Friday, December 2nd 2011

AMD Realizes That Bulldozer Has 800 Million LESS Transistors Than It Thought!

AMD's new flagship Bulldozer "FX" series of processors have turned out to be mediocre performers in almost every review and benchmark going, sometimes even getting bested by the existing Phenom II and certainly no match for their Intel competition. To add to this tale of fail, it now turns out that AMD didn't even know how many transistors they have! Anand Lal Shimpi of AnandTech received an email from AMD's PR department and this is the revelation he had to share with us:
This is a bit unusual. I got an email from AMD PR this week asking me to correct the Bulldozer transistor count in our Sandy Bridge E review. The incorrect number, provided to me (and other reviewers) by AMD PR around 3 months ago was 2 billion transistors. The actual transistor count for Bulldozer is apparently 1.2 billion transistors. I don't have an explanation as to why the original number was wrong, just that the new number has been triple checked by my contact and is indeed right. The total die area for a 4-module/8-core Bulldozer remains correct at 315 mm².

Yes, something as basic as how many transistors are in their flagship product wasn't known about until months after the launch! This kind of info would be common knowledge within the company by the time the first tape-out is ready during the design and testing phase, so surely this cannot be and there must be some other explanation? If this is an attempt to make the processor look better by showing it "doing more with less", then this PR stunt has backfired spectacularly and it would have been better to have left the "error" as it was. Paradoxically, FX processors are a sales success and are flying off the shelves as we just reported, here.Source: AnandTech
Add your own comment

142 Comments on AMD Realizes That Bulldozer Has 800 Million LESS Transistors Than It Thought!

#2
simlariver
This must have something to do with automated design.
I wonder if there is still anyone at AMD that knows how to design a cpu.

Bad management decisions having large scale backfire.
Posted on Reply
#3
Mistral
by: TheLaughingMan
People still use fax machines?
You'd be surprised... :banghead:

by: simlariver
This must have something to do with automated design.
I wonder if there is still anyone at AMD that knows how to design a cpu.

Bad management decisions having large scale backfire.
That article you're linking to uses BD having 2 billion transistor as basis for its speculations. The 2 being a PR BS and BD actually having 1.2 would suggest to disprove it.
Posted on Reply
#4
tigger
I'm the only one
Does this alter any performance per watt figures? As i said earlier, the power consumption is worse.
Posted on Reply
#5
TRWOV
No, performance per watt would be the same if it had 10B or 500 million transistors.
Posted on Reply
#6
xBruce88x
nah it'll be the same... regardless of how many transistors they say it has... we already know how much power it uses and how it performs...
Posted on Reply
#7
qubit
Overclocked quantum bit
by: tigger
Does this alter any performance per watt figures? As i said earlier, the power consumption is worse.
That's true, but think about it: what if Bulldozer would have had 800M more transistors? Then the performance per watt would be improved, but overall the CPU would be using even more resources to achieve less. Poor AMD can't win either way! :laugh:
Posted on Reply
#9
reverze
i think we need to have a word with wizard about these articles, TPU isnt worth the page visit anymore
Posted on Reply
#10
qubit
Overclocked quantum bit
by: reverze
i think we need to have a word with wizard about these articles, TPU isnt worth the page visit anymore
Have you ever considered that I write these articles with his approval? ;) And have you considered that a lot of people actually enjoy them?

You're always free to look somewhere else. I have no problem with that at all. None.
Posted on Reply
#11
trickson
OH, I have such a headache
WOW any they hyped there CPU so well too . What with having the worlds record OC . But wait what they did not even KNOW how many transistors were even on there CPU ? What ?
Posted on Reply
#12
Eva01Master
I like these articles qubit, keep up the good work, it's always nice to see polemic topics discussed by the community and not everyone following a chosen path like drones. Now regarding the body of the article itself, I do not think AMD didn't know how many transistors they had in their design but screwed badly when supplying the information out. How could you manage to work at PR of an electronics giant and know squat about the electronics they produce?
Posted on Reply
#13
qubit
Overclocked quantum bit
by: Eva01Master
I like these articles qubit, keep up the good work, it's always nice to see polemic topics discussed by the community and not everyone following a chosen path like drones. Now regarding the body of the article itself, I do not think AMD didn't know how many transistors they had in their design but screwed badly when supplying the information out. How could you manage to work at PR of an electronics giant and know squat about the electronics they produce?
Yes, it's weird how one hand didn't know what the other was doing, isn't it? Especially over such a long time, where something this basic would become common knowledge throughout the company.

Now, with my tinfoil hat on, I reckon it was deliberate in some way. I just have no idea how, lol. Or if they were really that disorganized, then AMD has some serious problems.

