AMD Benefitting from its Diminutive Size says CEO

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,532
AMD is getting its groove back hopefully, and maybe a little bit of spunk goes a long ways! All HardOCP readers certainly know that competition in our enthusiast industry is needed and has been sorely lacking for some time now.

“We understand we’re the smaller guy,” Su added, referring to rival Intel. “In some sense, you can view that as daunting. Yes, our competition may have significantly more engineers or significantly more R&D investment. On the other hand, we have significantly more freedom. We have the freedom to innovate.”

However, when it comes to VR, we have shown for months now that AMD does not come close to delivering on its RX480 "premium VR experiences" promise.

Of course, Su acknowledged that VR has yet to reach its full potential.

“What’s holding it back right now is, it’s still a little complex to use, still a little bulky,” Su told said. “With technological advances, there’s a very natural curve between cost and complexity and adoption. When the cost and complexity are high, the adoption rate is, let’s call it ‘modest.’”

Of course Su is referring to sales of HMDs rather than AMD's lacking performance. Good dodge. Here's hoping that AMD's new Vega GPU brings with it some sort of performance parity in the realm of VR. Certainly we will be investigating that with objective data.
 
Man, all this talk of grooves and spunk has me REALLY excited for Zen+Vega!
 
"Smaller guy" is only a positive for a start up, AMD you're hardly a small company and you're only so small because of your lack of competition in face of Intel even though that's what you were founded on. Also more engineers and R&D investment = freedom, freedom to blow money on experimental shit because your company can take a hit from a non-performer.

Nothing against AMD just hate buzz word bullshit.
 
Someone send Roy an email and link this thread!

All kidding aside, I think it's great that AMD is going start offering hardware we want with both CPU and GPU. Their new chipset is also fantastic. Competition is good for us. I just hope they can keep it up and continue innovating.
 
I'm amazed Lisa Su has held onto that job with all the flops during her tenure.
 
Last edited:
The book Thinner by Stephen King was fairly good and a good comparison of AMD's trajectory. I doubt AMD will be able to put its problems into a pie, but will effectively get to the same inevitability as the book did. :D
 
I'm amazed Lisa Su has held onto that job with all the flops during her tenure.

AMD needs a homerun now both CPU and GPU so that I can hopefully save a few bucks on my next Intel/Nvidia purchase.
Well, all the "flops" were well on the way long before she took control. Put that on Rory.
 
I'm liking that Su now essentially acknowledges that public PR was/is their biggest weakness, certainly in the consumer market.

Seems she is actually getting AMD's shit back together again! :cool:

Nice! But it must've been a hell of a lot hardwork to get them *just* here; so I hope it pays off...
 
Last edited:
I am back and forth between optimism and pessimism concerning VR. On one side I really, really, want it to take off. On the other the lack of investment and software have me chalking it up as another failing attempt at evolving the medium. I mean it has been poked at since 1995, perhaps earlier, with no mainstream adoption to speak of. Add to it that CES did not highlight any innovations or articles about it this year worries me further.

All in all there are a lot of hurdles that need to clear on the VR side before I need to worry about what video card my system is running.

I read into Su's comments as an eye roll at a topic that isn't exactly effecting the bottom line of their current GPU plan. The adoption is not there, the format is currently in toddler stages, and the cost of entry is just too steep. I cannot fault AMD 100% for failing to deliver on their "Premium VR" promise since the last time I even cared about donning a HTC headset was before Thanksgiving. If I had more reason to play VR I might care about my rig. Currently, my 4k monitor does the job.

Outside that, I can say she is definitely seeking that silver lining with the innovation comment, and in hindsight it is paying off. Everything else has been 'bleh' as a result of their smaller size.
 
I havent try VR, and i am not keen to try it yet, it does seem like there is a lot of road ahead.
 
Hopefully by the time I jump into vr with both feet, AMD will be more competitive.
 
I'm hoping that AMD will be competitive in the CPU segment, and really drive some competition across all segment, from budget to high end.
The price of top end consumer CPU like a 10 core chip is just retarded.
 
Video cards are not the barrier for entry in vr on the pc, it's the price of the HMD. If you can afford to shell out 800$ for a Vive or a Rift with remotes the video card is a non issue. They need to do two things to really drive the vr market, 1 make hmds wireless and more comfortable, 2 cut the price down to 3-400 while steadily increasing resolution with video card power.

