The rise and fall of AMD

[mrmt]

Its disingenuous & condescending to throw away AMD's accomplishments by the wayside just because they acquired some other firms & aren't a monolith like Intel in the semiconductor market !

Oh, a lot of other companies push/pushed innovation. I could say some of them, some monoliths like Intel, others far smaller. Let's pick a list IBM, Nvidia, Qualcomm, Transmeta, DEC, PA Semi, Seamicro, Calxeda.

See? It's not a matter of size. Every one of those companies picked a concept and transformed in a product by their own means. They did not have to acquire the company A or B to develop something.

But then I ask you, what did AMD develop, from concept to product, in house?

 

[R0H1T]

But then I ask you, what did AMD develop, from concept to product, in house?

How about an APU, & no it isn't a concept from Intel/Nvidia, with fully working(like Kaveri) HSA ?

 

[mrmt]

How about an APU, & no it isn't a concept from Intel/Nvidia, with fully working(like Kaveri) HSA ?

APU? The first APU was Intel (Westmere), not AMD, and AMD only became interested in GPU after losing the mobile market to Intel Centrino.

HSA? Hybrid systems are a reality far older than AMD. Even the old Cell processor would be an example of hybrid system. It's the level of integration that is something new, but so far a lot of hot air and no products. But yes, we can count this as one point for AMD.

 

[Charles Kozierok]

I’m sorry but you are talking like a fanboy. If AMD copied them from Intel and Intel didn’t invent them, then Intel copied them from someone else. Sorry but most of what you wrote in the list above Intel copied them from someone else, they were not invented by Intel.

Oh, that's BS.

Intel taking a concept that was used by Seymour Cray in the 1950s or whatever and making it a functional part of a modern microprocessor is not in any way comparable in terms of R&D effort to AMD just licensing the technology from Intel after all the work was already done.

I'm not an Intel "fanboy", I've just extensively studied these technologies and I know that the vast, vast majority of what is in modern x86 CPUs is there because of the efforts of Intel, not AMD.

 

[galego]

I think it's pretty self-evident to any objective individual that Intel has been responsible for the lion's share of innovations in this segment of the computer industry.

Being an objective individual (as you repetidedly accuse me) I can notice how you failed to provide a list...

 

[Exophase]

Oh, that's BS.

Intel taking a concept that was used by Seymour Cray in the 1950s or whatever and making it a functional part of a modern microprocessor is not in any way comparable in terms of R&D effort to AMD just licensing the technology from Intel after all the work was already done.

I'm not an Intel "fanboy", I've just extensively studied these technologies and I know that the vast, vast majority of what is in modern x86 CPUs is there because of the efforts of Intel, not AMD.

AMD hasn't licensed any uarch technology from Intel after 286, unless you count microcode. In which case nothing after 486.

They've licensed ISA but that went both ways so...

 

[R0H1T]

APU? The first APU was Intel (Westmere), not AMD, and AMD only became interested in GPU after losing the mobile market to Intel Centrino.

APU is just a marketing term, but I think you already know that.

HSA? Hybrid systems are a reality far older than AMD. Even the old Cell processor would be an example of hybrid system. It's the level of integration that is something new, but so far a lot of hot air and no products. But yes, we can count this as one point for AMD.

HSA, as you rightly said, purely as a concept is far older than much of this modern PC industry(as it stands now) but the only true(r) implementation has been from AMD & since this is the next "big thing" I'd say it counts more than just one in my book. Also if you're gonna strike every innovation as an "implementation of previous/dated concepts" then we might as well strike out the whole modern semiconductor & consumer electronics industry since alot of these concepts were from sci-fi flicks !

 

[mrmt]

HSA, as you rightly said, purely as a concept is far older than much of this modern PC industry(as it stands now) but the only true(r) implementation has been from AMD & since this is the next "big thing" I'd say it counts more than just one in my book. Also if you're gonna strike every innovation as an "implementation of previous/dated concepts" then we might as well strike out the whole modern semiconductor & consumer electronics industry since alot of these concepts were from sci-fi flicks !

A concept in project parlance is far different than a HQ, and it is this kind of concept I'm talking about.

As for HSA, we don't know what to expect from AMD yet. A lot of hot air, no products developed yet and no software support on the market (but this is SOP in AMD's case).

I'd suggest a wait and see approach, because while possibilities are infinite in the realm of the PPT-engineering, AMD has yet to launch an HSA product with suitable HSA software support to showcase their invention.

 

[Charles Kozierok]

AMD hasn't licensed any uarch technology from Intel after 286, unless you count microcode. In which case nothing after 486.

They've licensed ISA but that went both ways so...

There's been a lot of licensing in recent years, but of course they don't hand out that info to the public, so I don't really know exactly what it is.

 

[Exophase]

There's been a lot of licensing in recent years, but of course they don't hand out that info to the public, so I don't really know exactly what it is.

According to who?

We know a lot about CPU uarch of Intel and AMD CPUs and nothing stands out as cross-licensed technology.

 

[R0H1T]

A concept in project parlance is far different than a HQ, and it is this kind of concept I'm talking about.

