[ Bloomberg ] AMD Facing Bleak Future

[PPB]

Zen and K12 will be in whatever node GF has for them at their given mass production schedule. So whoever still wants competition in the x86 cpu space, light some candles for GF so they dont mess up yet again.

The stock marketers, PR disguised as forum posters and the everyday AMD haters can hold very tight to Kumar's words like they have been doing just recently, and hope their doom and gloom predictions in regards to AMD will finally come to fruition. Although those predictions have been missing the mark for so many years that it is hard for me to have a reason to believe them this time.

 

[Enigmoid]

Llano and it sucked.

Llano had process problems. A straight up port to 32 nm would have been quite good. The is no real reason for the CPU which had no problems at 45 nm 3.6 ghz to have trouble at 32 nm hitting the same frequencies.

One of the problems was the the igp bloated the die size and hurt cache. Llano has a 1 MB L2 per core arrangement which hurt vs. 4 MB shared of Trinity or the 512 KB/core + 6 MB shared of PII.

 

[cbn]

I have to just wonder what an evolution of Phenom II x 4/x6 would have been like on 32nm? Xtors capable of 5 Ghz coupled to a move to four wide front end (up from three wide) probably would have been pretty exciting. (With any luck we will see a return of this evolution).

Llano and it sucked.

I think in some ways that might have been a bit of a regression (rather than evolution) though. As I recall Llano lost the L3 cache of the Phenom II and the 32nm node was very new at the time. Furthermore, the front end stayed three-wide.

 

[ThatBuzzkiller]

Is this supposed to be news ?

I saw this coming from almost a decade ago when Intel released Conroe ...

I can only recount 3 good years from AMD out of their 45 years in the x86 microprocessor design industry ...

 

[Idontcare]

Is this supposed to be news ?

I saw this coming from almost a decade ago when Intel released Conroe ...

I can only recount 3 good years from AMD out of their 45 years in the x86 microprocessor design industry ...

3? C'mon man, there was way way way more than that going on for AMD.

The 486 processors they fielded had more cache than Intel, and they pushed the 486 platform past 100MHz. I owned a 133MHz and a friend of mine bought one of their 150MHz 486 chips. For the price they were a really good bargain.

And the K6? K6-2? K6-3? Those were great. Owned them as well, did superbly well.

But they really hit their stride with the K7. That was AMD's "conroe'd the market" moment. In fact when Intel released the K7, Anand really should have dubbed it "Intel just K7'ed the market" IMO.

And the K8 only got better from there.

AMD only started dying on the vine with the appointment of Hector as CEO and the subsequent disasterous "let's delay 65nm development, and Bulldozer, so we can milk the market more with our 90nm X2 chips" decision.

Then came Conroe and they had no backup plan or cash left (after their ATI acquisition, thanks Hector) to launch a backup plan the likes of Intel.

So here they are, screwed and on CEO #4 since Sanders left the helm of what was at the time a very healthily positioned company.

 

[ShintaiDK]

I miss Sanders. I wonder how he feels today after seeing one disaster after the other on the company he buildt. I hope he can displace himself from it, else it must be depressive for him to watch his lifework on a downhill spiral.

 

[Arachnotronic]

3? C'mon man, there was way way way more than that going on for AMD.

The 486 processors they fielded had more cache than Intel, and they pushed the 486 platform past 100MHz. I owned a 133MHz and a friend of mine bought one of their 150MHz 486 chips. For the price they were a really good bargain.

And the K6? K6-2? K6-3? Those were great. Owned them as well, did superbly well.

But they really hit their stride with the K7. That was AMD's "conroe'd the market" moment. In fact when Intel released the K7, Anand really should have dubbed it "Intel just K7'ed the market" IMO.

And the K8 only got better from there.

AMD only started dying on the vine with the appointment of Hector as CEO and the subsequent disasterous "let's delay 65nm development, and Bulldozer, so we can milk the market more with our 90nm X2 chips" decision.

Then came Conroe and they had no backup plan or cash left (after their ATI acquisition, thanks Hector) to launch a backup plan the likes of Intel.

So here they are, screwed and on CEO #4 since Sanders left the helm of what was at the time a very healthily positioned company.

Ah, Idontcare, you're making me all nostalgic

I had a K5, a K6-2 450MHz, and when I was really lucky, I got a hand-me-down K6-2 550MHz!

