Rumour: Bulldozer 50% Faster than Core i7 and Phenom II.

[AtenRa]

Anyway the ISSCC disclosures brought some information and they are from engineers no private talk from marketing guys.

From ISSCC we know the following:
Core performance drops by 10%. Overall performance increases by ~60%.
All in regard to current AMD cores.

Could you please be so kind and give us a link to support the above statement??

ps: except of the following BS
http://www.eetimes.com/electronics-...na-eyes-petaflops-IBM-hits-5-GHz?pageNumber=2

 

[Mopetar]

AMD are marketing gods!

They were able to create a vaporware product that has been talked about, discussed, disected, argued, etc, all over the internet for just about 3 years now. Now that is the most impressive thing about BD.

Don't know if you can really refer to it as vaporware as they have a moderately definitive shipping time established. If we'd already blown past that date by several months and AMD were saying, "Next year! Yeah that's right." you might have a more compelling argument, but right now it's pretty much in the 'unreleased product' category. I think they'd originally hoped to have it out much sooner, but adjusted their estimates long before it was closing in on the release date. Some time adjustments may have also dealt with fab companies being slow to get their new processes running. For instance, TSMC abandoned their 32 nm process which threw off a lot of time tables for several companies.

 

[Elixer]

We get higher IPC via the flux capacitor.

Now it all makes sense, AMD is using alien technology!

 

[Lonbjerg]

Don't know if you can really refer to it as vaporware as they have a moderately definitive shipping time established. If we'd already blown past that date by several months and AMD were saying, "Next year! Yeah that's right." you might have a more compelling argument, but right now it's pretty much in the 'unreleased product' category. I think they'd originally hoped to have it out much sooner, but adjusted their estimates long before it was closing in on the release date. Some time adjustments may have also dealt with fab companies being slow to get their new processes running. For instance, TSMC abandoned their 32 nm process which threw off a lot of time tables for several companies.

http://redirectingat.com/?id=673X542464&xs=1&url=http%3A%2F%2Fdownload.amd.com%2FCorporate%2FMarioRivasDec2007AMDAnalystDay.pdf&sref=http%3A%2F%2Faceshardware.freeforums.org%2Famd-bulldozer-sandtiger-delayed-vanished-from-2009-roadmap-t323.html

Page 21....

Or this:
http://enthusiast.hardocp.com/image.html?image=MTE4NTQ1OTQ5ME50OVZheTYzSTFfMV8xM19sLmpwZw==

 

[drizek]

So it seems like they replaced Sandtiger with Thuban/Magnycours.

That doesn't look like Bulldozer, but I could be wrong.

All I know is that if I read about what AMDs plans were for 2014 right now, I would take them with a giant grain of salt.

 

[Lonbjerg]

So it seems like they replaced Sandtiger with Thuban/Magnycours.

That doesn't look like Bulldozer, but I could be wrong.

All I know is that if I read about what AMDs plans were for 2014 right now, I would take them with a giant grain of salt.

http://www.dailytech.com/Bulldozing+the+Comptetion+in+2009/article8204.htm

Sandtiger’s eight cores consist of eight AMD Bulldozers. Bulldozer is the name AMD has given to one of the CPU cores for its M-SPACE architecture

 

[drizek]

lol, so they're a couple of years late.

"Expect AMD to introduce Fusion designs based on Bulldozer and Bobcat beginning in 2009."

 

[Lonbjerg]

lol, so they're a couple of years late.

"Expect AMD to introduce Fusion designs based on Bulldozer and Bobcat beginning in 2009."

Yup, to put this in perspective...if this was Intel (tick-tock) their Ivy Leauge wouldn't be a 2012 part...but a 2014 part.

 

[drizek]

But was Sandtiger supposed to be 32nm?


AMDs position right now is pretty crappy. I don't think they ever planned on having Phenom II competing with Sandy Bridge. They had a lot of problems over the past 4 years.

 

[AtenRa]

Can anyone tell me why we dont have the BD presentations from ISSCC yet ?? I cant find a single link

 

[IntelUser2000]

Maybe they aren't ready to release everything yet. I don't know if they got more to release since they showed so much info last year. AMD blog says we'll see more over this week though.

Could you please be so kind and give us a link to support the above statement??

ps: except of the following BS
http://www.eetimes.com/electronics-n...z?pageNumber=2

Probably not the thing you are looking for, but there's a slide by AMD saying 16 core Interlagos will be 50% faster in SpecInt_Rate and 80% faster in SpecFP_Rate compared to 12 core Magny Cours.

 

[Dresdenboy]

Now it all makes sense, AMD is using alien technology!

Could you please be so kind and give us a link to support the above statement??

ps: except of the following BS
http://www.eetimes.com/electronics-...na-eyes-petaflops-IBM-hits-5-GHz?pageNumber=2

 

[Dresdenboy]

Can anyone tell me why we dont have the BD presentations from ISSCC yet ?? I cant find a single link

http://pc.watch.impress.co.jp/docs/news/event/20110223_428720.html

 

[AtenRa]

Ahh nice,

thx a lot

 

[Phynaz]

lol, so they're a couple of years late.

