Bulldozer Possibly Pushed back till October

something which is interesting in a way that gains from single thread are limited by mhz and ipc they are both at a ceiling.
Click to expand...

This is not true for Intel. Or at least every revision for some time has generated at least a 10% improvement in single threaded ipc with at least the same frequency or higher. For AMD they seem to be a generation or two behind Intel in this which is the same time frame they are behind in process shrinks.
 
Last edited:
What marketing? We've already argued about that for a couple of pages. AMD doesnt think they need to advertise or market cause the majority of people that buy laptops and PC's supposedly dont know or care what brand of CPU it has in it. And the people like us that do know are such a small segment of the market that its not worth marketing to us.
 
conscript said:
you're grasping at straws...but i'll bite. Forget all the leaks, every one. You don't continually delay a product you will think be successful for your business...unless it has issues. You realize bulldozer has been delayed since before you were a member here?
Click to expand...

This is probably the biggest indicator of all.
It's possible there a bug in B1 revision that warranted holding it from release, but that's pretty unusual and not consistent with any of the rumors or leaked benchmarks that have been cropping up since spring when the B0/B1 chips were known to out there.
If B1 was competitive and performing up to where AMD wanted it, we wouldn't have seen a B2 revision before launch.
 
FreiDOg said:
This is probably the biggest indicator of all.
It's possible there a bug in B1 revision that warranted holding it from release, but that's pretty unusual and not consistent with any of the rumors or leaked benchmarks that have been cropping up since spring when the B0/B1 chips were known to out there.
If B1 was competitive and performing up to where AMD wanted it, we wouldn't have seen a B2 revision before launch.
Click to expand...

It's been admitted by AMD that b0 and b1 suffered from a defect that would lead to the hypertransport becoming flooded and killing performance.
 
MacLeod said:
What marketing? We've already argued about that for a couple of pages. AMD doesnt think they need to advertise or market cause the majority of people that buy laptops and PC's supposedly dont know or care what brand of CPU it has in it. And the people like us that do know are such a small segment of the market that its not worth marketing to us.
Click to expand...

Just what kind of marketing are you expecting? Where did they fail in letting you know about Bulldozer?
 
Meanwhile we ge extreme edition Llano cpus with unlocked multis:

A8-3870k.jpg
 
yup, 3.9g is the most the world record onboard gpu bencher got with it. however he could almost hit 5g but the mobo couldn't handle it with the video working.
 
Jellyfishpudding said:
BTW, Jeff's little rant is making the rounds. EVGA, Guru3D. Can't be a good thing.
Click to expand...

I'm in sales as well...

Oh O.K. business development aka marketing or whatever the shat you call it.

For Jim-Bo to make it as far as he did in sales then crash so hard... in public ... I think we do have a serious flop and thats not gigaflop wise.

Reason I say that in my early years of "marketing" I had doors slamed, people hang up, lead you on then laugh at you, promise you the world for lunch/games/treats/you name it, then say oh you have to talk to blahhh....

For him to break down like that I might just put my resume into AMD.. shoot could be a good run for a few months.

If you look at teh previous quote you can see alot of people dont understand marketing... it's not about letting people know what you have and when. It's about the sales anyway you can...

JM was leading us on keeping us fired up etc etc etc... now even he cracked.
 
Forgot to add WOW GREAT the

A8-3870K looks like a great chip (seriously I'm thinking abouty that one...)

Problem is it doesn't release until 2H of 2012 according to AMD's own anticipated Launch2



2 - note that the 4Q2011 basically means that we are going to try to keep you hanging on for 6 months. Or at least 3 revisions....
 
what's funny is how everyone is so offended, and how there are so many victims out there, in regards to that faq. really? you're feelings are hurt from an internet post? that is comedy.
 
Kluden said:
what's funny is how everyone is so offended, and how there are so many victims out there, in regards to that faq. really?
Click to expand...

Yeah, I don't get it either, but meh.