I also had an Athlon XP 2200+. The unstable Via chipset was a drag, but the CPU itself was pretty phenomenal.

 

[cytg111]

AMD 486 DX-4 100MHz .. was screaming back then... the K6-II with 3DNow, only game i played back then was Quake and Quake2 and with the 3DNow patch it was even giving the ppro a run for its money.

 

[Dribble]

I remember getting an old work machine no longer required with some Intel 150Mhz cpu but lots of memory and buying an adapter that let me put a K6-2 450Mhz in, added a voodoo 3 2000 and I was away. Great value for what it was.

 

[blake0812]

I sure hope they don't fail. Competition is always good.

 

[ThatBuzzkiller]

3? C'mon man, there was way way way more than that going on for AMD.

The 486 processors they fielded had more cache than Intel, and they pushed the 486 platform past 100MHz. I owned a 133MHz and a friend of mine bought one of their 150MHz 486 chips. For the price they were a really good bargain.

And the K6? K6-2? K6-3? Those were great. Owned them as well, did superbly well.

But they really hit their stride with the K7. That was AMD's "conroe'd the market" moment. In fact when Intel released the K7, Anand really should have dubbed it "Intel just K7'ed the market" IMO.

And the K8 only got better from there.

AMD only started dying on the vine with the appointment of Hector as CEO and the subsequent disasterous "let's delay 65nm development, and Bulldozer, so we can milk the market more with our 90nm X2 chips" decision.

Then came Conroe and they had no backup plan or cash left (after their ATI acquisition, thanks Hector) to launch a backup plan the likes of Intel.

So here they are, screwed and on CEO #4 since Sanders left the helm of what was at the time a very healthily positioned company.

There really was only 3 good years that I can only remember that AMD had and they were particularly the K8 years.

Their version of the 486 was indeed superior but that didn't matter when Intel had higher performance processors at the same time so the only ones that would look to AMD were OEMs looking to fill in for the low end.

Hector was indeed an issue but so are the AMD engineers in the CPU division ...

This was coming for a long time ...

 

[mrmt]

Zen and K12 will be in whatever node GF has for them at their given mass production schedule. So whoever still wants competition in the x86 cpu space, light some candles for GF so they dont mess up yet again.

No need for candles, 28nm seems to be fine.

 

[Arachnotronic]

Hector was indeed an issue but so are the AMD engineers in the CPU division ...

This was coming for a long time ...

Don't blame the engineers; they're just designing what they're told to design under a given budget/time constraint.

 

[ShintaiDK]

Don't blame the engineers; they're just designing what they're told to design under a given budget/time constraint.

The man on the floor yes, managers etc, no.

 

[Genx87]

This has been a long time in the making. AMD hasn't had a competitive high end processor since Conroe showed up nearly 8.5 years ago.

 

[Erenhardt]

This has been a long time in the making. AMD hasn't had a competitive high end processor since Conroe showed up nearly 8.5 years ago.

Their latest CPU was very competitive in terms of performance:

Or did you just evaluate CPU performance based on poorly performing software (games)?

I would also like to see how the beefed up cat cores perform.

here is IPC comparision of kaveri and piledriver to the jaguar kabini:

Basically a tie.

"We know from previous AMD disclosures that one Jaguar CPU core is 3.1mm sq. Our best estimates put Steamroller at 9.6mm sq, or more than 3x the size of a Jaguar CPU. Neither of these estimates include cache, which would tilt the equation still farther towards Kabini"

http://www.extremetech.com/computin...o-take-a-page-from-intel-and-dump-steamroller

Going wide and increasing frequency could make it a quite potent desktop chip - maybe the best in whole amd's lineup.

 

[III-V]

Their latest CPU was very competitive in terms of performance:


Or did you just evaluate CPU performance based on poorly performing software (games)?

I would also like to see how the beefed up cat cores perform.

here is IPC comparision of kaveri and piledriver to the jaguar kabini:


Basically a tie.

http://www.extremetech.com/computin...o-take-a-page-from-intel-and-dump-steamroller

Going wide and increasing frequency could make it a quite potent desktop chip - maybe the best in whole amd's lineup.

No, it's not competitive. Stop cherry picking already.

http://anandtech.com/bench/product/551?vs=697

Welcome to the real world.

Going wide and increasing frequency could make it a quite potent desktop chip - maybe the best in whole amd's lineup.