Start with those with some quality AAA games(you need at least 20 that are not gimmicky bullshit) and you might be getting somewhere
 
"Yes, our competition may have ...... significantly more R&D investment. On the other hand, we have significantly more freedom[to innovate].”

Phew, good thing then that "freedom to innovate" doesn't require a lot of research, eh?
 
I don't think people have as good of a handle on how hard accomplishing VR gaming is going to be. The way it stands now , all of the improvements are coming from hacks and cheat optimizations to lessen the fidelity so the bandwidth and processing needed to run well. no big VR stuff at CES for some good reasons.

It is so much like the early days of mp3 adoption were companies didn't know how to present the concept to the consumer market but still feel the need to push new product and tech at consumers in the hopes that demand will fuel the industry and they can somehow leverage the early adopters habits to shape the future.

The trick is that the content industry was already in dire need of a shift away from 1080p into "4K" UHD and HDR and larger color gamuts after the failure of pushing consumer "3D" that they didn't also want to have to re-imagine the entire process and industry again to push "VR"

All the bandwidth and hardware needed for accomplishing "4K" UHD and HDR was already fairly doable or easy enough to target for years now ( ie 10-12 bit displays and higher resolution screens) they juest needed a set of industry standards to rally around, but for VR to happen in a way consumers demand it become mainstream, it needs new technology to be invented so it can be implemented with low enough latency and it still needs all the display tech innovations that HDR and "4K" demand and consumers are still going to want content. Cheap smartphone based "VR" has almost undermined the industry before it has gotten off the ground

The you have AMD, desperate to be at the forefront of VR and trying every trick in the book and idea it can to stay competitive in the marketplace but still having to keep up with what gamers want (4K , HDR, freesync , good drivers, game streaming and capture, etc) and try to push back at Nvidia with the RX480 claiming it is a viable solution, I think it shows how being the "small guy" isn't advantageous when Nvidia can address all those same issues but also target AI ,super computing, the auto industry, console gaming , and set top boxes but still resist AMD's GPU efforts.

I want AMD to have a good product, but I don't want them selling the consumer the idea that they get a advantage of being the little guy when the end result is products like the RX480 and AMD saying it is good enough.

I hope ZEN is a good start to get them moving the company forward, and they look a lot better off than they did in years past and maybe investors will buy into the little guy PR.
 
You cant have more freedom in R&D when you cant pay for it. You never get the R&D you didn't pay for.

If anything, there is bigger freedom with bigger R&D budgets due more focus can be done outside the essential core areas.
 
VR is overhyped anyway. It's the next Jaguar, Sega 32x CD... *cough* Virtual Boy. Maybe they'll be able to throw enough money and hype at it that it actually becomes a thing (who knows when), but excuse me if I'm not on the bandwagon of our VR lord and saviour. Just another gimmick to milk more money, another miniscule "upgrade" (costing you an arm and a leg) which just adds yet another layer of bullshit on top the others ones.

AMD should just focus wholeheartedly, single-mindedly, and neurotically on making one Godzilla of a processor that slays Intel waffles morning, noon, and night, blindfolded, and with both hands and feet tied. That should be their singular mission; forget the gimmicks, forget the fluff.

:/
 
If R&D staff and $$$ was all it took, then we need to ask some very pointed questions: Why has Intel been an utter failure at competing with ARM in the mobile space? Why has Intel been an utter failure in developing a graphics processor, only just now coming out with a product that is "acceptable?" Why have we not seen anything truly progressive from Intel in years? How was it that Intel had to use their monopoly power to crush AMD in the mid 00's instead of simply innovating a better product?

Simply put, Intel is a monolithic company that has a very set way of doing things, and because of the Intellectual and Corporate culture, is unable to think "outside the box." A smaller company that must do things differently to compete has the flexibility to innovate. That company may fail, but they also have the option to produce in avenues that the larger company thinks impossible. There is definite merit to the argument that AMD's diminutive size can be leveraged to an advantage.
 