"Expect AMD to introduce Fusion designs based on Bulldozer and Bobcat beginning in 2009."

As was K10.

It seems every AMD cpu is two years late.

 

[Edrick]

Don't know if you can really refer to it as vaporware as they have a moderately definitive shipping time established. If we'd already blown past that date by several months and AMD were saying, "Next year! Yeah that's right." you might have a more compelling argument, but right now it's pretty much in the 'unreleased product' category. I think they'd originally hoped to have it out much sooner, but adjusted their estimates long before it was closing in on the release date. Some time adjustments may have also dealt with fab companies being slow to get their new processes running. For instance, TSMC abandoned their 32 nm process which threw off a lot of time tables for several companies.

I call Nvidia's Fermi of 2010 delayed. I call AMDs bulldozer vaporware from 2008 until now. Big difference.

I know that BD is real now and will come out. But all we have been shown for the past years is powerpoint slides. And I am faily certain 2009 was the original release date. We all know now that was never even possible. AMD has just been using these slides (and leaks) to try and fight the Nehalam/Westmere/SB chips that Intel has been releasing while we wait for BD.

Again, for the record, I really do hope BD is great. Competition in the market is great for consumers, and it forces technology advances. So I am rooting for AMD here. I just wish it was release last year, because if it were, we may have s2011 already available.

 

[Idontcare]

As was K10.

It seems every AMD cpu is two years late.

Not too unlike Intel and Itanium or Larrabee, or SUN and their SPARC chips, or IBM, or Fujistu...

There's really only one example we can cite for on-schedule releases and this Intel and even then we have to restrict the example to just their x86-based chips.

Which would kinda make the example more an example of the exception to the rule rather than everyone else's schedule slips being the exceptions.

We all know of the oft cited "Project Triangle" from project management where you can only pick two of three options:

You can make the product quickly (Fast) and have high performance (Good) but it will require a healthy R&D budget (won't be Cheap to design). <- Intel has this option, not many others in the industry do.

Or you can make the product have high performance (Good) on a shoe-string R&D budget (Cheap to design) but it is going to take a while to develop. <- this is about the only option for AMD, as well as many other IC designers

No one get's all three, not even Intel. There is a reason their R&D budget is some 4x more than AMD's.

 

[drizek]

Intel was slipping release dates up until 5 years ago with x86, before they implemented tick tock.

Did Yonah come out on time?

I know that BD is real now and will come out. But all we have been shown for the past years is powerpoint slides. And I am faily certain 2009 was the original release date. We all know now that was never even possible. AMD has just been using these slides (and leaks) to try and fight the Nehalam/Westmere/SB chips that Intel has been releasing while we wait for BD.

What if AMD announced back in 2007 that it will be using the Phenom X4 to compete with Intel in 2011? I would say that on the day of the Nehalem release, AMD would have gone bust.

 

[bryanW1995]

AMD are marketing gods!

They were able to create a vaporware product that has been talked about, discussed, disected, argued, etc, all over the internet for just about 3 years now. Now that is the most impressive thing about BD.

duke nukem forever baby!!!

 

[Vdubchaos]

Last time I even remotely considered AMD CPU was back in Athlon days (Winchester/Barton and Opteron). Back when AMD was actually EXTREMELY competitive in the gaming market.

Those days are LONG gone. With CPUs and GPUs to be honest.

They no longer compete price wise on the GPU end, Nvidia and AMD are pretty level performance/Price wise.

On the CPU front, they are WAY behind (at least in gaming). Every since Conroe showed up, it seems like AMD just gave up.

I would love to buy AMD Cpu at some point, but they need to deliver HARD and FAST.(I don't see it)

 

[bryanW1995]

on the gpu end they've been very strong for the past 18 mos or so, though much of that was due to NV miscues. on the cpu side, intel has been dominant for 5 years, though at least some of that has been due to AMD miscues. If AMD delivers this round then we'll have a much more competitive environment, not lease because of intel's recent stumble.

 

[sangyup81]

pricing, pricing, pricing!

 

[JFAMD]

Maybe they aren't ready to release everything yet. I don't know if they got more to release since they showed so much info last year. AMD blog says we'll see more over this week though.



Probably not the thing you are looking for, but there's a slide by AMD saying 16 core Interlagos will be 50% faster in SpecInt_Rate and 80% faster in SpecFP_Rate compared to 12 core Magny Cours.

I didn't approve those numbers, not sure where they came from.

Link?

 

[Idontcare]

I didn't approve those numbers, not sure where they came from.

Link?

http://forums.anandtech.com/showpost.php?p=29006492&postcount=401

And the posts that follow from there in the thread.

(edit1: it's slide 9 from Rick Bergman's analyst day presentation of Nov 2009)

edit2: Here is your post confirming

60-80&#37; over Magny Cours is accurate for Interlagos.

 

[ElFenix]

on the gpu end they've been very strong for the past 18 mos or so, though much of that was due to NV miscues. on the cpu side, intel has been dominant for 5 years, though at least some of that has been due to AMD miscues. If AMD delivers this round then we'll have a much more competitive environment, not lease because of intel's recent stumble.

it really has been since mid of 2006. i forget that core2 is that old. i built my old opteron 165 box only a couple months before that.