Yeah, because it's that simple.

 

[mrmt]

Or did you just evaluate CPU performance based on poorly performing software (games)?

Why should game developers bother to parallelize tasks if they run very well on over 85% of the processors sold?

 

[Abwx]

No, it's not competitive. Stop cherry picking already.

http://anandtech.com/bench/product/551?vs=697

Welcome to the real world.

Yeah, because it's that simple.

Are theses benchs realworld.??.

http://www.hardware.fr/focus/99/amd-fx-8370e-fx-8-coeurs-95-watts-test.html

Because a serie of 7 Sysmarks benchs is not what i would call real world, but i guess that some virtual world is necessary to inflate numbers thanks to Bapco/Intel.

 

[richierich1212]

AMD 486 DX-4 100MHz .. was screaming back then... the K6-II with 3DNow, only game i played back then was Quake and Quake2 and with the 3DNow patch it was even giving the ppro a run for its money.

Nostalgia, *wipes away tear*

 

[Erenhardt]

Stop cherry picking already.

Cherry picking? If someone thinks any software is optimized for amd cpus and crippling intel's performance then he is delusional.

AMD cpus are competitive in some cases:
Connecting the dots - AMD cpus have very competitive performance potential.

Going your logic:
Is RS4 faster than fiesta? Nope, both run 50km/h in city. RS4 is slower in 0-50 if you don't use full throttle.

Agree or disagree and move on. Stop continuing offtopic, discussing old tech.

 

[III-V]

Are theses benchs realworld.??.

Some of them are, yes. What is the problem with synthetic benchmarks, exactly?

Because a serie of 7 Sysmarks benchs is not what i would call real world, but i guess that some virtual world is necessary to inflate numbers thanks to Bapco/Intel.

Right. Because those are the only benchmarks ran in what I linked. You too, are living in an alternate reality.

 

[III-V]

Cherry picking? If someone thinks any software is optimized for amd cpus and crippling intel's performance then he is delusional.

Who cares? Why would anyone bother to optimize for AMD over Intel? It'd be illogical to do so.

AMD cpus are competitive in some cases:
Connecting the dots - AMD cpus have very competitive performance potential.

What? You've yet to prove this.

Going your logic:
Is RS4 faster than fiesta? Nope, both run 50km/h in city. RS4 is slower in 0-50 if you don't use full throttle.

Agree or disagree and move on. Stop continuing offtopic, discussing old tech.

Oh, okay. Let me just drop a bunch of garbage, and cower off after I'm proven wrong. To seal the deal, I'll brand anyone's responses to my argument as being off-topic.

 

[exar333]

Cut R&D budget - check
Large

Spoiler

layoffs

cost-cutting effort - check
Abandon cutting-edge node/processes - check

Only the absolute most hardcore AMD cheerleader will try to spin this in a positive way. I love AMD; they have been a great alternative and (at times) market-leaders with their offerings over the years. The 28nm announcement is scary and I would not be surprised to see AMD sell-off the graphics division and/or patents-IP now. They can cross-license what they already have and just ride those patents through console revenue for another 6-7 years. Maybe throw-in some ARM designs as well (on old tech), who needs anything else?

There has been a lot of 'the sky is falling' over the years, but this is scariest, by far to me. The writing is now [officially] on the wall.

 

[Abwx]

Some of them are, yes. What is the problem with synthetic benchmarks, exactly?

Right. Because those are the only benchmarks ran in what I linked. You too, are living in an alternate reality.

Ok, let s look at the benches you linked, let s start with Agisoft PhotoScan Benchmark wich is used 6 times with 4 subscores that are averaged for a total score, that s a mean to extract five favourable benches out of a single one.

Then let s look at the Mthreaded apps wich are also tested in single thread like CBench, do you use Cinema 4D or any other rendering engine in single thread.?
No but that s yet another mean to put a favourable although irrelevant bar.

Then we can finish with my favourite, 3D Particle movement.
Is that a real world bench.?
Dont know but what is sure is that Intel CPUs run up to SSE3 while the AMDs are stuck in X87, prove is that my Athlon XP, wich has no SSE2, run this bench with better ST IPC than the current AMDs CPUS.

What is left of this glorious serie of benches.?.

Cinebench MT, Winrar 5.01, H264, Agitsoft for the total score and so on, all bench that mirrors the ones i linked at hardware.fr...