If R&D staff and $$$ was all it took, then we need to ask some very pointed questions: Why has Intel been an utter failure at competing with ARM in the mobile space? Why has Intel been an utter failure in developing a graphics processor, only just now coming out with a product that is "acceptable?" Why have we not seen anything truly progressive from Intel in years? How was it that Intel had to use their monopoly power to crush AMD in the mid 00's instead of simply innovating a better product?

Simply put, Intel is a monolithic company that has a very set way of doing things, and because of the Intellectual and Corporate culture, is unable to think "outside the box." A smaller company that must do things differently to compete has the flexibility to innovate. That company may fail, but they also have the option to produce in avenues that the larger company thinks impossible. There is definite merit to the argument that AMD's diminutive size can be leveraged to an advantage.

ARM is about ecosystem and cost cutting. Same way why you cant touch x86 in the PC/Server segment with ARM or MIPS.

Define failure of a graphics processor. Do you mean the IGP? What extra value can you charge from IGP? *hint* Also from a perf/watt perspective they do quite well compared to AMDs IGPs dont they ;)

AMD in the mid 2005 delayed 65nm and thought K8 could never be beaten. Yet Pentium-M already beat the K8 in all metrics. And the CEO of AMD dreamt of only selling highend CPUs and server SKUs leaving the rest to Intel. Funny how that haunted and still haunts AMD today.

Why isn't VIA and Matrox the new best thing? They must have even more freedom than AMD. I mean R&D doesn't matter right? AMD will beat Nvidia and Intel on a 1/10th R&D budget or less due to this amazing "freedom".
 
Last edited:
If R&D staff and $$$ was all it took, then we need to ask some very pointed questions: Why has Intel been an utter failure at competing with ARM in the mobile space? Why has Intel been an utter failure in developing a graphics processor, only just now coming out with a product that is "acceptable?" Why have we not seen anything truly progressive from Intel in years? How was it that Intel had to use their monopoly power to crush AMD in the mid 00's instead of simply innovating a better product?

Simply put, Intel is a monolithic company that has a very set way of doing things, and because of the Intellectual and Corporate culture, is unable to think "outside the box." A smaller company that must do things differently to compete has the flexibility to innovate. That company may fail, but they also have the option to produce in avenues that the larger company thinks impossible. There is definite merit to the argument that AMD's diminutive size can be leveraged to an advantage.

I'd say it's that AMD cannot afford to be lazy, they have far more at stake, while Intel has all the time in the world to lay about. Being given unlimited resources is not necessarily a good thing, in fact, it can be contraproductive. It's the same way a challenging game can make you git good in a hurry, while playing the most casual weed-infused daydream of a game can atrophy whatever skills you had to begin with.

Competition doesn't drive innovation (or skill), challenge does. Competition can be part of the challlenge, of course. This also comes with its risks.
 
I'd say it's that AMD cannot afford to be lazy, they have far more at stake, while Intel has all the time in the world to lay about. Being given unlimited resources is not necessarily a good thing, in fact, it can be contraproductive. It's the same way a challenging game can make you git good in a hurry, while playing the most casual weed-infused daydream of a game can atrophy whatever skills you had to begin with.

Competition doesn't drive innovation (or skill), challenge does. Competition can be part of the challlenge, of course. This also comes with its risks.

You can reformulate it as: You never get anything you didn't pay for in R&D. However you may not get what you paid for either.
 
ARM is about ecosystem and cost cutting. Same way why you cant touch x86 in the PC/Server segment with ARM or MIPS.
Intel has tried to compete with and conquer ARM, and has so far failed. True, ARM cannot displace x86, but Intel actively has tried to displace ARM, and has failed.

Define failure of a graphics processor. Do you mean the IGP? What extra value can you charge from IGP? *hint* Also from a perf/watt perspective they do quite well compared to AMDs IGPs dont they ;)
This is only something that happened quite recently. Until the super expensive Broadwell, as an integrated chip, Intel's processors could not even come close to competing with AMD's APU line. Even now, to get anything that compares in the graphics performance requires the purchase of a processor that is 2-3 times more expensive. AMD has an integrated GPU selling for a fraction of the cost that at worst is equal to Intel's, and at best vastly outperforms. If R&D $$$ and staff was all it took, the situation should be reversed.

