AMD shows off 3.0 GHz Barcelona with 3x2900 XT?

Thank you for your inside knowledge and truly valid opinion :rolleyes:

While I dont agree with him on his points word for word, I do agree with the general geist of his points. The man does bring up valid points. While they may come off a bit extremist by his tone, they are certainly good points of concern.
 
Go AMD. Even if it is better then Intels selection, I probably won't go through all the changes to get this system (mobo, probably DDR3, etc.)

But hopefully what it will do is force Intel to attempt to make their chips run much cooler, because that tiny heatsink in that picture is impressive! Unless temps suck on it just like they do with Intel's stock on their quads.
 
Not quite. Each GPU client requires a CPU core to work, thus, if it were at all possible to run 3 GPU clients, you would only have one free CPU core. This scenario assumes that a GPU client compatible for the 2900 is even available, and that 3 such clients can be run simultaneously without issues. Highly unlikely. Best to run the SMP client alone or the SMP client + 1 GPU client providing there's one available/compatible with the 2900.

To all naysayers who are still insistent on discrediting AMD and hatching conspiracy plots, just one day ago most readers of this forum were speculating if 3GHz Barcelonas would even see the light of day this year in any form at all. I recall rumors not too long ago about 1.9/2.0GHz being the fastest stable Barcelona samples that AMD could currently produce. Well, Steve from this site witnessed one sample in action at a much higher clock. He viewed four cores being utilized in a game he actually played. His post is on the first page. Isn't that enough?

But 2.0 GHz is the top release speed of the Barcelona. What we saw was AMD demoing a Phenom at 3.0. One Phenom. There is a difference between producing one chip that hits 3.0 and producing that same chip at volume. For all we know, 2.0 is the fastest they can reliably get Barcelona right now. We just don't know and we won't know until they actually release the darn thing.

In other words, having one Phenom at 3.0 isn't enough. According to the latest info I could find, the top end Phenom will be released between 2.4 and 2.6, short of the 3.0 demonstrated. Anand says:

Anand said:
While Phenom won't be anywhere near that clock speed when it launches at the end of this year, AMD expects to be at 3GHz within the first half of 2008.

This tells us that volume of 3.0 is certainly not ready to go yet. They still have some refinements to make. What people saw at that demonstration was a golden sample.
IMO, AMD screwed the pooch with this one. They let no one have a stab at benchmarking the thing, as far as I can tell, they didn't set a firm initial clockspeed for Phenom, and if the coverage is any indication, they spent more time talking about products in 2009 than they did Barcelona, which they will be relying on to get them to 2009. Bulldozer does look exciting, to be sure, but their relative silence on a new product that's 1 month away doesn't bode well IMO.
 
^ I totally agree and share the same concerns Yes, critical mass will be reached when the day arrives for AMD to deliver the goods, but that's not what this thread is about. There are plenty of other threads about that specific topic.

My post was directed to all the individuals who doubted that 3.0GHz was possibly at all in 2007, let alone July . This proves that it is, as I hinted at in this post a week ago. AMD is not out of the running. Not yet anyway.
 
But 2.0 GHz is the top release speed of the Barcelona. What we saw was AMD demoing a Phenom at 3.0. One Phenom. There is a difference between producing one chip that hits 3.0 and producing that same chip at volume. For all we know, 2.0 is the fastest they can reliably get Barcelona right now. We just don't know and we won't know until they actually release the darn thing.

In other words, having one Phenom at 3.0 isn't enough. According to the latest info I could find, the top end Phenom will be released between 2.4 and 2.6, short of the 3.0 demonstrated. Anand says:

This tells us that volume of 3.0 is certainly not ready to go yet. They still have some refinements to make. What people saw at that demonstration was a golden sample.
IMO, AMD screwed the pooch with this one. They let no one have a stab at benchmarking the thing, as far as I can tell, they didn't set a firm initial clockspeed for Phenom, and if the coverage is any indication, they spent more time talking about products in 2009 than they did Barcelona, which they will be relying on to get them to 2009. Bulldozer does look exciting, to be sure, but their relative silence on a new product that's 1 month away doesn't bode well IMO.

Anand also posted that Barcelona is having a problem hitting high clocks, Agena will a launch at higher clocks then Barcelona
0.0
 
http://blog.pcmag.com/blogs/miller/archive/2007/07/26/2039.aspx

[. . .]Rick Berman, SVP, GM, Graphics Product Group said the technologies [3GHz Phenom + 2900XT] would be available this fall.