Still waiting on bulldozer to infinity (no choice, unless I want to buy new board).
 
tlc4tvm said:
Forgot to add WOW GREAT the

A8-3870K looks like a great chip (seriously I'm thinking abouty that one...)
Click to expand...

Actually if AMD dropped the whole LLano part and make only CPU part of it then sell unlocked chips like this for 80-100$ they would have killer in price/performance
 
Michaelius said:
Actually if AMD dropped the whole LLano part and make only CPU part of it then sell unlocked chips like this for 80-100$ they would have killer in price/performance
Click to expand...

They already do. They are called Black Edition chips. There are 3 of them under $100 right now. The 720BE, 555BE, and 560BE. Many of them unlock to quads also. The 955BE is getting awfully close to $100 through sales right now.
 
CaptNumbNutz said:
They already do. They are called Black Edition chips. There are 3 of them under $100 right now. The 720BE, 555BE, and 560BE. Many of them unlock to quads also. The 955BE is getting awfully close to $100 through sales right now.
Click to expand...

Yeah, but Llano is 32nm as opposed to the Phenom II 45nm and original Phenom 65nm, which means it should - if unlocked - clock MUCH higher all else being equal.

Seeing that any recent 45nm BE chip will hit 4ghz with a decent cooler and most (at least quads or below) will hit 4.2 fairly reliably, it's not unreasonable to think that 5+ ghz should be possible from an unlocked 32nm k10.5 core.

What those numbers actually are only AMD knows right now as there hasnt been an unlocked Llano yet.

The rumor mill has suggested that Llano is suffering from terrible yields in the GPU portion of the chip due to difficulties in adapting GPU designs to Silicon on Insulator (SOI) production lines.

This would be a great opportunity to release some BE chips that are essentially Llano chips with the GPU disabled, and adapted to AM3 sockets.

They wouldn't quite be as good as the upcoming Bulldozers, but they ought to he a great enthusiast bang-for-the-buck chip...

At the very least it ought to be better than tossing all those Llanos with bad GPU's in the trash...
 
Last edited:
Zarathustra[H];1037718893 said:
Yeah, bit Llano is 32nm as opposed to the Phenom II 45nm and original Phenom 65nm, which means it should - if unlocked - clock MUCH higher all else being equal.

Seeing that any recent 45nm BE chip will hit 4gjz with a decent cooler and most (at least quads or below) will hit 4.2 fairly reliably, it's not unreasonable to think that 5+ ghz should be possible from an unlocked 32nm k10.5 core.

What those numbers actually are only AMD knows right now as there hasnt been an unlocked Llano yet.

The rumor mill has suggested that Llano is suffering from terrible yields in the GPU portion of the chip due to difficulties in adapting GPU designs to Silicon on Insulator (SOI) production lines.

This would be a great opportunity to release some BE chips that are essentially Llano chips with the GPU disabled, and adapted to AM3 sockets.

They wouldn't quite be as good as the upcoming Bulldozers, but they ought to he a great enthusiast bang-for-the-buck chip...

At the very least it ought to be better than tossing all those Llanos with bad GPU's in the trash...
Click to expand...

They are actually producing locked quad x4 with gpu disabled but due to gpu transistors being still there it has 100W DP :mad:
While hypotheticall cpu only 32nm Llano would be much smaller than 45nm quads and compared to current Llanos much cheaper to produce (gpu takes more than half die size in Llano) and have much better yields as well as decent power consumption.

Get such thing to 4,5 Ghz and voila perfect budget friendly enthusiast chip.
 
Michaelius said:
They are actually producing locked quad x4 with gpu disabled but due to gpu transistors being still there it has 100W DP :mad:
Click to expand...

TDP doesn't mean the chip uses that amount of power when loaded. In fact, if the GPU is disabled, it will use maybe 1-2W, if that, more likely zero. So what's the problem?
 
iconiK said:
TDP doesn't mean the chip uses that amount of power when loaded. In fact, if the GPU is disabled, it will use maybe 1-2W, if that, more likely zero. So what's the problem?
Click to expand...

