The Rise and Fall of AMD

H

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
There is an editorial posted today at TechSpot titled "The Rise and Fall of AMD" that will no doubt spark a interesting debate. Here's a quote:

AMD has long been subject of polarizing debate among technology enthusiasts. The chapters of its history provide ample ammunition for countless discussions and no small measure of rancor. Considering that it was once considered an equal to Intel, many wonder why AMD is failing today.
Click to expand...
 
Did I miss out during reading (slightly feverish, lowered senses) or did that article not bring up Intel's anti-competitive activities against AMD and just go "hurr it's all their own fault"?
 
I don't think anyone will ever know what really happened. That deal that went down behind closed doors not long ago (a couple years I think? maybe a few), where essentially Intel had them by the balls and essentially could have put them out of business if they wanted (by halting all their x86 production) is probably what did it.
 
essentially twice in the same sentence... damn the lack of edit button! damn it!!!
 
MrLonghair said:
Did I miss out during reading (slightly feverish, lowered senses) or did that article not bring up Intel's anti-competitive activities against AMD and just go "hurr it's all their own fault"?
Click to expand...

Only 4 paragraphs dedicated to that on the second page. Surprised to hear that NetBurst started in the original Pentium chips, though. :rolleyes: The overall feeling was AMD biased, but an ok read.
 
MrLonghair said:
Did I miss out during reading (slightly feverish, lowered senses) or did that article not bring up Intel's anti-competitive activities against AMD and just go "hurr it's all their own fault"?
Click to expand...

They covered it, specifically mentioning things like Intel's deal with Dell. They also correctly pointed out that AMD didn't have capacity to produce more chips at the time even if Intel hadn't been screwing them over. AMD did get screwed in the settlement though because Intel knew they were in such a bad financial position that they needed money right then.
 
darkpaw said:
They covered it, specifically mentioning things like Intel's deal with Dell. They also correctly pointed out that AMD didn't have capacity to produce more chips at the time even if Intel hadn't been screwing them over. AMD did get screwed in the settlement though because Intel knew they were in such a bad financial position that they needed money right then.
Click to expand...

The cross-licensing agreements that came from that settlement were worth more than the cash sum.

http://www.tomshardware.com/news/intel-amd-settlement-antitrust-competition,9067.html

Good read, but he should have elucidated the monopolistic practices were a lot deeper than that, with paying OEMs over a period of years that amounted to a lot more than the 1.25B $$$.

Though it is true, under Ruiz -- and even before him -- AMD was slipping in their fab capabilities and R&D was essentially halted.
 
The article was pretty good. Though AMD's 386 and 486 offerings were faster in terms of offered clock speeds, they were slower than Intel's per clock. But other than I thought it was pretty much spot on.
 
If it weren't for AMD putting a fire under Intel's ass, we would not be where we are today.
 
They didn't mention the licensing of Zilog Z8000 because Intel didn't want to deal.
 
Dan_D said:
The article was pretty good. Though AMD's 386 and 486 offerings were faster in terms of offered clock speeds, they were slower than Intel's per clock. But other than I thought it was pretty much spot on.
Click to expand...

Well, yes.

But the clockspeed compensated for the IPC.(unlike BD and PD) Still not as fast but pretty close.
 
Even if AMD is "done," it'll be a long time before they shutter the doors. Asuming they follow through on their plan to restructure and move into markets where they can be competitive and profitable, they'll be around for a while.

As much as I hate to say it, I think the best thing they can do, for the time being, is back out of the desktop market.
 
To bad AMD may take ATI down with them... As in laying off the PEOPLE that make ATI, or selling patents to the highest bidder and breaking up the company that way.

Can only hope that doesn't happen... doesn't look good though.
 
I don't think AMD is done, just their desktop CPU 's. They're still making money in gpu's and APU's and this new deal with ARM looks promising. They may just lop off all Tue divisions where theyre not competitive and/or losing money and focus everything in the few areas where they're doing well.
 
Haven't read it yet because the link isn't opening at work for some reason, but does the article make any mention of possible sale of patents and such? I wonder how much interest there would be for some of the cross licensed technology with Intel, by parties outside of Intel and if this would keep Intel from lowballing AMD in the event of a sale.
 