AMD in the mid 2005 delayed 65nm and thought K8 could never be beaten. Yet Pentium-M already beat the K8 in all metrics. And the CEO of AMD dreamt of only selling highend CPUs and server SKUs leaving the rest to Intel. Funny how that haunted and still haunts AMD today.

AMD has certainly made their own missteps. But they were poised to make a serious dent in Intel's market share, and Intel could not come out with a product for years that could touch Athlon, and Athlon64. Intel had to resort to flexing their monopoly powers (got sued and lost for it) to keep AMD from becoming a serious competitor.

Why isn't VIA and Matrox the new best thing? They must have even more freedom than AMD. I mean R&D doesn't matter right? AMD will beat Nvidia and Intel on a 1/10th R&D budget or less due to this amazing "freedom".
These two companies are not the best examples. Instead of competing head to head, they decided to focus on a niche market where no one else was interested in competing. They do not attempt to innovate or do anything differently, and have no desire to do so. While they could change course rather easily, they do not wish to, and are content to remain in their small, niche markets.

I'm not saying that R&D doesn't matter, but those that think that it is everything are overlooking Intel's missteps, which have been many.
 
You never get anything you didn't pay for in R&D.

This is the second time you dropped that line in this very thread. You have a point but it's still too much of a blanket statement. Heard of viagra? It was developed - an actual paid for R&D - with something entirely different in mind. That's life.
 
This is the second time you dropped that line in this very thread. You have a point but it's still too much of a blanket statement. Heard of viagra? It was developed - an actual paid for R&D - with something entirely different in mind. That's life.

Hmm... I'd say that's more in line with Sony developing the PS3 and it ending up in supercomputing. Not developed for it, but still found a use. They still got pretty much what they asked for, a good concurrent processor/architecture. Just because the use differs from what was intended, doesn't mean they didn't get what they paid for.
 
Intel has tried to compete with and conquer ARM, and has so far failed. True, ARM cannot displace x86, but Intel actively has tried to displace ARM, and has failed.

So has ARM. The line of bankrupt companies is long ;)

Broadcom´s Vulkan was the last to be given up and all R&D lost.

This is only something that happened quite recently. Until the super expensive Broadwell, as an integrated chip, Intel's processors could not even come close to competing with AMD's APU line. Even now, to get anything that compares in the graphics performance requires the purchase of a processor that is 2-3 times more expensive. AMD has an integrated GPU selling for a fraction of the cost that at worst is equal to Intel's, and at best vastly outperforms. If R&D $$$ and staff was all it took, the situation should be reversed.

It happened before Broadwell. And its because of the TDP limits that forces AMD to lower clocks and/or throttle. In terms of R&D, take a wild guess on how much Intel spends on graphics. Its very low and due to the fact that "nobody" buys a CPU due to its IGP performance.

AMD has certainly made their own missteps. But they were poised to make a serious dent in Intel's market share, and Intel could not come out with a product for years that could touch Athlon, and Athlon64. Intel had to resort to flexing their monopoly powers (got sued and lost for it) to keep AMD from becoming a serious competitor.

AMD was production limited and the board at AMD loved it. Because then they didn't have to bother with discount chips for the plebs. AMD desired to spend it on ATI instead of fabs and/or R&D. It didn't affect AMD in a great way, hence why the fine was so low as well. I know the fans like to sugarcoat it much more than it was.

These two companies are not the best examples. Instead of competing head to head, they decided to focus on a niche market where no one else was interested in competing. They do not attempt to innovate or do anything differently, and have no desire to do so. While they could change course rather easily, they do not wish to, and are content to remain in their small, niche markets.

They have tried and failed due to R&D limitations. After that they seeked elsewhere. AMD however, being 2/3rds a semicustom company, still tries to focus all its R&D in areas that is very questionable for them. Their graphics division lost big time on this as well.

I'm not saying that R&D doesn't matter, but those that think that it is everything are overlooking Intel's missteps, which have been many.

And people tend to overlook the Pentium-M in that regard completely. And it included AMDs management too,
 
I'm probably just piling on at this point, but if AMD could funnel more resources to chip development maybe the Bulldozer might have been as "good" as mediocre and Zen and Vega development might not have suffered so many delays.

Almost forgot to add, isn't this what British would call "cheap and cheerful" as they spit it out through clenched teeth?
 