Sigh yes ok you are right. They gave it higher TDP than 45nm Phenom II quads have just for lols.
 
Michaelius said:
Sigh yes ok you are right. They gave it higher TDP than 45nm Phenom II quads have just for lols.
Click to expand...

It's the same sillicon as the ones with enabled GPU. But why should they spend several thousand dollars running power consumption tests on some crippled chips just to lower a number? Just give it the same TDP and call it a day, TDP is only important for heatsink designers, not for the average user. Not having the GPU enabled would probably drop 50-60W of full load power usage.
 
iconiK said:
It's the same sillicon as the ones with enabled GPU. But why should they spend several thousand dollars running power consumption tests on some crippled chips just to lower a number? Just give it the same TDP and call it a day, TDP is only important for heatsink designers, not for the average user. Not having the GPU enabled would probably drop 50-60W of full load power usage.
Click to expand...

Exactly. I wouldn't be surprised if the graphics portions of Llano and Brazos were responsible for 55-75% of the power consumption and heat generation.
 
Zarathustra[H];1037718893 said:
This would be a great opportunity to release some BE chips that are essentially Llano chips with the GPU disabled, and adapted to AM3 sockets.

They wouldn't quite be as good as the upcoming Bulldozers, but they ought to he a great enthusiast bang-for-the-buck chip...
Click to expand...

why not just sell them for socket fm1 as phoenom II's. thats really what they are. the whole reason they couldn't use am3 for llanos is because of the extra lanes needed for the apu.

better of keeping it for fm1, at least that ways there an upgrade path (however i believe fm1 is only going to be around mabey 2 years) after this generation of bullzoders am3 is end of life, its being replaced with fm2 when trinity is released in a year. the new socket will be needed for the apu.
 
Kluden said:
what's funny is how everyone is so offended, and how there are so many victims out there, in regards to that faq. really? you're feelings are hurt from an internet post? that is comedy.
Click to expand...

I don't think anyone's offended. We're just pointing out how out-of-character it is for someone who holds his position. It definitely puts him beneath whatever standard AMD holds him to.
 
Poisoner said:
It's been admitted by AMD that b0 and b1 suffered from a defect that would lead to the hypertransport becoming flooded and killing performance.
Click to expand...

I stand corrected.
I thought that was resolved with B1; since not the extra stepping wouldn't directly be attributed to performance issues.
 
Jellyfishpudding said:
I don't think anyone's offended. We're just pointing out how out-of-character it is for someone who holds his position. It definitely puts him beneath whatever standard AMD holds him to.
Click to expand...

Yeah and heres what I dont understand.
JF states that he is a server guy. Well thats fine and dandy so WHY is he commenting on the client side of things? On Enthusiets forms to boot? I could see if he was posting on walmarts computer forurms trying to inform the cpu unware but hes not. Intstead hes trying to make comments to enthusiests about how small of a market we are and that having a halo product is usless??? Again were enthusiets... Like telling a corevett owner that a minivan is better. He just seems like hes frequenting the wrong forums... Or at least giving the wrong impresion. And with his recent "FAQ" it seems that hes finally seeing it himself. Again this is just my own opinion. I will give JF credit for sticking through all of this even if hes unfortunatly starting to crack.:rolleyes:
 
Again were enthusiets... Like telling a corevett owner that a minivan is better.
Click to expand...

Except he only said Llano was better from a profit perspective, not a performance one.
 
I am genuinely interested in the idea that product announcements "stall" demand. Do you have any data that suggests that this really happens?

