Bulldozer expected to be 50% faster than i7 CPUs

Dude this has already been brought up in two other threads (the original source, of course). For now just consider it complete hogwash that they made up (for better or for worse).
 
"I'll believe it when I see it." is my attitude at this point.
 
LOL, the rumored performance just doubled.

I mean, OMG AMD finally did it. 22 months late (so far) and it's still on top! Everyone wait for this chip and when it's released IT WILL BE THE BESTEST THING EVER. With a name like Bulldozer, whose details are coming up on a 4 year anniversary now, it can't possibly disappoint.

YEP 50% faster than i7. Anything much slower would surely be a big disappointment. :p
 
LOL, the rumored performance just doubled.

I mean, OMG AMD finally did it. 22 months late (so far) and it's still on top! Everyone wait for this chip and when it's released IT WILL BE THE BESTEST THING EVER. With a name like Bulldozer, whose details are coming up on a 4 year anniversary now, it can't possibly disappoint.

YEP 50% faster than i7. Anything much slower would surely be a big disappointment. :p

Same thing happened with the 6970 build up. It went from beating a gtx580 by %5 then the rumor mill starting churning out ridiculous %30 guesses.
 
Same thing happened with the 6970 build up. It went from beating a gtx580 by %5 then the rumor mill starting churning out ridiculous %30 guesses.
Which proves yet again that the overeager fans are their own worst enemies. If people had the patience and sense to wait for real, confirmed benchmarks instead of wild speculation based on early reports and "insider" leaks they wouldn't find reality so depressing.
 
Even going by the information though this is not as impressive as it might seem. The article states a 8-core bulldozer will be faster 50% faster then a i7 950 (not SB) in "games, rendering and multimedia application." I mean if you take a optimized multithreaded rendering benchmark, being 50% faster is certainly what you would expect if anything.

For instance I believe the six core i7 980X itself exceeds a 50% performance improvement in many rendering benchmarks compared to the i7 950.

Also any claims of actual gaming performance improving that much from a CPU (assuming current trends hold) is just unrealistic itself.
 
I heard that AMD found some Ompa Lumpas and they made the Bulldozer taste 75% better than any other chip, except sour cream and cheddar Ruffles.
 
Last edited:
Didn't AMD also claim that Phenom was going to be 40% faster than Core 2?
 
It's going on five years now that AMD has been out of the high-end CPU game, it's time to put up or shut up.
 
It's going on five years now that AMD has been out of the high-end CPU game, it's time to put up or shut up.

I was thinking it must have been something like that.

Almost five years ago exactly I finally went (and stuck*) with an AMD system with an FX-60. The dual-core was slightly clocked under the other FX single-core chip, but I was anticipating new games would start making use of both cores. Unfortunately, it didn't happen fast enough.

I didn't think there had been anything between then and now from AMD. I remember shortly after putting together my AMD system, the Q6600 hit the market with its almost guaranteed 50% overclock, and Intel's never looked back(?)


(*and stuck - in 2000 I built an AMD system but the VIA motherboard sucked so bad that I took it back and instead of the Athlon(?) 1GHz, I went with a P3-933... Unfortunately that mobo also had its problems too, but it wasn't absolute show-stopping problems... A decade ago was not a great time for mobo QA...)
 
Yeah, if I remember the desktop Core 2's came out in June 2006 and just rocked AMD's world. I was running an Athlon 64 x2 4800+ at the time. I held on to that system WAY longer than I should have, got the Q6600 In June 2007 and OMG, the jump was incredible.

I admit on that day I lost all my love for AMD CPUs and just haven't been interested in them since.
 
Yeah, if I remember the desktop Core 2's came out in June 2006 and just rocked AMD's world. I was running an Athlon 64 x2 4800+ at the time. I held on to that system WAY longer than I should have, got the Q6600 In June 2007 and OMG, the jump was incredible.

I admit on that day I lost all my love for AMD CPUs and just haven't been interested in them since.


lol i went from the x2 6400+ to the x2 BE-2350 @ 3Ghz(and no it wasnt a downgrade i actually liked the be-2350 more then the 6400+), then to the phenom II 940 @ 3.5Ghz.
 