I don't know what rise and fall they are talking about. AMD has always been small and never "equal" to Intel in any respect. I love AMD, but title is a stretch. Intel is about 70x larger than AMD.
 
KrylonArt said:
Haven't read it yet because the link isn't opening at work for some reason, but does the article make any mention of possible sale of patents and such? I wonder how much interest there would be for some of the cross licensed technology with Intel, by parties outside of Intel and if this would keep Intel from lowballing AMD in the event of a sale.
Click to expand...

No, but if there's a change of ownership the cross-licensing agreement would be void thus require Intel playing ball, and that just doesn't seem likely.

Granted, AMD leaving the x86 market would be a death knell for x86 as the dominant player in the PC and entire Computing industry. Company's wouldn't be as willing to dive into a monopoly if there's stiff competition elsewhere (see ARM)
 
MikeTrike said:
I don't know what rise and fall they are talking about. AMD has always been small and never "equal" to Intel in any respect. I love AMD, but title is a stretch. Intel is about 70x larger than AMD.
Click to expand...

I thought that when I read the first part of it but it doesn't ever talk about them being financially anywhere near as successful. It just talks about the products themselves. It goes on to further talk about how AMD was held back by Intel's business practicies, but that even if Intel hadn't done that, they'd never have been as successful as people might think. AMD never had the capacity to produce CPUs in the quantities that HP, Dell etc. would have demanded.
 
I can't help but point to the 1999 like reviews of the FX as the reason for their latest stock tanking.

Lots of reviews came out and basically said "FX sucks." but I'm not sure we're all on the same planet, or maybe I'm just odd.

See, even when playing a game I usually have teamspeak open, the game of course, probably a browser in the background (and modern browsers themselves often split up threads) and run a VLNA server from my desktop PC. In addition I have an antivirus, the creative bloatware, and a probably a gazillion other things I'm not thinking of all running in the background.

I don't reboot my 7 PC like I used to back in the day either, and often there's another account logged in, usually running my son's mindcraft and who knows what other crap he's got.

Now you take all that crap, and add a multithreaded game. Something has to give. Sure, I could reboot. But half that crap in 2012 is just a fact of life.

I don't know, an 8 core CPU sounds pretty decent. It seems like no one ever mentions this, and it's actually the reason I'm getting an FX. I haven't dropped big money on video cards in years. And unless you do, I don't think you'd see a big performance difference.

Then what about future titles that better multithreaded?

I don't know, it just seems like the FX was a well conceived processor, or maybe i'm not the typical user in 2012.
 
Even during the "good times" for AMD when they were pumping out Athlon64s, Opterons, and dual-core processors, everyone I ever talked to about it knew it was only a matter of time. AMD never had resources even close to what Intel had.

The only reason AMD did so well at that point was because Intel was slacking on the other end. If they hadn't been so invested in netburst, and other side projects like the Itanic, AMD never would have stood a chance.

I wouldn't count AMD completely out however. Every time I go into Fry's it's a pretty clear picture. Cheap <$300 computers using AMD E-series or A-series APUs constantly flying off the shelves while the more expensive Intel systems largely collect dust. Even the cheap celeron systems tend to still carry a price premium over the AMD systems, and it seems pretty clear that most people care about price more than anything.
 
^ I said crap a lot in that post, which ironically is also probably crap.
 
A few historical points:

The 2900 series mentioned existing in 1975 was more of a "bit slice computer" than just a series of chips. String enough chips together and you could have a real computer. Some of the more famous computers from that era used this.

Also the story played down the pentium-pro/II/III rivalry with the K6[-2]. The K6 might have competed just fine against the Pentium[1/MMX], but it wasn't nearly up to the mighty P6. Despite AMD managing 300-450MHz off some K6-2s they really had to get the K7 out.

Jerry Sander's famous "real men have fabs" quote seemed to drive AMD's fab policy. It may well be they needed those fabs (it looks from outside like they are having issues with GobalSemi's process, but can't do anything about it), but they could never afford enough. This story basically begins and ends with fab capacity.

A sidebar discussing just what Intel did to make sure AMD never amounted to anything would help. Also some sort of mention about Intel suing Cyrix into oblivion: through all the lawsuits, intel never won once. Cyrix had to employ 10 lawyers for every engineer, and eventually they either couldn't build a chip, or cut down plans to far (any cyrix fans remember what the story was with Jalepeno?)
 
