We've been looking forward to this....

... for the true benchmarks ... Hope, although I'm a avid proponent of Intel, that it is good news for AMD; hope for their sakes no delays and at least be able to compete with Woodcrest... Nehalem is on its way. Again, although I do not support AMD at all, I truely hope they make it; this chip seems to be a crunch-time moment in their history

2Ghz? Pathetic. They are ten months behind Intel.

Will both be present? We must know!

i am sure this will not happen...
probably they will be ready by nov ...

Unfortunately, even though AMD may have better and more efficiently designed CPUs than Intel, AMD's lack of interest toward competing head-to-head with Intel will likely cost AMD dearly for years to come.
Shame on you!
Being second best used to mean you would work harder to earn the purchaser's respect by providing us with better products at great price breaks.
Fast forward to today, August 8, 2007.......unless this next CPU release is big, a solid performer and available to us end users "on time", I think that you will see lots of your long time customer supporters jumping ship and moving to Intel.
Frankly, I couldn't wait forever so after 10 years and many built PCs later (all AMDs, up until now), sadly, I did cross over to Intel and frankly evaluating the evolving PC CPUs availability in the now, I am glad that I did!
I have a great running, lightening fast Intel E6850 3.0GB core dual machine, which I built at a reasonable price.

AMD boys and girls: I don't know what you have up your sleeves, but if you don't wake up soon and realize that Intel is "kicking your ass!", it will be too late!!

I don't expect this comment to mean much to your company, but this is my truth and reality as I see it.
Mike Fry

It comes :)

it'll suck. amd lost their competitive edge and we all know it. it's just boring to hear about a company competing in the budget space.

unless they have some insane price drop, like an X2 6000 at a $60 price point, I know I won't be considering them for a future build. From the looks of things They'll be out of it for the next 2 years.

fusion? fusion flavors are not going to be a high end anything. They're going to be energy saving, low cost solutions. Not sexy for the builder market. Plus, with all the R&D it'll take to get there, it won't be a cheap solution = Not sexy for the average joe market.

logically, it's not reasonable to combine the GPU die of a high end graphics and a CPU die w/ out a massive die shrink AND a massive performance hit. Sure, I think they might be able to get today's tech to shrink / fit / and run stable, but who wants the best of today in 2 years?

I want 3.2Ghz Barcelonas!!

AMD needs to get that FAB process under control and spin out some good fast silicon so we can take on the Intel boys. The AMD Team wants to take back the lead in the Folding@Home wars.

Isn't that a little insensitive?

Im waiting for Phenom at 3Ghz or better, not upgrading my 6000 till then. But its working just fine for me till then.

Let's be honest folks.... isn't this the ONLY premier AMD has in 2007, unless of course you consider the graphics card FIASCO? What exactly has AMD "premiered" in 2007?

I know AMD is trying to get better at the whole PR thing (cooked up monopoly profit report, ridiculous "AMD break-free" website), but they are going to demo some 2.6GHz and 3.0Ghz chips and then at the very end say, oh by the way we will only be launching up to 2GHz "value" chips at this time. You know, because that's what our customers are requesting and we are nothing if not customer-centric...

Demos on the 2.0GHz chips? Did I mention they have very high platformance/watt capability?

The launch is September 10th. They said shipments should begin that day or the next, September 11th. Even if it was the 11th, I don't see where a processor launch has a thing to do with New York or the tragedy of that day. Things happen, we must move on and keep going. I don't expect all companies to be closed that day, so why shouldn't a company launch something that day? Would you suggest we shut the world down that day and all just sit around at home watching memorial specials on CNN crying?

Onto the topic of the article, CAN'T WAIT!!!

Please tell me you're not serious? However tragic 9/11 was, you cant expect the whole world to stop on that date.

Is this another "forecast" / paper launch ala 65nm, 4x4, R600... or will there be actual product for PURCHASE at launch.

This whole we're shipping thing is getting kind of weak... remember the 65nm "shipping" in Dec (or was it Nov?)... no actual product could be had until the end of January.

Let's hope AMD does a better job living up to it's "we don't do paper launches" better than they did when they "launched" the XT2X00 graphics cards.