One sad thing with AMD is their 64 bit tech was so much better than Intel's and HPs plan for 64 bit itanium, that Intel ended up implementing AMDs x86-64, they can innovate from a underdog position, but I don't want it to only happen when Intel makes a misstep.

I'm excited for what ZEN is bringing to the table, but if AMD is playing games with VEGA and has it ready but didn't release anything at CES 2017 because Nvidia didn't announce a 1080Ti or whatever, it isn't thinking like the little guy, it is thinking like a PR and Marketing drone.

AMD seems to be in a good position to dominate DX12 gaming if it gets out ahead of Nvidia just as games are moving to DX12 with a better product, but delaying stuff for marketing reasons would be a sad reason to lose out on a potential lead so I hope it is a matter of AMD wanting to do a "hard" launch with product at retailers rather than a paper launch at CES is Nvidia didn't announce anything either.
 
VR is overhyped anyway. It's the next Jaguar, Sega 32x CD... *cough* Virtual Boy.

Given the amount of sales, research, development, and products already out there for VR (including multiple HMDs from large companies)...you are already wrong.
 
One sad thing with AMD is their 64 bit tech was so much better than Intel's and HPs plan for 64 bit itanium, that Intel ended up implementing AMDs x86-64, they can innovate from a underdog position, but I don't want it to only happen when Intel makes a misstep.

I'm excited for what ZEN is bringing to the table, but if AMD is playing games with VEGA and has it ready but didn't release anything at CES 2017 because Nvidia didn't announce a 1080Ti or whatever, it isn't thinking like the little guy, it is thinking like a PR and Marketing drone.

AMD seems to be in a good position to dominate DX12 gaming if it gets out ahead of Nvidia just as games are moving to DX12 with a better product, but delaying stuff for marketing reasons would be a sad reason to lose out on a potential lead so I hope it is a matter of AMD wanting to do a "hard" launch with product at retailers rather than a paper launch at CES is Nvidia didn't announce anything either.

They are not delaying anything because of marketing, they are still on their time table of the 1H of 2017. I will not be looking to get a new card for a while even then because I will be upgrading both my computers to Zen. :) That and my 2 x Sapphire Furies will tide me over for quite a while for what I do and play.
 
One sad thing with AMD is their 64 bit tech was so much better than Intel's and HPs plan for 64 bit itanium, that Intel ended up implementing AMDs x86-64, they can innovate from a underdog position, but I don't want it to only happen when Intel makes a misstep.

I'm excited for what ZEN is bringing to the table, but if AMD is playing games with VEGA and has it ready but didn't release anything at CES 2017 because Nvidia didn't announce a 1080Ti or whatever, it isn't thinking like the little guy, it is thinking like a PR and Marketing drone.

AMD seems to be in a good position to dominate DX12 gaming if it gets out ahead of Nvidia just as games are moving to DX12 with a better product, but delaying stuff for marketing reasons would be a sad reason to lose out on a potential lead so I hope it is a matter of AMD wanting to do a "hard" launch with product at retailers rather than a paper launch at CES is Nvidia didn't announce anything either.

x64 isn't better. It got legacy support however. Something we have to suffer with the next 30 years or more. Same reason why ARM and x86 is entrenched and cant touch one another and no other instruction set got a chance.

AMD doesn't have a better DX12 product, they however are good at sponsoring DX12 titles ;)

Vega is about as ready as Polaris was in February 2016. 6-9 months lead BS and all.
 
Given the amount of sales, research, development, and products already out there for VR (including multiple HMDs from large companies)...you are already wrong.

Justin Bieber has sold more than 15 million records and has a net worth of 200 million USD. Just because it sells doesn't mean it's any good.
 
Justin Bieber has sold more than 15 million records and has a net worth of 200 million USD. Just because it sells doesn't mean it's any good.

You just had to go there. :D :) ;) The only reason I would bother with VR is flight simulation and I will not do it because of guaranteed motion sickness for me. :(
 
Justin Bieber has sold more than 15 million records and has a net worth of 200 million USD. Just because it sells doesn't mean it's any good.
I have been doing this for 20 years and I flatter myself for knowing what is a fad and what is forward-looking technology. VR is here to stay.
 
Back
Top