The article was very spot on and after reading it I believe that there are quite a lot of changes coming to AMD. Will it be viable enough to get AMD working again where two previous CEOs-- Ruiz and Meyer-- have failed (miserably at that)? Can Read's success at Lenovo be carried over to AMD? Only time will tell. In my opinion, I see Read as the necessity to AMD, and the one most willing to take risks and chances to get AMD up and running again. That means if it takes layoffs and cutting staff to just break even by next year, then it's a start even if a small one.

Read looks like he is shifting focus away from the non-competitive, unprofitable and gradually slowing desktop processor market and moving it towards mobile and server market. And, if the article is to be believed, that means desktop processors will be delayed by an extra year. Opteron processors will probably be delayed or released before their consumer desktop variants. So any AMD diehards will have to wait longer now for a desktop processor upgrade. I would not be surprised a second stepping of Piledriver will come out between now and 2014 to ease the wait.

The way I see it on AMD's product roadmap, everything that I and many others have seen in a previous roadmap should be completely disregarded. By my estimations, it looks like this:
  • Kaveri APU-- first combined "fusion" of CPU and GCN cores-- moved from Q1 2013 as per pre-Rory Read roadmap to early 2014.
  • Steamroller-based FX/Opteron processors moved from Q1 2014 to probably late 2014.
  • AMD HD Radeon 8000-series "Sea Islands" remains on course for late Q1 2013.
  • Mobile and embedded APUs-- Brazos and 5-Watt Temash-- remain on course for 2013 on the post-Rory Read roadmap.
  • Second stepping of Trinity desktop and mobile APU-- Richland-- moved into Kaveri's time slot.
  • Opteron processors based on Piledriver modules paired with ARM cores for server markets moved into 2013 release.
  • Updated architecture Excavator FX/Opteron processors are now delayed probably indefinitely, and may or may not see a release. If it does release, it will probably not see a release until 2015 or 2016 at the earliest if going by how they're delaying the desktop variants by an extra year.
It's a bold if not also risky move by AMD to shift the products around.

As for video game console Radeon-based GPUs-- Wii U, possibly PS4 and Xbox Next/720-- are still on course with whatever their release schedule is at. Wii U is already out, but PS4 and Xbox Next/720 should be out by late 2013 or late 2014 during the Holiday season in either years. That is if both consoles will use Radeon-based GPUs. As for a PS4 and Xbox Next/720 console using an AMD APU based on Piledriver modules or even Steamroller, that's still highly speculative. Both consoles will need the latest PowerISA-based POWER processor from IBM to maintain backwards compatibility if either Microsoft or Sony wish to include it in their next consoles.
 
Dan_D said:
AMD never had the capacity to produce CPUs in the quantities that HP, Dell etc. would have demanded.
Click to expand...

This.

Back in the AMD golden age. AKA Athlon64. Many believed that AMD could become a real contender and steal a good chunk of the pie.

Unfortunately AMD just couldn't keep up with demand and failed to secure deals with the big boys. And of course Intel wouldn't let them.
 
MikeTrike said:
I don't know what rise and fall they are talking about. AMD has always been small and never "equal" to Intel in any respect. I love AMD, but title is a stretch. Intel is about 70x larger than AMD.
Click to expand...

There was a time when AMD surged and Intel faltered, and for that time, they were equals. It didnt last long. Intel came out with killer designs and AMD just couldnt keep up with the giant.
 
Spire3660 said:
There was a time when AMD surged and Intel faltered, and for that time, they were equals. It didnt last long. Intel came out with killer designs and AMD just couldnt keep up with the giant.
Click to expand...

Centauri said:
No, there wasn't... AMD was always playing in Intel's shadow.
Click to expand...

AMD and Intel have never been on equal footing. Even when AMD had the faster performing processors Intel still outsold them and out performed them financially. Only the enthusiasts knew or cared which was the better processor. And even at their best I'd argue that AMD still had chipset and platform shortcomings. Intel's chipsets have always been second to none.
 
