AMD Bulldozer Thread

[ajaidev]

People are really nutty here, why would an employ be put on death row if its employer did something wrong. From what i gather JF is not that up in the food chain to be accountable for BD fiasco.

If it was not so i would have been liable for Intel's P4 problems and so would all other employs. Its never a good idea to interact with consumers directly, people at Intel interact with server level consumers all the time and we have a forum setup specially for that purpose where question on future products and other questions are answered . If normal run of the mill consumers were allowed to enter they would spam the hell out of the forum not to mention the nightmare the online staff will need to handle heheh.

Ideas that AMD has put out again and again like they did with Phenom and BD are wonderful and really progressive. The only problem is execution and the so called expose by Ex- AMD worker is well hogwash. Intel works with automated computer design program all the time, this is how we select a specific layout. Then the design is redone via hand and this is also done via computer assistance "Just imagine a doctor with a laser pointer pen with clickers in-front of a huge screen".

[AliG]

Looking at his comments, I think he's just being passive aggressive about the whole situation. While he does take a bit of responsibility saying that he was just wrong, he also makes it seem that its the fault of the consumer for being upset with them. That's just rediculous, you should expect a lot of angry investors during the launch of a bad product if you're the horse's mouth - that's just how it works. And yes people who waited forever to get a BD build over SB are investors as well, as they invested a lot of time without an upgrade not to mention probably spent a lot of energy trying to justify their waiting.

I'm willing to bet he'll come back. It's like a massive fight in a relationship, lots of yelling, but in most cases (especially if it's just the first one) both sides will want to reconcile. Give him some time, my guess is he'll apologize for "unknowingly" (we'll never know if this is true or not, just have to trust him) misleading people and forgive the haters.

[demonkevy666]

If anything JF should be apologizing to the people here for all the delays and the subpar performance. That would be the honorable thing to do. We were promised by AMD very specific increases that we absolutely did not see; if anything he specifically earned all the venting as he is in charge of marketing (regardless of server or desktop, BD doesn't have a 50% increase of thuban at same clocks)

NO he should not need to, it is not his fault for getting crap info out of an engineer.

[Sgt.McRuff]

And don't forget the fact he kept saying this is for server products, yet everyone wants to yell at him for desktop client side.

[Evantaur]

kinda like the video on youtube of interlagos and people are yelling "LOL FAILDOZER!"

[AliG]

NO he should not need to, it is not his fault for getting crap info out of an engineer.

*sigh*

This is the problem with our generation, there's no sense of responsibility. You take care of your actions, regardless of why they occurred. So 1 engineer gave him bad info, that's his fault for not checking up on it with the rest of the team. Not to mention he could have just looked at the data himself instead of listening to what the guys designing it are saying. Would you ever tell your boss you are designing a product that will flop? Of course not.

JF mislead the people here, whether intentional or not. As head of marketing it's his responsibility (as he gets paid lots of money to do this) to make sure all information he puts out there is accurate. When you say IPC increases with a sense of authority, then clearly you are insinuating to the people that you know something from your position and it is no longer just an opinion of yours.

[BoredByLife]

*sigh*

This is the problem with our generation, there's no sense of responsibility. You take care of your actions, regardless of why they occurred. So 1 engineer gave him bad info, that's his fault for not checking up on it with the rest of the team. Not to mention he could have just looked at the data himself instead of listening to what the guys designing it are saying. Would you ever tell your boss you are designing a product that will flop? Of course not.

JF mislead the people here, whether intentional or not. As head of marketing it's his responsibility (as he gets paid lots of money to do this) to make sure all information he puts out there is accurate. When you say IPC increases with a sense of authority, then clearly you are insinuating to the people that you know something from your position and it is no longer just an opinion of yours.

I tend to agree with you. But still it doesn't justify sending hatemails. Hatemails is just an animal like approach to dealing with something not meeting your expectations. Sure, you can be mislead, but being destructive about it is not very rational behaviour...

[AliG]

Of course my good chap. I never said anywhere that people should be allowed to send hatred. I'm just saying he's not being a man about this situation and should have expected backlash from the weaker minded. I would have a lot more respect for the guy if he would make a few statements here apologizing about performance, then i'm almost positive the forum would have self-moderated those causing trouble by making it socially unacceptable.

[verndewd]

Phenom I was crap. Phenom II was actually pretty good.

History repeating itself once again it seems... piledriver should clean up this mess.

BD 2 will clear it up and likely windows 8.

[verndewd]

Of course my good chap. I never said anywhere that people should be allowed to send hatred. I'm just saying he's not being a man about this situation and should have expected backlash from the weaker minded. I would have a lot more respect for the guy if he would make a few statements here apologizing about performance, then i'm almost positive the forum would have self-moderated those causing trouble by making it socially unacceptable.

wassup ali ? hope all is well.

