AMD manager speaks about Bulldozer, admits failure

[boxleitnerb]

Andrew Feldman, Corporate Vice President and General Manager of AMDs server division made some interesting comments about how AMD views Bulldozer in retrospect:

Feldman candidly acknowledged that the Bulldozer failure cost AMD some credibility in servers. But the company is looking to rebound with a revamped management team led by CEO Rory Read and a new server roadmap comprising x86 and ARM chips for multiple server categories.

“Bulldozer was without doubt an unmitigated failure. We know it,” Feldman said.

“It cost the CEO his job, it cost most of the management team its job, it cost the vice president of engineering his job. You have a new team. We are crystal clear that that sort of failure is unacceptable,” Feldman said.

AMD bridges road to ARM with new low-power x86 server chips | PCWorld

I honestly didn't expect that. How goes the saying? A fault confessed is half redressed. I like the honesty, it's truly refreshing and gives me hope that AMD is on the right track.

 

[Deleted member 4644]

F servers.... what about Gammmmesssss!

 

[mrmt]

Oh dear, what's left for the fanbois now?

Maybe Mr. Feldman is just an Intel shill.

Seriously. You guys can only go 2 posts before trolling? Knock it off.
-ViRGE

 

[Rvenger]

Oh dear, what's left for the fanbois now?

At least they still have moar cores!

 

[Ventanni]

Maybe they'll divulge some performance details about Steamroller along with their honesty?

 

[mrmt]

You have to ask what kind of future the products related to that "unmitigated failure" have inside the company. If anyone had any doubt that AMD will shift away from Bulldozer-related products in the future, Mr. Feldman could not be clear enough.

 

[mrmt]

Maybe they'll divulge some performance details about Steamroller along with their honesty?

I don't think "unmitigated failure" leaves much hope for Steamroller, Excavator, Concrete Mixer or any other chip from this family.

 

[Ancalagon44]

As I've said before, Steaming Pile of Bull....

Its good that AMD admits it, but what is the AMD Defense Force going to say? I dont think they will like this.

 

[boxleitnerb]

I don't think "unmitigated failure" leaves much hope for Steamroller, Excavator, Concrete Mixer or any other chip from this family.

I wouldn't go so far. The design is very new, there are many spots where it can be tweaked. Combined with a new 28nm process, Steamroller could be really competitive.

 

[Rvenger]

I don't think "unmitigated failure" leaves much hope for Steamroller, Excavator, Concrete Mixer or any other chip from this family.

I think acknowledging it was a step in the right direction actually and I think they will prevent another failure from happening such as BD again. PD was greatly improved so it can only get better for AMD. I want to see PCI-E 3.0 support and an updated chipset to start.


Also, I really don't think AMD wanted to release BD at all.. they knew from the start it was a piece of garbage.. they just didn't have anything else to release.

 

[Abwx]

I don't think "unmitigated failure" leaves much hope for Steamroller, Excavator, Concrete Mixer or any other chip from this family.

How to inflate the sayings...

 

[inf64]

I wouldn't go so far. The design is very new, there are many spots where it can be tweaked. Combined with a new 28nm process, Steamroller could be really competitive.

He(poster) is just dreaming. The big core will be going forward and it will be 15h+ based. After XC core it's possible we have some sort of convergence of families ( like ARM based on one side and some in-between big-small concept on x86 side). SR is launching and will be in server segment as APU first, similar goes for Excavator.

 

[mrmt]

I think acknowledging it was a step in the right direction actually and I think they will prevent another failure from happening such as BD again. PD was greatly improved so it can only get better for AMD. I want to see PCI-E 3.0 support and an updated chipset to start.

As I said, I don't think "unmitigated failure" leaves much room for anything. He essentially said that the entire Bulldozer family is FUBAR and specifically in servers, it will be scrapped.

 

[mrmt]

He(poster) is just dreaming. The big core will be going forward and it will be 15h+ based. After XC core it's possible we have some sort of convergence of families ( like ARM based on one side and some in-between big-small concept on x86 side). SR is launching and will be in server segment as APU first, similar goes for Excavator.

Do you still expect future products from an "unmitigated failure"?

 

[sushiwarrior]

I don't think "unmitigated failure" leaves much hope for Steamroller, Excavator, Concrete Mixer or any other chip from this family.