Thanks for the kind words. :toast:
Posted on Reply
#14
theoneandonlymrk
so if i dont like what ya sayin i can ef off then right, good TPU frontin their man my issue isnt the news .more the woohoo slant you put on it:toast:
Posted on Reply
#15
qubit
Overclocked quantum bit
by: theoneandonlymrk
so if i dont like what ya sayin i can ef off then right, good TPU frontin their man my issue isnt the news .more the woohoo slant you put on it:toast:
People that take a pop at me with lame personal attacks can take a walk, yes. What's so wrong about that? Don't forget that personal attacks are expressly against the forum rules, too.

If you're not too happy about my writing style, then I'm always willing to listen to constructive criticism and suggestions and debate them with you. :toast:

Finally, these should be made in the comments section, or PM, not the news thread. If you want to make sure that I see your comment in that section, then feel free to PM me the link and I'll discuss things with you. :)
Posted on Reply
#16
3volvedcombat
Bulldozer is the best architecture in the world. Bulldozer is better then any other processor on the market.
Posted on Reply
#17
Yellow&Nerdy?
No wonder they fired their PR. Why would you hype your product to be so much better than it is? Yes, having 2B transistors might close the performance gap, but the power consumption would of been off the charts. Bulldozer failed in both design and marketing. Kinda makes you think about, if there are any competent people working at AMD anymore.
Posted on Reply
#18
FeuchterFutzi
OLD NEWS its from November 16th and german sorry

http://www.planet3dnow.de/cgi-bin/newspub/viewnews.cgi?id=1321455683

in googlish:

Wednesday 16 November 2011
16:01 - Author: Dr @
AMD Bulldozer Schrumpfkur missed - a virtual

Who now expected, it would be a new revision of the miracle bulldozer in the wings who will be disappointed. For the launch of the first "Bulldozer" processor-based desktop AMD FX (codenamed "Zambezi") had AMD communicated to the press through a die size of 315 mm ² and a transistor count of around 2 billion. As The Register then in the context of the launch of the new Opteron 6200 ("Interlagos") and 4200 ("Valencia") is only 2.4 billion transistors for the "Interlagos" spoke, asked some users cope in our forum, such as this number to the previous statement fits. After all, this figure is 40% below the original value of about 4 billion




Therefore we have asked AMD, after which it has been confirmed that the statement of The Register is accurate. Because all processors of the first "Bulldozer" generation on the same "Orochi" The building is for all versions 315 mm ² die size and 1.2 billion transistors to estimate. Since the Opteron 6200 is constructed from an MCM (Multi Chip Module) from each of two dies, each processor must here per the duplicate values ​​are estimated. The virtual downsizing in the number of transistors is built so simply due to a communication error. As it happened, you could not tell us yet.




Source: AMD
Posted on Reply
#19
erocker
by: qubit
People that take a pop at me with lame personal attacks can take a walk, yes. What's so wrong about that? Don't forget that personal attacks are expressly against the forum rules, too.

If you're not too happy about my writing style, then I'm always willing to listen to constructive criticism and suggestions and debate them with you. :toast:

Finally, these should be made in the comments section, or PM, not the news thread. If you want to make sure that I see your comment in that section, then feel free to PM me the link and I'll discuss things with you. :)
Please don't bother responding to it then. Show them where to post if anything. Remember, news articles are about the story, not you. Any problematic posts should be reported, not responded to as that creates more work for everyone. The topic is the news artice STAY ON TOPIC.

This whole news story has me scratching my head. Has it always been 800 million transistors less? Did AMD keep with the original number and not know about it? Did they disable the chip in a last minute attempt to bring it to market? Funky stuff...
Posted on Reply
#20
hellrazor
Can we get those 800 million and put them into FPUs so that we don't lose a core when each module is being used for floating-point ops?
Posted on Reply
#21
v12dock
So Bulldozer patch 1.1

Patch Notes:

Added 800 Million more transistors should see 40% performance increase.
Posted on Reply
#23
Crap Daddy
Wait. Are there really 800 million missing? Were there 2 bill. in the first place? L.A. Noire
Posted on Reply
#24
qubit
Overclocked quantum bit
by: erocker
Please don't bother responding to it then. Show them where to post if anything. Remember, news articles are about the story, not you. Any problematic posts should be reported, not responded to as that creates more work for everyone. The topic is the news artice STAY ON TOPIC.
I'm sorry, thanks for the good advice. :toast: It's kinda hard not to say something sometimes... :shadedshu I'll try harder next time. :)

by: erocker
This whole news story has me scratching my head. Has it always been 800 million transistors less? Did AMD keep with the original number and not know about it? Did they disable the chip in a last minute attempt to bring it to market? Funky stuff...
It's downright weird, isn't it? It's a bit like a car manufacturer not realising the capacity of the engine on a new model, only to correct the official spec later. It'll be funny if it really is 2bn transistors and they had to fuse off parts of the chip to bring it to market, perhaps due to a design issue, isn't it? That would imply extra performance is locked up in there...

To the best of my knowledge, a mistake like this has never been made before in the semiconductor industry.
Posted on Reply
#25
DarkOCean
This explains it somehow why it's so weak.
Posted on Reply
Add your own comment