AMD shows off 3.0 GHz Barcelona with 3x2900 XT?

Leepox said:
osbourne effect

if it really is that good then amd consumers would stop buying their products. I think the reason they have been so vague and that they have delayed this a lot is because maybe they want to close the gap between the time of release for barcelona which is for servers and the phenom which is for desktops. Imagine, if they released barcelona now and people see how good it is, everyone would stop buying their current line up. Since they cannot buy barcelona as it is a server processor, then that halts a lot of revenue. And if the phenom comes out months after, imagine how much financial damage that could make to AMD.
Click to expand...
Very true, and as I mentioned in another thread this time it's a do or die situation for AMD, so the OE can cripple them to the point of possibly never attaining the market share they enjoyed in the recent past. That scenario became manifest upon Conroe's release, and Intel was stuck with its previous product line. AMD maybe very conscious of that happening to them, but with more serious consequences due to its current vulnerable financial position. It is indeed a very crucial year for them, and nothing should be overlooked that could cause potential harm.

Interesting that the AMD detractors fail to mention this and other possibilities and instead foment rumors of an impending Barcelona/Phenom stillbirth.

So i do think that all of this delay and teasing is purely for marketing reasons. If it wasnt because their cores are so bad, then I swear im gonna force myself to buy a bad AMD product even though intel has got the lot :D lol
Click to expand...
LOL, funny.
 
Leepox said:
osbourne effect

if it really is that good then amd consumers would stop buying their products. I think the reason they have been so vague and that they have delayed this a lot is because maybe they want to close the gap between the time of release for barcelona which is for servers and the phenom which is for desktops. Imagine, if they released barcelona now and people see how good it is, everyone would stop buying their current line up. Since they cannot buy barcelona as it is a server processor, then that halts a lot of revenue. And if the phenom comes out months after, imagine how much financial damage that could make to AMD.
Click to expand...

This is a non-issue.

Intel's Woodcrest and in particular, Clovertown, has already relegated AMD's lineup to brown banana status. It's not as if people are still buying AMD products. If anything, impressive Barcelona benchmarks would hit Intel hard, not AMD.

Intel went on the offensive, successfully selling off their NetBurst inventory by moving the Pentium D name down to the Celeron space and making it extremely competitive against single-core Athlon 64. Intel showed off Conroe benchmarks, took a hit, but knew Core had no competitor and that their ASPs would eventually rise and help make up for NetBurst. With the introduction of the Phenom brand, it's clear AMD is trying to follow the same model. The difference is the lack of pre-release benchmarks, which says quite a bit. Instead of going offensive attacking Kentsfield/Clovertown with benchmarks, they're clinging onto K8, increasing clock and still trying to compete with Conroe. It suggests AMD isn't confident Barcelona will deliver the same shock and awe and recover ASPs the way Conroe did.
 
JackPack said:
This is a non-issue.

Intel's Woodcrest and in particular, Clovertown, has already relegated AMD's lineup to brown banana status. It's not as if people are still buying AMD products. If anything, impressive Barcelona benchmarks would hit Intel hard, not AMD.

Intel went on the offensive, successfully selling off their NetBurst inventory by moving the Pentium D name down to the Celeron space and making it extremely competitive against single-core Athlon 64. Intel showed off Conroe benchmarks, took a hit, but knew Core had no competitor and that their ASPs would eventually rise and help make up for NetBurst. With the introduction of the Phenom brand, it's clear AMD is trying to follow the same model. The difference is the lack of pre-release benchmarks, which says quite a bit. Instead of going offensive attacking Kentsfield/Clovertown with benchmarks, they're clinging onto K8, increasing clock and still trying to compete with Conroe. It suggests AMD isn't confident Barcelona will deliver the same shock and awe and recover ASPs the way Conroe did.
Click to expand...


Some of the things JackPack has said recently:
Wow, this is sad. Basically, the demo systems might or might not have been running at 3 GHz. That shot of 'System Properties' is almost worthless.

I doubt it. On one hand, AMD only allows journalists to take a boring photo of "System Properties." And on the other, he's playing Stronghold and looking at Task Manager? No one else is reporting...

"Uncle! I said Uncle!"
- Hector Ruiz, July 22 2007

I'd say Penryn's launch speeds are right where they should be given AMD's 2.0 GHz disaster.

Intel's got the knife in, and now they're twisting it.

You confuse Intel's excellent operational execution with "rushed" product.

BE for Boring Edition.

The benchmark results aren't that impressive anyhow. A pair of X5355 still handily beats those supposed scores

Customers are not buying Athlons or Opterons. AMD's Q4 and Q1 results already prove that.
Click to expand...