Actually this is quite a brilliant strategy. Come out saying your new processor line will be 50% faster than the competition.....then wait so long to actually release the new processor line that we're all senile from old age and dont remember anything.

So Sandy Bridge rumors start floating around April an the processor drops barely 9 months later. Bulldozer rumors start floating back sometime in 09 and we're STILL waiting for an IDEA of when theyll be released.

Anybody else find that frustrating.....or is it just me?
 
Yeah, if I remember the desktop Core 2's came out in June 2006 and just rocked AMD's world. I was running an Athlon 64 x2 4800+ at the time. I held on to that system WAY longer than I should have, got the Q6600 In June 2007 and OMG, the jump was incredible.

I admit on that day I lost all my love for AMD CPUs and just haven't been interested in them since.

Not knocking your 980x when I say this, but I mainly mention the mark of the Q6600 (vs. just a C2D) as the Q6600 for me made it clear you could have a high-end gaming system without having to spend $1,000 on the CPU.

Again, not knocking you, but for just a pure gaming system only, the new 2600K is certainly a very worthy value compared to the 980x. It's crazy that I'm reading 5Ghz OCs on the thing... Minus the GPUs, you could put together a new 2600K system for the price of a 980x alone, that's really hard to ignore...

(I'm comparing Intel's newest mainstream chip to the top-of-the-line one they currently have, just like when I bought the FX-60. You running a 980x I swear is coincidence in the discussion... But I do admit I highly doubt ever paying $1,000 for a CPU again. FX-60 was a one-and-done deal...)

But yeah, I'm considering upgrading my 920 system to a 2600K system just for the extra OC alone... I'll think I'll wait 'til I'm convinced I'm CPU-bound on something. Would the 2600K already help Metro 2033?
 
AMD has been long on promises, and delivered nothing of interest since the Athlon64 X2s.

I like AMD, I still run an X2, but we've heard these promises before. And every time they've been late, slow and disappointing. Honestly, I'd be perfectly OK with them saying they've got no where near the budget or the resources of Intel and are still producing a CPU that does 99% of what you want for a lot less money, but these claims they're going to pull another giant killer like K7 out of nowhere are getting tiresome.
 
AMD has been long on promises, and delivered nothing of interest since the Athlon64 X2s.

I like AMD, I still run an X2, but we've heard these promises before. And every time they've been late, slow and disappointing. Honestly, I'd be perfectly OK with them saying they've got no where near the budget or the resources of Intel and are still producing a CPU that does 99% of what you want for a lot less money, but these claims they're going to pull another giant killer like K7 out of nowhere are getting tiresome.


except these aren't promises. this whole thread is based on bogus information from a source/site that did it just for the extra hits. JF_AMD has already said none of this information comes from AMD. it comes right back to the fud started on the AMD HD6970 having 1920 SP's. AMD never said it had 1920 SP's, yet when the 6970 was released with less than 1600 everyone went ape shit about it based on information that never came from AMD. just add it to the list of crap to take with a grain of salt til we get some real numbers which should be the opteron based bulldozers sometime later in this quarter.
 
Last edited:
Not knocking your 980x when I say this, but I mainly mention the mark of the Q6600 (vs. just a C2D) as the Q6600 for me made it clear you could have a high-end gaming system without having to spend $1,000 on the CPU.

If I were building a gaming rig today I'd go SB but I bought the 980x at launch so I've had it ten months and I do a lot more than game on this puppy and I do use the extra cores a bit for things like video transcoding.

I have very pleased with this rig and the component choices I've made and got a good amount of time.
 
except these aren't promises. this whole thread is based on bogus information from a source/site that did it just for the extra hits. Jeff has already said none of this information comes from AMD. it comes right back to the fud started on the AMD HD6970 having 1920 SP's. AMD never said it had 1920 SP's, yet when the 6970 was released with less than 1600 everyone went ape shit about it based on information that never came from AMD. just add it to the list of crap to take with a grain of salt til we get some real numbers which should be the opteron based bulldozers sometime later in this quarter.

