. If Intel again uses the MCM approach it could be possible that they again beat AMD to market. Fusion is planned for 2009 while Nehalem is coming out in 2H08. It is not entirely impossible that Nehalem+GPU comes out before Fusion but it is too early to say anything for certain.

Well, a few comments...
1. I believe the model AMD has for Fusion is to put several GPUs on each chip, so high-end may be closer than you think...I don't know, but it certainly is possible.

2. I personally don't think Intel is going to introduce graphics OR MCMs to Nehalem in it's first year...it's a radically different architecture compared to Core, and as you said, "before you learn to walk you need to learn to crawl".

3. While Intel is the largest graphics manufacturer (from their on-board graphics), they have never had success with discrete high-end graphics. Their only real stab at it was a dismal failure...and while I have heard of their new upcoming attempt, I can't say I'm very optimistic.
The same can't be said for ATI/AMD however, and while your point on integration is well taken, it's apparent that discussion of this project began well before the buyout occured.

1. It simply isn't possible to get the kind of bandwidth a high-end GPU needs into a single detachable socket processor. Not to mention the power requirement of running both a high-end CPU and GPU under the same cooler. It would have to be soldered right to the motherboard and have a 512-bit memory controller with the RAM soldered to the board as well. That is the only way to get the high frequencies the graphics RAM operates at. GPUs need huge amounts of bandwidth and that won't change until the programming model for graphics changes to something radically different (like raytracing possibly, I believe it is much more friendly to caches)

Based on current GPU models, I would agree. However when Fusion is released, that may not be the case.
Firstly, the bandwidth when on-die doesn't need to be near as high because of the much lower latency (the GPU doesn't have to go through the PCIx bus and is right there).
Secondly, if AMD does use Z-Ram L3, they can have HUGE quantities of cache (thus reducing the bandwidth requirements even more).
Third, with multiple GPU cores operating on bare-metal coding, the bandwidth requirements are reduced even further...

2. I never said we should expect Nehalem+GPU in 2008. Just that there may be sufficient time between the release of the first Nehalem (probably a straight quadcore with integrated memory controller for the high-end desktop) and the release of Fusion, for Intel to come out with their own integrated CPU-GPU device. Even if it comes out around the same time it could still be considered a win for Intel considering how much AMD has trumpeted it's Fusion project. Intel could come in totally under the radar here (although I suspect we'll hear more about it once it gets closer to release.)

I understand...but my understanding of the 2 roadmaps is that Nehalem is to be first released at the end of 2008, and Fusion at the end of 2009. So I can buy a simultaneous release (though I tend to doubt that given Intel's experience and the fact that AMD/ATI have been working on the project far longer), but certainly not Intel releasing first (at least not anything decent).

3. We'll see how Larrabee pans out. I have my doubts as well.

 

[Viditor]

Originally posted by: Phynaz

"It's not going to replace high-end discrete graphics chips coveted by gamers, and it's not going to deliver the ultimate in CPU performance."

Phil Hester, at the 2007 hot chips conference.
Funsion is low-end, mass market stuff.

As I said, the first iteration certainly is not...it's the ?accelerated computing? versions that might (please note that I used the word "might" there...).

BTW, you are aware that fusion is just an idea, and hasn't been designed yet, right?

Huh? If you mean it hasn't been manufactured yet, then you're right. Of course the same is true for Nehalem...
If you mean that they don't have sample reticules (thanx to IDon'tCare for the term...), then you are incorrect. In fact, if you scroll back up in this thread, you'll find a die shot of one...

 

[Viditor]

Originally posted by: Nemesis 1

Really good stuff here. Since I don't know alot about AMD I need to ask a few questions.

1) Viditor on the Z ram thing . Z-ram works only with SOI is that true. So your saying that AMD on 45nm will be SOI.
2) Back on subject . aren't ATI GPU on Bulk and AMD is on soi . So how will the 2 play nice together.

1. Yes, Z-Ram is SOI only...as to whether they will keep using it, I don't think even AMD engineers can answer that yet. I do know that the 45nm process that AMD and IBM developed together at East Fishkill and presented the paper on in Dec used SOI with HKMG (High-K Metal Gates).