While I didn't hear this directly when listening to the presentation, PC-Mag claims to have heard this. While it is true that Barcelona will launch at 2GHz, Phenom will be launched a good few months later. Further, Phenom is simpler, it only has one HT link instead of three, and qualification for desktop chips is much more forgiving than for server processors. Before I get flamed, I'm not saying this is true. I'm just quoting PC-Mag and mentioning the possibility. I wouldn't be surprised either way.
 
what if they dont pan out! dear god, amd is gonna lose you as a customer. thats amd's biggest fear, if you didnt know. :rolleyes:

amd is gonna be fine, they could care less about enthusiasts, we are a small market share. the big money is in servers and desktops and laptops. they could care less about overclocking headroom, or any of that.
 
so AMD managed to get a single quad core system running at 3ghz sample into their analyst call in order to blow everyone in the crowd away because of the high clocks....

but what I don't get is, if this thing is supposed to be as awesome as they say it is, why no benches *at all*? Why only the experience score? Why not a Far Cry/encoding demo/etc? Why no performance metric so we know how it performs at 3ghz?

If this thing is supposed to be a killer, then it should be ok the release the benches. The tired excuse that it will give Intel time to react is complete bullshit, since they shouldnt be able to "react" in any meaningful way to reclaim the performance crown (if it is even lost), and if they are able to react when the numbers are finally released, then all this stalling by AMD meant nothing anyway....

seriously AMD, I'm buying my system in December, you have till then to give me a Q6600 @ 3ghz match, and by match I mean:

price range as low or *lower* than Q6600,
performance greater or MUCH greater than Q6600 @ 3ghz,
power consumption as good or better than Q6600

yes, the price must be as good, possibly lower, and the performance MUST be better .... why? Because I'd buy that Q6600 NOW and would have enjoyed it for months by December... so seriously, for making us wait so long, and without giving us ANY benches, this thing better be *better* or it will leave a sour taste in my, and many other people's mouths.


Those 40% better clock for clock claims had *better* pan out.
 
so AMD managed to get a single quad core system running at 3ghz sample into their analyst call in order to blow everyone in the crowd away because of the high clocks....

but what I don't get is, if this thing is supposed to be as awesome as they say it is, why no benches *at all*? Why only the experience score? Why not a Far Cry/encoding demo/etc? Why no performance metric so we know how it performs at 3ghz?

If this thing is supposed to be a killer, then it should be ok the release the benches. The tired excuse that it will give Intel time to react is complete bullshit, since they shouldnt be able to "react" in any meaningful way to reclaim the performance crown (if it is even lost), and if they are able to react when the numbers are finally released, then all this stalling by AMD meant nothing anyway....

seriously AMD, I'm buying my system in December, you have till then to give me a Q6600 @ 3ghz match, and by match I mean:

price range as low or *lower* than Q6600,
performance greater or MUCH greater than Q6600 @ 3ghz,
power consumption as good or better than Q6600

yes, the price must be as good, possibly lower, and the performance MUST be better .... why? Because I'd buy that Q6600 NOW and would have enjoyed it for months by December... so seriously, for making us wait so long, and without giving us ANY benches, this thing better be *better* or it will leave a sour taste in my, and many other people's mouths.


Those 40% better clock for clock claims had *better* pan out.

Haha ok...:rolleyes: You know AMD can't hear you rant.
 
I have to post here after reading all the bullshit in the last 5 pages.
"It's not a quadcore cpu, I can't see the cores running"
A room full of journalists played games on the thing dickhead.

"Why won't they run benches on it"
For christ sakes, this is the first time it has even been seen, give them time.

"It can't be running at 3ghz with that little HSF"
Short memory, remember Intel showing conroe with locked boxes? No one could look at them because they were watercooled.

"Those cards can't be crossfired because there is no bridge"
Are you blind? I can see a bridge. As for the third card, ever heard of physics?

Give them a break fellas.
This is indeed wonderful news.
 
I have to post here after reading all the bullshit in the last 5 pages.
"It's not a quadcore cpu, I can't see the cores running"
A room full of journalists played games on the thing dickhead.
Which proves nothing. Don't insult people btw.
"Why won't they run benches on it"
For christ sakes, this is the first time it has even been seen, give them time.
How were they time-limited to run a benchmark? They have plenty of time to run a benchmark which wouldn't take more than 5 minutes, but they won't.
Give them a break fellas.
This is indeed wonderful news.
AMD is basically teasing customers like a prostitute would tease a client.
 