The first page is a gigantic mess. It falsely gives the impression that AMD was always nipping at the heels of Intel's products, which was simply not true. There were most often years between Intel's last generation release and when the "equivalent" AMD product was released. Higher speed AMD 286 chips (which were licensed clones of the i286) at higher speeds were competing against similar or faster clock speed i386 chips, for example. I've timelined it before and can post the link to the prior post, if desired.

Overall it's pretty amateurishly written and offers close to zero insight into AMD's many woes, primarily poor business decisions and delays.
 
Intel spends more on R&D then AMD's total sales each year. There was never any way that AMD would be more then a bit player long-term. I'm sure AMD (or their designs) will be around for a long time.

Dell is losing its shorts right now, why don't they merge together?
 
Spire3660 said:
There was a time when AMD surged and Intel faltered, and for that time, they were equals. It didnt last long. Intel came out with killer designs and AMD just couldnt keep up with the giant.
Click to expand...

I see an AMD 10 year peak of 12.36 Billion

For comparison:
I see an Intel 10 year peak of 209.35 Billion

I don't think faltered means the same to you and I when you compare those two numbers... They played in the shadows, they did a good job, but it was still in the shadows.
 
pxc said:
The first page is a gigantic mess. It falsely gives the impression that AMD was always nipping at the heels of Intel's products, which was simply not true. There were most often years between Intel's last generation release and when the "equivalent" AMD product was released. Higher speed AMD 286 chips (which were licensed clones of the i286) at higher speeds were competing against similar or faster clock speed i386 chips, for example. I've timelined it before and can post the link to the prior post, if desired.

Overall it's pretty amateurishly written and offers close to zero insight into AMD's many woes, primarily poor business decisions and delays.
Click to expand...

Agreed and I've said as much. The AMD am386 and am486 CPUs may have been offered at higher clock speeds but they weren't as fast per clock cycle. The difference was small when all was said and done but the AMD CPUs were often released very late in comparison. By the time the am486 DX4 120 came out Intel was selling Pentium CPUs as I recall. The K5 was a disaster. Not only was it late but it's performance was terrible. The PR75+ didn't compate all that well against the Pentium 60/66. The PR90+ matched closer to the Pentium 75MHz at most. And as I recall Intel had pretty much discontinued everything under the Pentium 100MHz at that point.

Cyrix had better offerings than AMD did at the time. Both were weak FPU wise compared to Intel's offerings even if they were faster in a handfull of other benchmarks. K6 was certainly much better but it wasn't quite the match for Pentium II they'd hoped. It was certainly a good value and offered a lot of bang for your buck but it wasn't until the Athlon than AMD had a "Pentium Killer" on their hands. And it wasn't even until the Athlon 64 that AMD had truly distanced themselves from Intel in every performance metric aside from video encoding which was still close most of the time.
 
What I think slowed them down was them buying ATI, but without it I think there wouldn't even be the A-Series CPU. I think they just need to focus more on the A-Series and make a more powerful version of it. It actually performs quite well for the price, especially if you pair it up with a dedicated video card.
 
I wonder which of the rise-and-fall companies will go bankrupt first..

amd, rimm, or nokia
 
Dan_D said:
Cyrix had better offerings than AMD did at the time. Both were weak FPU wise compared to Intel's offerings even if they were faster in a handfull of other benchmarks. K6 was certainly much better but it wasn't quite the match for Pentium II they'd hoped. It was certainly a good value and offered a lot of bang for your buck but it wasn't until the Athlon than AMD had a "Pentium Killer" on their hands. And it wasn't even until the Athlon 64 that AMD had truly distanced themselves from Intel in every performance metric aside from video encoding which was still close most of the time.
Click to expand...

The problem was that AMD was superior to Intel between the Athlon and Athlon 64 chips. That's a wide range of chips, but Intel still sold more then they did. Shouldn't have been the case, since AMD was also cheaper then Intel.

The only reason anyone ever sold an AMD machine was to piss off Intel and drop their prices. When Intel released the Pentium M, it was down hill for AMD. By the time Intel Core chips were released, AMD wasn't relevant anymore.

Without a competitive x86 chip, AMD will be sold off. Without AMD, x86 won't last much longer. As great as Intel chips are, they do have a monopoly on x86. ARM is now a viable alternative to x86.
 
You must log in or register to reply here.
Back
Top