2. You need to understand that Fusion will be the very first time that they will be combining ATI designs with AMD manufacturing. This will require a ground-up design anyway, so what they use now makes very little difference...

 

[Viditor]

Originally posted by: Gikaseixas
nice topic/tread

a friend told me that the cpu/gpu combo will enable AMD system to run much faster than the competition while saving more energy.

While most of the rest of what we are talking about is speculative, what you're friend told you is pretty much in the bank. In fact, it's probably the only thing we can be sure of in all of this at this point...

 

[Phynaz]

Originally posted by: Viditor

Originally posted by: Phynaz

"It's not going to replace high-end discrete graphics chips coveted by gamers, and it's not going to deliver the ultimate in CPU performance."

Phil Hester, at the 2007 hot chips conference.
Funsion is low-end, mass market stuff.

As I said, the first iteration certainly is not...it's the ?accelerated computing? versions that might (please note that I used the word "might" there...).

BTW, you are aware that fusion is just an idea, and hasn't been designed yet, right?

Huh? If you mean it hasn't been manufactured yet, then you're right. Of course the same is true for Nehalem...
If you mean that they don't have sample reticules (thanx to IDon'tCare for the term...), then you are incorrect. In fact, if you scroll back up in this thread, you'll find a die shot of one...

That's exactly what I'm saying. Fusion hasn't been designed yet, so there has never been a test mask made.

According to a interview on 7/16/2007 with Giuseppe Amato, Fusion is still at the idea stage.
They don't even know how they will build it yet.

 

[Nemesis 1]

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: Viditor

Originally posted by: jones377
I should add that it has also been reported that Intel is planning to integrate a GPU in some of their Nehalem line of products. If Intel again uses the MCM approach it could be possible that they again beat AMD to market. Fusion is planned for 2009 while Nehalem is coming out in 2H08. It is not entirely impossible that Nehalem+GPU comes out before Fusion but it is too early to say anything for certain.

Well, a few comments...
1. I believe the model AMD has for Fusion is to put several GPUs on each chip, so high-end may be closer than you think...I don't know, but it certainly is possible.

2. I personally don't think Intel is going to introduce graphics OR MCMs to Nehalem in it's first year...it's a radically different architecture compared to Core, and as you said, "before you learn to walk you need to learn to crawl".

3. While Intel is the largest graphics manufacturer (from their on-board graphics), they have never had success with discrete high-end graphics. Their only real stab at it was a dismal failure...and while I have heard of their new upcoming attempt, I can't say I'm very optimistic.
The same can't be said for ATI/AMD however, and while your point on integration is well taken, it's apparent that discussion of this project began well before the buyout occured.

1. It simply isn't possible to get the kind of bandwidth a high-end GPU needs into a single detachable socket processor. Not to mention the power requirement of running both a high-end CPU and GPU under the same cooler. It would have to be soldered right to the motherboard and have a 512-bit memory controller with the RAM soldered to the board as well. That is the only way to get the high frequencies the graphics RAM operates at. GPUs need huge amounts of bandwidth and that won't change until the programming model for graphics changes to something radically different (like raytracing possibly, I believe it is much more friendly to caches)

Based on current GPU models, I would agree. However when Fusion is released, that may not be the case.
Firstly, the bandwidth when on-die doesn't need to be near as high because of the much lower latency (the GPU doesn't have to go through the PCIx bus and is right there).
Secondly, if AMD does use Z-Ram L3, they can have HUGE quantities of cache (thus reducing the bandwidth requirements even more).
Third, with multiple GPU cores operating on bare-metal coding, the bandwidth requirements are reduced even further...

2. I never said we should expect Nehalem+GPU in 2008. Just that there may be sufficient time between the release of the first Nehalem (probably a straight quadcore with integrated memory controller for the high-end desktop) and the release of Fusion, for Intel to come out with their own integrated CPU-GPU device. Even if it comes out around the same time it could still be considered a win for Intel considering how much AMD has trumpeted it's Fusion project. Intel could come in totally under the radar here (although I suspect we'll hear more about it once it gets closer to release.)

I understand...but my understanding of the 2 roadmaps is that Nehalem is to be first released at the end of 2008, and Fusion at the end of 2009. So I can buy a simultaneous release (though I tend to doubt that given Intel's experience and the fact that AMD/ATI have been working on the project far longer), but certainly not Intel releasing first (at least not anything decent).