This conspiracy theory stuff is always so idiotic. It happens every time. The fan boys crawl out of every corner. Seriously, they have nothing to gain from lying, except jail time. Investors would only be pacified for a very short time after which the company would loose all credibility. And the idea that they would stage this so that forums goers would hold off on their next purchase is equally laughable. The sales from people who actually read this article wouldn't make even a dent in their overall sales.

Seriously, why is it so surprising that companies come out with new products? These are not fourteen year old fan boys showing off in teh forumz. These are business, and this is an official announcement, not a rumor from the Inquirer. :rolleyes:
 
It was a technology day. AMD were never intending to show this as a release or let people benchmark it.
 
is this not a technology forum

go somewhere else if the concept of a forum doesn't make much sense to you


No its not,its the AMD adoration society.All that same I pray this thing launches at decent clocks,in a realistic amount of time. :(

http://www.overclockers.com/tips01184/

Let's get this straight: You're a company hemorrhaging red on the financial pages like a victim in a slasher movie.

You call a financial meeting of analysts to try to convince them that this is going to stop someday soon.

To impress them, you find the sweetest, most-blessed engineering sample you can find and get it to run stably at 3GHz. You then cobble together a beast box with plenty of video cards.

You bring it to the meeting where you desperately need to impress people, and what do you do with your "killer" box? Why, you

hide the benchmark of the game demo you're running and
not release any other serious measurement of what this "killer" box can do.

If you have the real deal, an Intel-killer, in your hands, is this what you do? Like hell you do!

If a 3GHz K10 quad-core can't impress, what can AMD possibly come out with in the next year that will do better?

No one, I repeat no one, in AMD's position would pass up a golden, I repeat golden PR opportunity to impress everyone if they had the real deal.
 
Now that Conroe has been out for over a year.... the only ones still running AMD in their systems are those who truly believe the AMD of 2007 can make a successful, market leading product launch.

OR those of us running high end 939 x2s or optys. :) my 4800+ w/ 2G ram still doesn't break a sweat with anything i throw at it.

I've built 3 conroe systems since, but i think i'm liekly positioned to end up back with an AMD (not intentionally - just happenstance)

And for those of you claiming AMD rigged a presentation like this with fake stats, dummy cards, or XML edited winstat, you're idiots.

zv
 
Now that Conroe has been out for over a year.... the only ones still running AMD in their systems are those who truly believe the AMD of 2007 can make a successful, market leading product launch.

Or they have no reason to upgrade. I don't really see my CPU being overwhelmed by anything for at least another year, and I'm not really into e-penis upgrades every two months.
 
Originally Posted by dook43
Now that Conroe has been out for over a year.... the only ones still running AMD in their systems are those who truly believe the AMD of 2007 can make a successful, market leading product launch
.

Heh - there are still SOME of us with "old" skt 754 stuff, too. Amazing as it sounds I still don't see the need to upgrade just yet. Besides, if someone likes AMD and wants to buy a new AM2 system what is that to anyone else? If all you have time to do is look down your nose at someone else for the choice they've made concerning their computing needs you need to go get another hobby - and quite possibly a life. Since this IS the AMD SECTION OF THE FORUM why is it so surprising that most of the people here are excited by this news? Seems pretty simple to me. For all you Intel enthusiasts who have a different opinion - well, there is always the INTEL SECTION you can post in. Gee - how how quaint. Bet you never thought of that.
 
[+Duracell-];1031321239 said:
To be honest, I thought Steve said that he saw all four cores being utilized in an earlier post.

Yeah, I thought so as well. So much for him putting it in bold lettering.
 
I think the Peeps Here that don't Believe what Steve Posted On this CPU, are Probably the same people that Believe that Man did not land on the Moon......ROFL
Cmon People, Cut AMD some slack will ya?
Nice Pictures though, man i can't wait!!!!
 
I think the Peeps Here that don't Believe what Steve Posted On this CPU, are Probably the same people that Believe that Man did not land on the Moon......ROFL
Cmon People, Cut AMD some slack will ya?
Nice Pictures though, man i can't wait!!!!

Is your capitalization randomizer on?
j/k :D

On topic:
Both machines are running a yet-to-released Phenom processor based on the Agena core. One machine is running Call of Juarez and the other is running Stranglehold. I have personally played stranglehold on an Agena system and watched it fully utilize all four Agena cores!

Note the yellow.
 