Bulldozer suffers from poor cache performance and poor FP performance, both of which are not architecture problems but rather implementation issues. I have no doubt Steamroller or Excavator could fix either of these issues (especially based on those leaked die shots) and give pretty big performance leaps.

 

[guskline]

I think acknowledging it was a step in the right direction actually and I think they will prevent another failure from happening such as BD again. PD was greatly improved so it can only get better for AMD. I want to see PCI-E 3.0 support and an updated chipset to start.


Also, I really don't think AMD wanted to release BD at all.. they knew from the start it was a piece of garbage.. they just didn't have anything else to release.

Excellent point Revenger. Though I wish it wasn't so, I doubt my AM3+ mbs will be the preferred chipsets for SteamRoller, if it does get released.

My hunch is the SteamRoller core will be released with a new socket and chipset much the same way as Intel is doing with Haswell.

 

[boxleitnerb]

As I said, I don't think "unmitigated failure" leaves much room for anything. He essentially said that the entire Bulldozer family is FUBAR and specifically in servers, it will be scrapped.

Bulldozer is the first installment only. He doesn't say anything about completely abandoning the concept of Bulldozer or that there is no hope for improvement.

 

[mrmt]

Oh dear, not even AMD management calling Bulldozer an "unmitigated failure" can demolish the convictions of the Failthful that Bulldozer will some day shine, and that the company will still invest on the thing.

Those will be the guys calling out AMD for the huge mistake in scrapping the unmitigated failure for good.

 

[guskline]

Bulldozer is the first installment only. He doesn't say anything about completely abandoning the concept of Bulldozer or that there is no hope for improvement.

Maybe so but why do I now hear "Dandy Don Meridith" singing "turn out the lights, the party's over"? (A classic on Monday night football!)

 

[mrmt]

Bulldozer is the first installment only. He doesn't say anything about completely abandoning the concept of Bulldozer or that there is no hope for improvement.

And obviously you would fire the CEO, the engineering VP, the engineering team and most of your senior management because the first installment of your wonderchip didn't work, but the second and the third still had very good prospects, right? Why fire the old management if they had a winner on their hands?

 

[Abwx]

As I said, I don't think "unmitigated failure" leaves much room for anything. He essentially said that the entire Bulldozer family is FUBAR and specifically in servers, it will be scrapped.

Moar inflation...
Now mrmt switched from financial expert to uarch expert.

I guess the OP did throw a welcomed bone for the usual
AMD downplayers...

 

[Ancalagon44]

Oh dear, not even AMD management calling Bulldozer an "unmitigated failure" can demolish the convictions of the Failthful that Bulldozer will some day shine, and that the company will still invest on the thing.

Those will be the guys calling out AMD for the huge mistake in scrapping the unmitigated failure for good.

It will be like the Pentium 4 - in a few years time, those who supported it will be too embarrassed to admit it.

I just love the fact that, to fix BD, they end up undoing some of the changes, making an SR core more like a Phenom II core.

 

[Exophase]

Admitting Bulldozer was a failure doesn't mean that its successors are viewed as dead-ends within the company, or that Steamroller is now cancelled. We don't know that he's referring to anything more than the product launched late 2011 launch.

Easy to confess a fault made by the company before they acquired you. What's more surprising is that AMD let him. I bet he wanted to say it before the acquisition was even on the table, but was more afraid of offending his supplier than he is now of offending the board of directors.

 

[Ajay]

I wouldn't go so far. The design is very new, there are many spots where it can be tweaked. Combined with a new 28nm process, Steamroller could be really competitive.

I wouldn't go that far I still think AMD will offer an SR based Opteron for tradition markets. Since Kaveri is designed, it would have only required and incremental cost to create an Opteron (though I don't know how much server verification adds to the development costs). AMD can probably beat Intel on price/performance, but I don't see how they can even match Intel on power/performance, if the recently "released" Kaveri die shot are real, and with IVB-EP coming online later this year.

 

[boxleitnerb]

And obviously you would fire the CEO, the engineering VP, the engineering team and most of your senior management because the first installment of your wonderchip didn't work, but the second and the third still had very good prospects, right? Why fire the old management if they had a winner on their hands?

You don't need to exaggerate. The risk was that they screw it up a second time in the future, so they were let go. It doesn't necessarily say anything about the future prospects of a product, though.