3. We'll see how Larrabee pans out. I have my doubts as well.

How sure are you that Fusion has been worked on longer than Larrabee? Or are you confusing the name with other projects.

http://blogs.intel.com/technol...e_instilling_ener.html

Noyice the blogger states larrobee is the first terrascale project to be released. There is more info on this. Many have said the terrascale project is the reason AMD bought ATI. Just a little food for thought. Off course if you could show me a link that fusion was before terrascale I would be indebted to you.

 

[Viditor]

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

"It's not going to replace high-end discrete graphics chips coveted by gamers, and it's not going to deliver the ultimate in CPU performance."

Phil Hester, at the 2007 hot chips conference.
Funsion is low-end, mass market stuff.

As I said, the first iteration certainly is not...it's the ?accelerated computing? versions that might (please note that I used the word "might" there...).

BTW, you are aware that fusion is just an idea, and hasn't been designed yet, right?

Huh? If you mean it hasn't been manufactured yet, then you're right. Of course the same is true for Nehalem...
If you mean that they don't have sample reticules (thanx to IDon'tCare for the term...), then you are incorrect. In fact, if you scroll back up in this thread, you'll find a die shot of one...

That's exactly what I'm saying. Fusion hasn't been designed yet, so there has never been a test mask made.

According to a interview on 7/16/2007 with Giuseppe Amato, Fusion is still at the idea stage.
They don't even know how they will build it yet.

Well then he either misspoke (what he may have meant to say is design stage which would last right up until tape-out) or is wrong...we've already seen a die shot of a test chip. Besides, if it's coming out anywhere near 2009, it HAS to be in the design stage.

However, I would very much like to read the interview...could you link it please?

Cheers...

 

[Phynaz]

Originally posted by: Viditor

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

"It's not going to replace high-end discrete graphics chips coveted by gamers, and it's not going to deliver the ultimate in CPU performance."

Phil Hester, at the 2007 hot chips conference.
Funsion is low-end, mass market stuff.

As I said, the first iteration certainly is not...it's the ?accelerated computing? versions that might (please note that I used the word "might" there...).

BTW, you are aware that fusion is just an idea, and hasn't been designed yet, right?

Huh? If you mean it hasn't been manufactured yet, then you're right. Of course the same is true for Nehalem...
If you mean that they don't have sample reticules (thanx to IDon'tCare for the term...), then you are incorrect. In fact, if you scroll back up in this thread, you'll find a die shot of one...

That's exactly what I'm saying. Fusion hasn't been designed yet, so there has never been a test mask made.

According to a interview on 7/16/2007 with Giuseppe Amato, Fusion is still at the idea stage.
They don't even know how they will build it yet.

Well then he either misspoke (what he may have meant to say is design stage which would last right up until tape-out) or is wrong...we've already seen a die shot of a test chip. Besides, if it's coming out anywhere near 2009, it HAS to be in the design stage.

However, I would very much like to read the interview...could you link it please?

Cheers...

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

 

[Viditor]

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

 

[ZOXXO]

Originally posted by: Viditor

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

Thats a photo of the bulldozer core. Not much fusing happening there.

 

[Phynaz]

Originally posted by: Viditor

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

Correction, that's the photo of a die. I see nothing that proves it's a fusion die.

 

[Stoneburner]

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

Correction, that's the photo of a die. I see nothing that proves it's a fusion die.

You don't see Ruby painting her toenails in there?

 

[intangir]

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

Correction, that's the photo of a die. I see nothing that proves it's a fusion die.

That's not even a photograph. That's a screenshot of a CAD tool plot of a processor that is presumably Bulldozer. That picture, by itself, does not remotely imply that actual silicon samples exist....

Compare this Barcelona die plot, released last June over two months before AMD announced Barcelona's tape-out, with an actual Barcelona die photo. See the difference?

 

[Viditor]

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

Correction, that's the photo of a die. I see nothing that proves it's a fusion die.

And I note that you still haven't provided a link...so we are left to compare and contrast this unprovable photo vs your unprovable recollection?
Sorry mate, the photo wins in my book...

 

[Acanthus]

I think the memory bandwidth limitation will prevent a high end solution for a long time.