I like the 8 core idea I just think a revision or two is needed like the whole phenom thing.

people here tend to whine a bit about whos beating who but AMD makes chips that get the job done for less.
Most pc users dont care about fps as much as a smooth and relatively fast experience. its not even worth mentioning here that amd's release dates are investor pressured . that cant be fun in any way, definitely has lost them points in the geek community.

that and HMK on SOI was a freaking nightmare for a while. add to that 32nm. not an easy task.

id buy one if i had money to burn, then replace it in a year when better revisions were out.

[BoredByLife]

http://semiaccurate.com/2011/10/17/w...er-underwhelm/

http://semiaccurate.com/2011/10/17/b...ingle-problem/

Interesting read

[ajaidev]

http://semiaccurate.com/2011/10/17/w...er-underwhelm/

http://semiaccurate.com/2011/10/17/b...ingle-problem/

Interesting read

Stop posting the same thing in both the AMD BD threads...

[BoredByLife]

Stop posting the same thing in both the AMD BD threads...

They are relevant in both threads... What are you so mad about?

[freeloader]

JF clearly commented that IPC would increase on BD and we all know what happened with that comment. Stop sticking up for him, he made mistakes and I'm sure he's learned from them. Let's just all move forward now and put this whole fiasco behind us and look forward to BD's variants.

[Mechromancer]

http://muropaketti.com/artikkelit/pr...8150-zambezi,5

X-2011101209080269034.jpg

X-2011101209080468738.jpg

AMD is a little too ahead of the times with this design. It seems we have to wait for Windows 8 for things to improve a little. Still, it is a server design first and desktop second, but not a complete failure. I guess most of us are butthurt because it isn't the WTFPWNASAURUS for gaming we hoped it would be. The AMD platform is still good and cost effective overall. Maybe we can look towards Trinity (designed for desktop) with a little hope now.

[AliG]

I mean, AMD clearly advertised in their slides that it would offer "similar" to 970x performance in gaming. I think that's fair reason for the consumer to actually expect it to be there

[verndewd]

http://muropaketti.com/artikkelit/pr...8150-zambezi,5

X-2011101209080269034.jpg

X-2011101209080468738.jpg

AMD is a little too ahead of the times with this design. It seems we have to wait for Windows 8 for things to improve a little. Still, it is a server design first and desktop second, but not a complete failure. I guess most of us are butthurt because it isn't the WTFPWNASAURUS for gaming we hoped it would be. The AMD platform is still good and cost effective overall. Maybe we can look towards Trinity (designed for desktop) with a little hope now.

thats not saying much for an 8 core. small gains.

[STaRGaZeR]

AMD is a little too ahead of the times with this design.

You mean AMD just launched a CPU that needs specialized code not to suck, right?

Usually new CPUs need specialized code to shine instead. Big difference.

[gosh]

JF clearly commented that IPC would increase on BD and we all know what happened with that comment. Stop sticking up for him, he made mistakes and I'm sure he's learned from them. Let's just all move forward now and put this whole fiasco behind us and look forward to BD's variants.

If there is a problem with L1 cache writes and they fix this then it might be increased IPC

It also depends on what type of software that is executed, server software might work better on bulldozer compared to synthetic tests on desktops

[AliG]

Now that might be a legitimate issue, perhaps transistor density played into it? I can see process having a big influence there

[verndewd]

You mean AMD just launched a CPU that needs specialized code not to suck, right?

Usually new CPUs need specialized code to shine instead. Big difference.

no it needs a specialized code to suck less

[verndewd]

Now that might be a legitimate issue, perhaps transistor density played into it? I can see process having a big influence there

it could be a deposition issue. I remember reporting on HMK SOI compatibility issues at xcpus suck.com. I also remember JJ showing a cross section of an intel chip and the comparison to amd chips was pretty striking. AMD looked like lumps of metal and intels looked neat and clean. also another issue could be soi itself as the nodes lower soi becomes problematic.
Tose possibilities aside Transistor density wouldnt account for that I think, other than the possibility of requiring more latency to process.

[jogshy]

A question....

Can you disable Turbo Core to force the CPU to work at a fixed frequency independently of the workload? I suspect if you make that and you disable all the energy-saving options like C1E you'll notice a great speed boost.

[eRazorzEDGE]

A question....

Can you disable Turbo Core to force the CPU to work at a fixed frequency independently of the workload? I suspect if you make that and you disable all the energy-saving options like C1E you'll notice a great speed boost.

Yes you can. Visit the AMD section to see what we're all up to with these chips.

[DilTech]

thats not saying much for an 8 core. small gains.

It's not quite an 8 core. This is more or less AMD's stab at hyper-threading. It's essentially 4 full cores that can handle 8 threads. A few review sites touched on that fact, but most ignored it.