Hmmm... I smell something....
suspect.gif
 
The Osborne Effect: Sometimes What Everyone Remembers Is Wrong

"Adam Osborne was accused of pre-announcing the DOS version of his CP/M line -- in fact, he told reporters there wouldn't be a DOS version for at least a year because it was too expensive for his price-sensitive line. He announced the Executive -- the follow-on model to his successful first Osborne -- one month before delivery, and sales dropped in half for a couple of weeks while inventory cleared out. Then the Executive actually went on sale -- and sales dropped, from about 10,000 a month for the original model, to essentially zero for the new model.

The reason for the drop was that the Osborne Executive was not competitive with the Kaypro, a slapped-together rival priced at a couple hundred less ($1,795 instead of $1,995 for the Osborne 1), but it had a much larger screen -- 9 inches compared to O1's five inches.

The Executive came out, much better built, more manufacturable, but with a mere 7-inch screen. There was deep disappointment among Osborne fans.... Buyers walked away."


There are lessons to be learned from what really happened with Osborne. But the mythical Osborne effect, even if interpreted the way people think they remember it, doesn't apply to this situation. If people found out AMD had an excellent product on it's way, high end buyers and enthusiasts, most of whom buy Intel, would stop buying Intel and wait.
 
All this AMD showing off and still NOTHING on the shelf. What a joke,step it up already AMD,im tired of waiting.
 
Customers are not buying Athlons or Opterons. AMD's Q4 and Q1 results already prove that.
Click to expand...

That's pretty funny in light of the fact that AMD actually gained back a small percentage of market share in Q2.:rolleyes:
 
BigMacAttack said:
That's pretty funny in light of the fact that AMD actually gained back a small percentage of market share in Q2.:rolleyes:
Click to expand...
Hmm... Seems to me a lot of people are unaware that AMD has its hands in far more markets than simply CPUs or graphics. Someone should run down the list for JackPack and the other prophets of doom. Care to take a stab, morfinx?
 
APOLLO said:
Hmm... Seems to me a lot of people are unaware that AMD has its hands in far more markets than simply CPUs or graphics. Someone should run down the list for JackPack and the other prophets of doom. Care to take a stab, morfinx?
Click to expand...
Nah, I usually try not to get into "discussions" like that :p
 
BigMacAttack said:
That's pretty funny in light of the fact that AMD actually gained back a small percentage of market share in Q2.:rolleyes:
Click to expand...

Sure, a 2.5% gain in revenue share at the cost of a $600m net loss. At the same time, AMD had to write off $30m worth of old inventory.

Feel free to interpret that as demand for AMD products instead of a fire sale of brown bananas to emerging markets.
 
APOLLO said:
Hmm... Seems to me a lot of people are unaware that AMD has its hands in far more markets than simply CPUs or graphics. Someone should run down the list for JackPack and the other prophets of doom. Care to take a stab, morfinx?
Click to expand...

Such as their consumer electronics division? Losses shot up from $4m to $22m
 
Sure, a 2.5% gain in revenue share
Click to expand...
Ok let's stop the quote right there. Actually they gained in MARKET SHARE. You previously said "Customers are not buying Athlons or Opterons. AMD's Q4 and Q1 results already prove that." Now when proof comes to light that you are wrong in your assertion you change the subject. I'm refuting what you said..."Customers are not buying Athlons or Opterons..." I'm sorry to say you are wrong. Now, operating losses and all that are no picnic to be sure, but that's a separate subject and not the subject of your assertion. Stay on topic. The fact is people ARE buying their products as evidence by an increase in market share in Q2. It remains to be seen if they can continue to keep the momentum going. The problem of profitability is a concern, I'll not deny that.

At the same time, AMD had to write off $30m worth of old inventory.
Click to expand...
This is not a unique situation to AMD. It seems to me that Intel had to write off a major loss on their old surplus some time back also. Both companies do it so please don't sound so smug.
 
JackPack said:
Sure, a 2.5% gain in revenue share at the cost of a $600m net loss.
Click to expand...
The ATI debt ceases by Q3 or Q4 this year. I'm not exactly sure when, but that will take a few hundred million off that figure.

JackPack said:
At the same time, AMD had to write off $30m worth of old inventory.
Click to expand...
That was 939 parts that will never be sold now. The reason P4 parts were written off is totally different. ie. Conroe screwed them.
JackPack said:
Feel free to interpret that as demand for AMD products instead of a fire sale of brown bananas to emerging markets.
Click to expand...
No thanks.
 
There are people who will constantly buy the best product from any company when the product comes out without waiting. There are people who will wait to buy a product from a single company only. There are also people who will buy something when they need it and will get the best product for their money from any company.
 
JackPack said:
...Feel free to interpret that as demand for AMD products instead of a fire sale of brown bananas to emerging markets.
Click to expand...

It seems one man's "brown banana" is another man's banana pudding. Better to sell bananas for pudding than throw them out. :D
 
whys the 3 card not connected to the 1st and second?
 
f1y said:
whys the 3 card not connected to the 1st and second?
Click to expand...