CPU, GPU, and Physics all sharing the same memory bandwidth just seems bad.

Unless they start moving to Quad channel DDR3...

 

[Phynaz]

Originally posted by: Viditor

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

Correction, that's the photo of a die. I see nothing that proves it's a fusion die.

And I note that you still haven't provided a link...so we are left to compare and contrast this unprovable photo vs your unprovable recollection?
Sorry mate, the photo wins in my book...

I'm sorry, I thought I was following your example, since after repeated requests you didn't provide any substantiation to your remark that Henri Richard was going to resign from AMD a year ago.

"Fusion is little more than a name on a roadmap and there are still a lot of design and concept decisions to be made."

Right here.

"So far AMD has not made a final decision solidifying the future of Fusion"

Here.

Your "photo" has been trumped.

So much for my recollection being "unprovable"

 

[Viditor]

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

Correction, that's the photo of a die. I see nothing that proves it's a fusion die.

And I note that you still haven't provided a link...so we are left to compare and contrast this unprovable photo vs your unprovable recollection?
Sorry mate, the photo wins in my book...

I'm sorry, I thought I was following your example, since after repeated requests you didn't provide any substantiation to your remark that Henri Richard was going to resign from AMD a year ago.

"Fusion is little more than a name on a roadmap and there are still a lot of design and concept decisions to be made."

Right here.

"So far AMD has not made a final decision solidifying the future of Fusion"

Here.

Your "photo" has been trumped.

So much for my recollection being "unprovable"

As I thought...it was NOT a quote. The only quote that comes close actually says:
"With Fusion we?ll be able to offer single chip and multichip products, which will also make it possible to have parallel GPUs in the same architecture. So far AMD has not made a final decision solidifying the future of Fusion, but this is a possible route we might take"

How you get from this statement to your statement of

"According to a interview on 7/16/2007 with Giuseppe Amato, Fusion is still at the idea stage.
They don't even know how they will build it yet"

is beyond me...
The other link is just some guy's opinion...he doesn't attribute it to anyone else.
It makes no sense that a chip to be released in 2009 isn't already outputting test chips...

Bottom line, your recollection is incorrect...

 

[Phynaz]

Originally posted by: Viditor

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

Originally posted by: Viditor

Originally posted by: Phynaz

Wow, now you know more about Fusion than AMD's Europeon head of marketing.
And yet it was one of his quotes that you used to "prove" that Barcelona was K10.

Or maybe you read the quote wrong, or he was misquoted, or the interview is a reprint...
In any event, I don't see your point as this is a photo of the die...

Correction, that's the photo of a die. I see nothing that proves it's a fusion die.

And I note that you still haven't provided a link...so we are left to compare and contrast this unprovable photo vs your unprovable recollection?
Sorry mate, the photo wins in my book...

I'm sorry, I thought I was following your example, since after repeated requests you didn't provide any substantiation to your remark that Henri Richard was going to resign from AMD a year ago.

"Fusion is little more than a name on a roadmap and there are still a lot of design and concept decisions to be made."

Right here.

"So far AMD has not made a final decision solidifying the future of Fusion"

Here.

Your "photo" has been trumped.

So much for my recollection being "unprovable"

As I thought...it was NOT a quote. The only quote that comes close actually says:
"With Fusion we?ll be able to offer single chip and multichip products, which will also make it possible to have parallel GPUs in the same architecture. So far AMD has not made a final decision solidifying the future of Fusion, but this is a possible route we might take"

How you get from this statement to your statement of

"According to a interview on 7/16/2007 with Giuseppe Amato, Fusion is still at the idea stage.
They don't even know how they will build it yet"

is beyond me...
The other link is just some guy's opinion...he doesn't attribute it to anyone else.
It makes no sense that a chip to be released in 2009 isn't already outputting test chips...

It says the same thing, you can't know how you will build something if you don't know what you're building.

The "other guys opinion" was somebody that heard Phil Hester speak at the hot chips conference. It's called reporting. Phil Hesters' talk was paraphrased in simular fashion on many web sites. I'm sure you could use Google to check it out.

It's nice you think my recollection has gone from unprovable to mearly incorrect.

Got link for Henri Richard's resignation yet? (that's four requests now, btw).