Exactly it's like the dailytech link i provided before, this is consumer chips not server stuff, which is indeed BETTER tbh, because I don't care if Barcelona can do 4GHz if it's only server orientated. However until benchmarks and further proof there is always going to be a lot of doubt
 
No its not,its the AMD adoration society.All that same I pray this thing launches at decent clocks,in a realistic amount of time. :(

http://www.overclockers.com/tips01184/

While all credibility washes away in the after math of your overclockers link... You talk about realistic, and then in the next sentance you post a link to that site... a shame really. I once thought you had some dignity....
 
Yeah, what happened to Overclockers? They are the biggest haters on earth. Did AMD burn them with some sort of deal? I know AMD chips don't clock to the moon like these new C2D/Q's, but what's their problem? I've been reading that site for a while just to gain some perspective on the anti-AMD fanbois, but they seem downright jaded. Maybe one of the old t-bred's burned down a house of someone they knew, lol.
 
Despite the fact that I can't wait to see benches of K10, i do wanna wait till launch time and get a surprise when I get home from work at midnight of September 10th - just like I did with K8. It was too good.
 
Yeah, what happened to Overclockers? They are the biggest haters on earth. Did AMD burn them with some sort of deal? I know AMD chips don't clock to the moon like these new C2D/Q's, but what's their problem? I've been reading that site for a while just to gain some perspective on the anti-AMD fanbois, but they seem downright jaded. Maybe one of the old t-bred's burned down a house of someone they knew, lol.

Every week for the past 6 years Ed writes an article explaining how AMD is going out of business. I dont know what the catalyst was, but he has done it like a religion for the past 6 years.
 
While all credibility washes away in the after math of your overclockers link...


Your words,not mine Duby :D

And you are right,all the crediblity AMD had with that 'killer' box went right out the proverbial window after the common sense in that article was read.

Whats funny is if it was Intel or any other company,most everyone would expect full disclosure,and yet here its all swept under the rug,and everyone is just expected to swallow it whole,hook,line,and sinker.Dont shoot the messanger and all that.Ed,nor
anyone else on this,or any other forum that decides to frame AMD's pr actions in a negative or 'un spun' light,should not have to be subjected to persoanl insults.If its raining outside day after day,I'll be the last one to call Ed or you or anyone else a
'hater' for merely telling the truth.

I have dumped thousands into this company,and need not apologise to anyone for my opinions,or my views on this corporations actions,past or present. :)

As Anand said,they are doing thier damndest to sugar coat a faulted launch.
 
i dunno abt overclockers and whatnot, but this lack of benches is just inexcusable...

it just totally reeks of the 2800xt launch.... we got no performance numbers, and so many people held off upgrading because they were waiting to see what AMD would come to the table with before they bought..... only to be horridly let down.....

I mean, just give us the numbers already! and not spec numbers, anything but those.....
 
Both machines are running a yet-to-released Phenom processor based on the Agena core. One machine is running Call of Juarez and the other is running Stranglehold. I have personally played stranglehold on an Agena system and watched it fully utilize all four Agena cores!

Note the yellow.
It reads like he has played the game on a similar Agena system in the past, perhaps at the AMD CTO Summit. But not during this 3 GHz event.
 
Or they have no reason to upgrade. I don't really see my CPU being overwhelmed by anything for at least another year, and I'm not really into e-penis upgrades every two months.
Yeah I think a lot of people around here would be surprised by how infrequently some of us upgrade. I'm still running a 2600 AXP (on my primary desktop), and have no intention of upgrading until sometime in mid 2008.
 
It reads like he has played the game on a similar Agena system in the past, perhaps at the AMD CTO Summit. But not during this 3 GHz event.

Yeah, reading it again it does sound weird. You'd think he'd say he played it on the Agena machine and not an Agena machine. It does read like he didn't actually play it on that machine. I seriously doubt AMD was faking the demo but I wonder if anyone was allowed to touch it. Oh well, all will be revealed with time.
 
Why does everyone try and read between the lines, some times a fucking duck is a fucking duck and not a screwy kangroo in disguise.
 
Why does everyone try and read between the lines, some times a fucking duck is a fucking duck.

I don't think any reading between the lines is required. He said he played on an Agena machine. Now, I'm guessing it was just excitement near the end of the paragraph and he was boasting that he can now claim to have played a game on an Agena machine. However, it does read odd at first glance. Not that it really matters though, the proof of the pudding is in the performance and we won't know for another couple months.
 
hmm, upon second reading that is sort of odd.

I think he meant to say "the Agena system" instead of "an Agena system?"
 
Back
Top