I don't want to say, or even really speculate, whether or not these numbers are real. And you're correct in that blaming AMD entirely for the hype around their products is unfair. The internet is full of misinformation and speculation, but really how hard is it to say 'that's not a real estimate' or 'those aren't official numbers.'
Just on the CPU side, the later iterations of the X2s, the original Phenoms and the Phenom IIs have all been late and fairly disappointing compared to the hype.
And really, there's nothing wrong with the Phenom II or X4s today, they're great CPUs for a price. I just don't want to hear about how much better than reality the next one will be.
 
If I were building a gaming rig today I'd go SB but I bought the 980x at launch so I've had it ten months and I do a lot more than game on this puppy and I do use the extra cores a bit for things like video transcoding.

I have very pleased with this rig and the component choices I've made and got a good amount of time.

Which is why I said high-end gaming system, meaning dedicated gaming system.

I've just learned the hard way with my purchase of the FX-60 that a $1,000 CPU for a dedicated gaming machine is not good value. And unfortunately like my experience, not necessarily even the best money can buy, either.

980x isn't a bad CPU, and can obviously beat i7/SB on specific benchmarks and thus for specific usages, like video encoding. But for a pure dedicated game machine, the 980x is somewhat like my FX-60: the extra core(s) aren't being used to their advantage currently in games that the single (now quad)-core systems are doing. So, perhaps best to save the ~$700 on the cheaper CPU and put that money towards GPUs or the next round of CPUs.


Nothing AMD's done the last half-decade makes me excited that Bulldozer may be 50% better than what I am running right now...
 
If I were building a gaming rig today I'd go SB but I bought the 980x at launch so I've had it ten months and I do a lot more than game on this puppy and I do use the extra cores a bit for things like video transcoding.

I have very pleased with this rig and the component choices I've made and got a good amount of time.

Hey, why not join our folding team and put that bad boy to work? Intel hex cores absolutely eat folding@home work units for lunch, especially if you put a good overclock on it. If you feel like adventuring, you can even put those Fermis to work folding GPU work units :D
 
I don't want to say, or even really speculate, whether or not these numbers are real. And you're correct in that blaming AMD entirely for the hype around their products is unfair. The internet is full of misinformation and speculation, but really how hard is it to say 'that's not a real estimate' or 'those aren't official numbers.'
Just on the CPU side, the later iterations of the X2s, the original Phenoms and the Phenom IIs have all been late and fairly disappointing compared to the hype.
And really, there's nothing wrong with the Phenom II or X4s today, they're great CPUs for a price. I just don't want to hear about how much better than reality the next one will be.

But why must AMD say anything. To comment on a Rumor gives it more headway. So far one report not in English halfway across the planet made one rumor. Should a company be expected to have a PR person dedicated to every web rumor/article any hack can create? Also how do you battle it, to be official you have to have a presser, to have a presser you have to spend money, waaaaay more then it takes for some hack to post wild news on their site?
 
But why must AMD say anything. To comment on a Rumor gives it more headway. So far one report not in English halfway across the planet made one rumor. Should a company be expected to have a PR person dedicated to every web rumor/article any hack can create? Also how do you battle it, to be official you have to have a presser, to have a presser you have to spend money, waaaaay more then it takes for some hack to post wild news on their site?

You are correct. I can't tell you if this is right or wrong, all I can say is that it did not come from us.
 
Also how do you battle it, to be official you have to have a presser, to have a presser you have to spend money, waaaaay more then it takes for some hack to post wild news on their site?
The internet has made advertising relatively cheap and quick. Marketing or at least informing the enthusiast crowd here is also pretty easy. JF or AMD could just give a official release every other week or so. There'll always be some bad info. floating around no matter what, but that would go a very long way towards discrediting the fake rumors and people who spread them.
 
The internet has made advertising relatively cheap and quick. Marketing or at least informing the enthusiast crowd here is also pretty easy. JF or AMD could just give a official release every other week or so. There'll always be some bad info. floating around no matter what, but that would go a very long way towards discrediting the fake rumors and people who spread them.

Again and this goes back to the middle ages. Responding to a fake rumor gives that person a voice. Not only that but there is a separation from what JF says here and what his department officially states. A Presser is a true statement from the company, it is written up by important people (or their secretary) analyzed by lawyers and published for the sake of their board and stockholders. JF on the other hand here is given a set up of guidelines of what not to talk about and becomes and official/unofficial person to the users.