Wasn't launch originally schedule for the end of Q2'07 or was that only an "introduction", I forget all of this marketing speak....sure does give a lot of wiggle room though - at least Intel now puts dates on their roadmap with product available now (they were horrendous in the past with "launches")

Its kinda late, but we need to see real benchmarking. Intel 45 nm wolfdale already has a benchmark, which shows an increase in performance compared to 65 nm core 2 duos (conroe)... If amd phenom delivers like 50-75% increase in all aspects maybe it would be really phenomenal, phenominal it is, we need to know the pricing... intel plans to price the 45 nms at 999 which is WTF.. amd needs to have lower priced phenoms to compensate with the very late introduction to the market...

AMD is going to make people buy Barcelona's in bulk, to prepare themselfs against the Terrorists(Intel), Red Alert and everything included.

If you read the article you'll see that it says the desktop version of the architecture (phenom) will be released 60-90 days after the initial launch.

Ditto on what pink said, yes it was tragic but where not going to all put our brakes on for a day because of it.

The timing says it all; that the launch will not be noticed as much as a breakthough should be, which leads me to believe that the product will not be very compelling at this time.

Will it be Setember 9 or Septemeber 12? They should be able to launch 3Ghz Wolfdale with no problems out of the gate which will kill this product on almost everything except "pure" multi-core tests that never go outside the caches.

Someone let GIANT OYSTER in on the secrect that it's not all about clock speed. Hell the Core 2 Duo at 1.8ghz was easily out performing 3.4GHz chips. It's not about the Hz anymore.

It doesn't suck intel dust, which it will, Intel will just use some of that massive headroom and kick AMD a$$, and AMD will have no overclockability, so they'll be screwed. AMD sat on their X64 bit behinds too long and no amount of "Barcelogna" or "Phenom will catch up. Their is no way that it is going to be that much faster, not to catch up. Just luuk at their excuse for a graphics card and see...

...i've read somewhere.

For the experts out of you saying "AMD will be out of (high end) business in 2 years..." i may just recall the past development pattern.

With the Pentium II and Pentium III, they were superior to AMD's K8 in everything.
Then the Thunderbird appeared out of nothing and it had superior game performance at quite less cost than Intel's CPUs. This was probably the time most AMD supporters got on the ship. It took Intel nearly a year to come out with a new technology, when they released the Pentium 4, the Athlon XP was almost introduced. Again no chance for intel to re-conquer the enthusiastic home-users hearts. Only it was Intel's luck that they had good exclusive contrats with dell and many more resellers. With the introduction of the Opteron CPUs, AMD also seemed to seriously outpower Intel also on the server-market with a much better cost/performance ratio.

It took as long as to the Core launch for intel to really get the #1 back. AMD fought well competing against the Core 2 Duos with they 2 years older architecture of the athlon x2, i may just note that the x2 6000 (with 90nm) is nearly at the same level with the C2D E6600 in *most* benchmarks (i know there are things like video encoding in which it isnt) but i find that quite noticeable. Where the C2D are the better choice, AMDs CPUs arent the total looser, as which many of you seem to see it. AMD was second, not more or less to meaning of this. There will always be a second one. And if AMD can't change it with the K10 over the penryn (which i frankly don't believe they will) they might do it with the 45nm in a year.

If AMD will then get the #1 back, nobody will ask how long it took to get the crown back. No one asks now how long it took Intel to get it back with the C2D...

The title says it all.

Stop hating AMD. I'm fascinated with the fact that while so many people are saying that AMD has lost its competitive edge don't realize what they're saying. AMD 6000 is only slightly behind E6600 and even E6700 with about a 35% price advantage (way more for E6700). Also, please come back to reality, most of the buyers of either processor WILL NOT overclock them. So there goes Intel's advantage in terms of overclocking. Then, if you also consider recent reports and benchmarks of a number of different GPU's and CPU's, you also realize that a lot of current games are more GPU intensive (gamers appear to be the loudest speakers of all in the mess). On top of that, AMD builds it all on 90nm technology, while still delivering decent power usage. So if Barcelona and Phenom DO live up to the expectation, than AMD is back in the game (for me it is hard to say that it’s not in the game right now). And with more price cuts coming within next few days, Intel will begin feeling even more pressure.

On the personal note, I have postponed building a new system for over 6 months, because I wanted the market conditions to stabilize. Hopefully Barcelona and Phenom will bring that as well.

What is this "Too Little too late?"

What does everyone listen to JoJo.

Amd better be coming back into the lead with this cpu.