The concepts that you're talking about have come to fruition more because of breakthroughs in the fields of physics & chemistry rather than any true innovation. You wouldn't get the power of an IVB if it weren't for 22nm FinFET likewise E=mc^2 is perhaps the last original concept from the 20th century, all else is just derivation !

 

[Fox5]

AFAIK, Intel's TSX extensions are the first commercial implementation of transactional memory. That's kind of cool. AMD had announced such a thing years ago, but never made a product.

 

[Sleepingforest]

The concepts that you're talking about have come to fruition more because of breakthroughs in the fields of physics & chemistry rather than any true innovation. You wouldn't get the power of an IVB if it weren't for 22nm FinFET likewise E=mc^2 is perhaps the last original concept from the 20th century, all else is just derivation !

How are breakthroughs in physics and chemistry not innovation?

 

[Exophase]

Argh double post can someone please delete this or something

 

[Exophase]

AFAIK, Intel's TSX extensions are the first commercial implementation of transactional memory. That's kind of cool. AMD had announced such a thing years ago, but never made a product.

IBM has had it in products for a while, afaik so did Sun's Rock processor. Transmeta's processors had a form of it.

 

[R0H1T]

How are breakthroughs in physics and chemistry not innovation?

Did I say that ? But they wouldn't fit in one's definition of an "original concept" & if you carefully follow this page you'll see how innovations are quantifiable in some weird context i.e. X outweighs Y !

 

[Sleepingforest]

Then nothing is really "innovation" because it's all ultimately based on logic. You have to draw a line somewhere. A new idea, like FinFET, deserves to be called innovation. A marginal "improvement", like rounding rectangle corners, is not innovation.

Did I say that ? But they wouldn't fit in one's definition of an "original concept" & if you carefully follow this page you'll see how innovations are quantifiable in some weird context i.e. X outweighs Y !

Also, you say:

more because of breakthroughs in the fields of physics & chemistry rather than any true innovation.

Which implies "breakthroughs" are not innovation.

 

[mrmt]

The concepts that you're talking about have come to fruition more because of breakthroughs in the fields of physics & chemistry rather than any true innovation. You wouldn't get the power of an IVB if it weren't for 22nm FinFET likewise E=mc^2 is perhaps the last original concept from the 20th century, all else is just derivation !

As I said, I'm talking about concept in project parlance, don't come here with a straw.

 

[R0H1T]

As I said, I'm talking about concept in project parlance, don't come here with a straw.

And like I replied earlier the implementation had more to do with how much you could afford such stuff, cause implementation is not only dependent on the R&D cost but also if you can make money out of it or market it like P4, & AMD cannot match Intel in this dept ! Don't go around in circles here, you're saying I'm avoiding the argument but its not a sound argument in the first place, cause even if you can improve say a hybrid car doesn't mean you have the money for R&D & sales later on !

 

[R0H1T]

Then nothing is really "innovation" because it's all ultimately based on logic. You have to draw a line somewhere. A new idea, like FinFET, deserves to be called innovation. A marginal "improvement", like rounding rectangle corners, is not innovation.

I would call this a breakthrough not innovation, you can put SOI in the same category cause they're flip slides of the same coin called efficiency.

Also, you say:

Which implies "breakthroughs" are not innovation.

We can discuss this at great lengths, but to no avail, however I'd call a smartphone an innovation & certainly not a breakthrough. I'd like to get your opinion on this ?

 

[AnandThenMan]

I'm not an Intel "fanboy",

Then why are you turning this thread into all about Intel and how great they are and how they invented pretty much everything in your computer?

 

[AtenRa]

But then I ask you, what did AMD develop, from concept to product, in house?

AMD Patents from 21st June 2012 to 18th April 2013, I’m sure you will find thousands for the last 20 years. A lot of those were integrated/used in to the products we all know all these years.
http://stks.freshpatents.com/AMD-sym.php

Also, a glimpse from the past.

Edit: Just to add,

AMD has been in the TOP 100 Global Innovators for 2012

 

[aigomorla]

And what has AMD done that's really been innovative? Not much. They mostly just copy Intel.

*looks at AMD64 on my Windows 7 64-bit OS*

I guess 64 bit instructional set which my intel uses isnt much?

looks at my space heater err prescott cpu and is thankful of
AMD telling intel PUAHAHAAHAHAHAHAHAHA??

I think AMD has done some nice stuff... it just all went downhill on bulldozer.

 

[mrmt]

And like I replied earlier the implementation had more to do with how much you could afford such stuff, cause implementation is not only dependent on the R&D cost but also if you can make money out of it or market it like P4, & AMD cannot match Intel in this dept !

And here comes again the R&D/size excuse... Look, what can you say about NexGen, PA Semi, Calxeda, Transmeta and others, you know, companies with far smaller budget than AMD but that at least could bring something from concept to product?

Face it: AMD is not innovative as you think it is. R&D is not in their DNA, not in their business model. They are quite good in bringing in other's researches and get it on the market, but they are not an innovation engine.

 

[AtenRa]

Face it: AMD is not innovative as you think it is. R&D is not in their DNA, not in their business model. They are quite good in bringing in other's researches and get it on the market, but they are not an innovation engine.

You really have to take off the ignore :whiste:

Edit: Someone quote my post (#271)