I would be interested in an examination of the data. If you announce you will release your product six months from now, does demand stall? 3 months? 2 weeks? I highly suspect six months has little effect (frankly, I doubt none of these scenarios have much effect in the desktop space, it really does take an enthusiast to realize that BD even exists). I could perhaps see three months or two weeks, but stalled demand is still demand. If I'm buying low margin AMD products now, is it not in AMD's best interest to convert all production to high-margin supply for future nominal and newly "stalled" demand? You would think on one side of every stall, there would be an immediate surplus of demand post-release.

We see that in macroeconomics (more my area), where actors will reduce consumption in a recession but inevitably have to make up, at least somewhat, for those expenditures as technology still advances and items still wear out. No item is perfectly durable.

So I guess my question really has three parts: do people really stall in a measurable way, how close to the expected release date do we observe this behavior, and, if it exists, do you see a subsequent increase in demand post-release as the pent up demand is released?

I can't hardly think there's much of an impact here as, as was stated, only ~6% of buyers are considered enthusiasts and only some portion of them will engage in stalling behavior.

Honestly, I was always under the impression that release dates were vague not because of some mythical concept of stalling, but because it reduced the company's liabilities in meeting those targets. Far better to say Q4 than October 2nd lest you be forced to constantly return to your customers and stockholders explaining a delay and exposing yourself to personal and legal liabilities. If I were C-level, I'd certainly be as vague as possible about release dates so I wasn't constantly putting my career on the line as I met with the board from Oct to Dec explaining why the product was delayed and deadlines missed.

FWIW, JF, I wasn't at all offended by your FAQ. I think cooler heads understand you can't really say anything substantive about performance or dates. I think the attacks you receive here, elsewhere, or via email are completely inappropriate. I do not deride you or your previous targets or statements. However, I do not believe that AMD is holding back the release date because of demand stalling. I highly suspect the release date has not been made public because you honestly don't know when it's going to be ready. Whether the last spin was ineffective at improving yields, performance, power usage, etc, is irrelevant, a delay is a delay. If I had to take a personal shot in the dark, I would say BD won't be seen this month. Usually, with a major release like this, there would simply be more "action" if the release was only two weeks away.

I sincerely hope AMD releases a performer, preferably this year ;).
 
Buying that 1090t x6 during one of thr [H]ard Newegg deals as a "wait for bulldozer" CPU is really starting to pay off as the right choice.

If I had waited for bulldozer I would have gone nuts by now.
 
BigDH01 said:
I am genuinely interested in the idea that product announcements "stall" demand. Do you have any data that suggests that this really happens?
Click to expand...

Im having a hard time understanding their entire marketing or PR or whatever you want to call it.

Right now, demand for AMD processors is pretty stalled on its own. The $125 i3-2100 is beating the $180 Phenom X6 in most benchmarks! Currently there is no reason at all to buy an AMD processor if your cpu budget is over $100 and when youve got AMD fanboys like me recommending Intel procs on all the forums, demand has stalled. Having at least an estimated release time might keep a lot of people that are about to pull the trigger on a Sandy Bridge to hold off a while and at least wait and see how BD stacks up.

I realize Im a dumb blue collar schmuck but I just dont get it. Just like I dont get their stance on advertising and thinking its ok when the majority of average consumers have never heard of AMD but can sing the Intel jingle on demand. Or going on multiple enthusiast forums just to tell people youre not going to tell them anything and it doesnt matter because we're such a small, irrelevant segment of the market anyway.
 
BigDH01 said:
I am genuinely interested in the idea that product announcements "stall" demand. Do you have any data that suggests that this really happens?
Click to expand...

There is a textbook example of it taught to anyone who studies business in college.

It is called the Osborne Effect.

It is named after Adam Osborne who founded the Osborne Computer Corp in 1981. Their first computer, Osborne 1, was a success. He announced 2 new products they were working on as successors to the Osborne 1, sales immediately stalled on currently shipping Osborne 1's. Due to fierce competition in that market and drastic drop in sales as people anticipated the product, Osborne Computers went out of business soon after.