It is, there just isnt any good pictures of it, the door is just blocking the view of the front xfire connector. VID0 and VID1 are connected with the rear connector and VID1 and VID2 are connected with the front one.

BigMacAttack said:
It seems one man's "brown banana" is another man's banana pudding. Better to sell bananas for pudding than throw them out. :D
Click to expand...
Brown bananas are for banana bread :yum:
 
debello_64 said:
It is, there just isnt any good pictures of it, the door is just blocking the view of the front xfire connector. VID0 and VID1 are connected with the rear connector and VID1 and VID2 are connected with the front one.
Click to expand...
I read somewhere that the third card communicates through software.
 
APOLLO said:
I read somewhere that the third card communicates through software.
Click to expand...
Yeah I heard that AMD was having the pc play the game, then grabbing it back, encoding the video, then replaying it, with the replay being very close to real time.

If that was actually happening it would be unbelievable. It would probably mean that one of those cards was doing the encoding. (I think they can encode)

I would like to know more about it.

Anyone?
 
strange that I still haven't read an independent review yet. shouldn't the engineering samples be out by now?
 
not just NDA, IMHO there probably aren't any review sites with sample chips at all.... there are no benchmarks out there for us to look at other than what AMD has released
contast with Intel, where even Penryn has been benched already (see anandtech), and you have to wonder if this is just a case where there arent any benchmarks because there aren't any, or *very* few chips out there....
 
brucedeluxe169 said:
not just NDA, IMHO there probably aren't any review sites with sample chips at all.... there are no benchmarks out there for us to look at other than what AMD has released
contast with Intel, where even Penryn has been benched already (see anandtech), and you have to wonder if this is just a case where there arent any benchmarks because there aren't any, or *very* few chips out there....
Click to expand...

It has to be a stranglehold NDA. ES are definately out there. As Anand said, though, the lower clocked chips with early BIOS revisions had heavily patched BIOS' so the performance was shit. I wouldn't think that's the case now. There are certainly samples out there with working BIOS', they've validated their server chips FFS. The NDA must come with a beartrap attached to the junk of whoever signs it ;-)
 
brucedeluxe169 said:
not just NDA, IMHO there probably aren't any review sites with sample chips at all.... there are no benchmarks out there for us to look at other than what AMD has released
contast with Intel, where even Penryn has been benched already (see anandtech), and you have to wonder if this is just a case where there arent any benchmarks because there aren't any, or *very* few chips out there....
Click to expand...

No one really knew anything about G80 before it's launch , and it was the same with R600 too. Just because Intel has loose lips doesn't mean everyone does too.

If I remember right, reviewers couldn't talk about Conroe before a certain date either (July 14th?). Intel was just talking enough to make up for the silence of the reviewers.
 
ryan_975 said:
No one really knew anything about G80 before it's launch , and it was the same with R600 too. Just because Intel has loose lips doesn't mean everyone does too.

If I remember right, reviewers couldn't talk about Conroe before a certain date either (July 14th?). Intel was just talking enough to make up for the silence of the reviewers.
Click to expand...

I think he means that people had benchmarks despite the NDA. We already have Penryn benchmarks and we had Conroe benchmarks (sanctioned by Intel) before the official launch date. I think people are expecting that someone should have leaked benchmarks by now on a chip that everyone is waiting for and should be at least somewhat available to launch partners this close to launch.
 
Yes, thanks for the link Manny, I don't know how I missed it.

Hmm, maybe AMD realizes it would be very difficult to compete with clockspeeds hovering around 2.0GHz. In an indirect sense, AMD might just be able to keep up with the new chips from Intel on a clock-for-clock basis if these benchmarks are accurate. Looks like an average of 8% increase in performance of Penryn over Conroe. I was expecting a minimum of 10-15% increase. If true, it might make for tighter competition combined with the higher release speeds of Barcelona.
 
Apparently AMD are going to release some information about a new instruction set when Barcelona is officially released.
So far it is just a rumour. There are 5 instructions in this new set.

If this is what I think it is, then AMD will be around for a long time yet and Intel will be passing over the performance crown. :)
 
as355 said:
If this is what I think it is, then AMD will be around for a long time yet and Intel will be passing over the performance crown. :)
Click to expand...
Where did you hear of this? Do you know what the instruction set might be? Virtualization optimizations perhaps?
 
APOLLO said:
Where did you hear of this? Do you know what the instruction set might be? Virtualization optimizations perhaps?
Click to expand...

I was thinking reverse hyperthreading... possibly.

I have no idea really, it's just a rumour.

The instruction set is definite, it will happen, but it's just speculation on what it might be. That's why I don't really want to say where I heard it from.
Just keep your eyes out for some news.
PM sent.
 
as355 said:
Apparently AMD are going to release some information about a new instruction set when Barcelona is officially released.
So far it is just a rumour. There are 5 instructions in this new set.
Click to expand...
You mean SSE4A, which is already well-documented?
 
You must log in or register to reply here.
Back
Top