He already said it was not from an AMD source. Now that isn't enough? But that is exactly what your asking for. Which now you want them to give performance numbers that can change at any time for a product that hasn't yet gone into production. Where does it end,? If JF says its not 50% faster only 25%, will you believe it then? Probably not, next you would want something official. Would an a Presser work? You would say yes but then again it would not be, you would want info from HardOCP or Anandtech and so on, as third party overseers.

As a company that is struggling, that won't sell 10 billion in equipment becuase no one ever gets fired for purchasing their product, they have to be really careful. Get hopes up and their current products will no longer sell as people save up, admit they are slower and people just look for the latest alternative from another company. Say nothing but, its done when its done, then they can still sell current product stock, and not worry about whether they hit performance goal expectations that may or may not have been established by them prior.

No the best thing for every company specially an underdog, is to let the production product speak for itself and not comment on unannounced products and random fake rumors. Because in the end of the day, once you start doing that, it becomes expected and all of a sudden the company has to respond to nutjob with a domain name.
 
If I posted up that bulldozer was XX% faster than ABC product, I'd be innundated with "what about benchmark 1" "what about benchmark 2" "what was the configuration?" "What if you ran it with 8MB of memory?"

"WHY CAN'T YOU RUN EVERY BENCHMARK I ASK FOR? AND RUN THEM NOW MONKEY BOY!!!"

Trust me, it is either no benchmarks or all benchmarks. There is no such thing as just a little bit.

It is a no win situation. Waiting until launch works better for the business in the long run.
 
Again and this goes back to the middle ages. Responding to a fake rumor gives that person a voice.
Random internet jack ass vs JF-AMD or some other offical AMD source isn't even a competition.

Not only that but there is a separation from what JF says here and what his department officially states.
Yet here he is talking to us. He can obviously say some things.

He already said it was not from an AMD source. Now that isn't enough?
People are complaining about there being a lack of info. in general close to launch, not necessarily just this threads' original article or his response to it.

Which now you want them to give performance numbers that can change at any time for a product that hasn't yet gone into production.
If AMD doesn't have performance numbers now then a Q2 launch is impossible. The chip should've been done quite a while ago, the only thing we should be waiting on is for them to fab up enough chips for a good supply at launch.

If JF says its not 50% faster only 25%, will you believe it then?
If he cited the benchmarks and gave some screens with exact numbers vs. other AMD or Intel products sure. I doubt he'd straight up lie.

Probably not, next you would want something official.
He is offically a AMD employee. Good enough for me and most people I'd bet.
 
If I posted up that bulldozer was XX% faster than ABC product, I'd be innundated with "what about benchmark 1" "what about benchmark 2" "what was the configuration?" "What if you ran it with 8MB of memory?"

"WHY CAN'T YOU RUN EVERY BENCHMARK I ASK FOR? AND RUN THEM NOW MONKEY BOY!!!"

Trust me, it is either no benchmarks or all benchmarks. There is no such thing as just a little bit.

It is a no win situation. Waiting until launch works better for the business in the long run.
3-4 real world non-synthetic benchmarks would satisfy most people's curiosity I would think. Obviously some are going to want to know how it'll perform for their custom app or hardware config. no matter what but at least they'd have a rough yard stick to judge by.
 
but does it play Crysis?


Seriously though, If it has IPC parity to the Intel i7 and can clock to at least 3.5Ghz when using 4 cores I'm going to upgrade from my QX9650 and go get one. Or put another way.... Compared to my QX9650 If I am going to gain 30% performance boost in most of my apts and a greater than 50% performance boost in my scalable apts then I am definitely going to buy bulldozer.
 
Seriously. We release benchmarks for a product launching next quarter. Sales come screeching to a halt.

OEMs have hudreds of thousands of systems in the channels that start getting returned by their retailers.

OEMs come to us and ask for compensation for lost sales.

Satisfying the curiosity of people who are anxious does not cover the cost of disruption in the supply chain. I have been in this business for 20 years now, getting too far ahead of your launch with performance information does not help you, and in most cases it hurts you.

This is a buisness folks, it's not some "mine is bigger than yours" compeition. Sorry to bring the dose of reality in.
 
Back
Top