If the company is large enough and has enough capital, they can survive the Osborne effect. Apple and Intel are two companies that fit this description, AMD is not. Apple sort of has their own mini self inflicted version of the Osborne effect because they have a habit of releasing particular products (e.g. iPhone) at near the same time every year. This year was an exception for them, but I'm sure they see drops in sales in the couple months leading to the new release of their product.
 
CaptNumbNutz beat me...

BigDH01 said:
I am genuinely interested in the idea that product announcements "stall" demand. Do you have any data that suggests that this really happens?
Click to expand...

From this article on the front page, there is the case of the Osborne Computer Corporation, it was the 2nd story in the article. Guy announces new products before they are ready for launch...

However, there is one particular event in computer history in which Mr. Osborne’s name will forever be associated with:The Osborne Effect.

What happened to the Osborne Computer Company after the announcements of the “Executive” and the “Vixen” is now classic business school material. Due to the pre-announcement of the newer, better products while the current inventory in the reseller channel was still full, buyers were no longer interested in current products.
Click to expand...

...and the company goes bankrupt because sales of it's current products disappear. So apparently it does happen in real life. :confused: Makes sense.
 
That would make sense for the last generation of i7's. Those were a home run for processors and still had high demand so I can get how announcing a release date for Sandy Bridge could make people wait and skip that new i7-850. But like I said, Phenom is flailing away at the bottom of most benchmarks and I dont see too many people wanting a new Phenom X4 980 especially when they find out they can get a cheaper Intel DUAL CORE and have better performance. My theory is that giving out a BD release ETA would keep those about to buy an Intel from jumping ship just yet.
 
CaptNumbNutz said:
There is a textbook example of it taught to anyone who studies business in college.

It is called the Osborne Effect.

It is named after Adam Osborne who founded the Osborne Computer Corp in 1981. Their first computer, Osborne 1, was a success. He announced 2 new products they were working on as successors to the Osborne 1, sales immediately stalled on currently shipping Osborne 1's. Due to fierce competition in that market and drastic drop in sales as people anticipated the product, Osborne Computers went out of business soon after.

If the company is large enough and has enough capital, they can survive the Osborne effect. Apple and Intel are two companies that fit this description, AMD is not. Apple sort of has their own mini self inflicted version of the Osborne effect because they have a habit of releasing particular products (e.g. iPhone) at near the same time every year. This year was an exception for them, but I'm sure they see drops in sales in the couple months leading to the new release of their product.
Click to expand...


Only problem with that comparison is that AMD has already announced their replacement, they just won't announce an actual launch date. You have to wonder if that uncertainty is better or worse than the drop in sales caused by announcing a firm release date (assuming there is one, since announcing a product and announing a release date for an already announced product are two different things). Witness how many "what should I buy now" threads lead to responses like "wait and see what BD brings" - so they've already impacted sales - I just don't see how pinning a launch date is going to change that. Heck, announcing a date farther in the future may even help their sales - if you know BD isn't coming until Jan, there isn't much point waiting and you might go ahead and buy an X6 now.
 
All I know is that AMD has shed a ton of fans and this time it may be greater then the time(s) before. Waiting for the Phenom 1, me and my friends were waiting, but all the delays forced us to go with Intel (and glad we did) with the Core 2 series...Now, I couldn't wait to upgrade and went with the i7 2600K because AMD has been sitting on their hands for 6 years.

Friends and family want to build gaming machines, and I want to recommend waiting for BD, but that makes no sense now. I usually say pick up an i3 2100 if your on a budget, or get the 2600k if your not. Even without the ability to OC the i3 is potent and inexpensive. AMD has nothing...These people don't need 6 threads when 4 will do fine, especially with the higher IPC on the SB parts. An i3 2100 and a GTX460 and your good for a while...With no news or performance figures, it makes no sense waiting.

Hell, I'm starting to step away from the Radeon products due to this BS with BD...Get your Crap together AMD!!!
 
You must log in or register to reply here.
Back
Top