K-TRON

The average ASP for AMD is now $66... that's average across desktop, mobile AND server. (Intel is @$118 for comparison)

They may be in the game technically, but financially they are not viable at this point in time. Even if Barcelona does well, server shipments account for ~5% (by volume) for AMD - factor in ramp rate and Barcelona will DO NOTHING for AMD this year (other than PR)... AMD (Ruiz) has even acknowledged this. Until volume in desktop arrives (Q2-ish?), AMD ASP's will continue to be in the dumps...

As for Bald Eagle - Intel has cut Core 2 prices twice in the year since it's been released which is about normal; market conditions are unstable as AMD has cut prices on desktop countless times and then claims they are stopping the price war - only to cut prices AGAIN not more than 2 weeks later! Is it better to sell 2 chips at a small profit or 5 at a loss? AMD's ASP has fallen ~33% in 6 quarters, including 50% on desktop. A 2.0GHz Bracelona is not going to stabilize the market (a 3.0GHz sometime in H1'08 might, but that's more than 6 months away and not a certainty)

As the title says, the X2 6000 is now competing directly (from a price perspective) with the E6750, which trounces it.

the admins are censoring today....

[We remove only foul language and insults. Be respectful to others and your post will be posted. It's as simple as that -ed]

plenty of posts "slip" through with "wtf" or f#$@, so it's pretty selective and arbitrary..

Firstly, I really think that AMD will not be able to ramp up to 45nm within a year or even two: Well known fact is that Intel is 18 months ahead in manufacturing against competition, especially volume production (Look how fast the switch was from 90nm to 65nm: They said they would be releasing 65nm at a given date and it almost seemed they stop producing 90nm at that date as well); AMD haven’t even properly ramped up to 65nm. And the retooling required for 45nm would cost an enormous amount of money, something AMD simple do not have now (And are so bad in debt, that I don’t think they will get a loan for that either, which is a shame though)

Furthermore, yes, most users will not overclock their CPU’s, especially server CPU’s (And I which people would stop commenting on overclocking server CPU’s), but being able to overclock Intel CPU’s to the extent that they are, means Intel can easily scale clock-wise to surpass a threat by AMD with only Core 2.

Lastly, with the shrinkage going on at Intel, what prevents them of adding 2 or 4 more cores to a CPU package like they did with Quad Core 2 until arrival of Nehalem which is a native 8 core (Even if it is not native and just a quick shortcut to undermine Barcelona)? Even if AMD do catch on with both 45nm and scaling to 3 Ghz, at that time, Nehalem would be just around the corner…

PS: Even though the AMD 6000 is only 90nm and the E6600 is 65nm, the AMD has a clock advantage of 600MHz (Which has a far greater effect on performance than only one feature size generation gap between 90nm and 65nm)

Your statement: “And with more price cuts coming within next few days, Intel will begin feeling even more pressure” is quite off: AMD burns cash to stay in the game, Intel on the other hand makes a profit. Quite a nice one too, high manufacturing yields at 65nm (Probably already at 45nm also), have a LOT of cash in the bank… I fail to see the “pressure”… There’s nothing at all!?

If Intel were to start selling at a loss like AMD, they can do it for MUCH longer than AMD can hold out for and certainly then completely and utterly obliterate AMD financially (Especially considering AMD’s financial state now). Pressure!?!

is posting, I can't see how they can continue. They are a year behind Intel in terms of development and it seems this will remain so. The only way they can stay in business is to find a niche market, such as provide a quad-core chip with 2 CPU cores and 2 ATI- type GPU cores on the same chip. This way, at least they can stay solvent while producing CPU's for economy level PC's.

I believe someone at AMD already mentioned they wanted to manufacture these types of CPU's. This is actually a good concept. As long as they have high quality mobo maker partners who can provide 2 (dual) analog/digital and 1 HDTV/S-video out connections, I think this concept will be successful. The customer would not need to purchase 1 or 2 video cards, thereby saving at least $100 and reducing system power requirements. Also, there would be no video card fan, reducing noise levels. Since the memory controller and 2 GPU's would be right on the CPU chip, there should be a speed increase, even if the ATI GPU's were not the fastest type.

Since most PC users are not fanatical gamers, the above system should be an acceptable setup. Average PC users on a budget would give serious consideration to such a system. The only problems are, can AMD produce such a quad-core CPU in time before they bleed to death with the massive losses they are incurring and would mobo makers be willing to provide a high quality product?

What many people fail to realize here is the true benefit of the AMD quad cores is that they perform as one logical core, this means that single threaded applications (around 90% of programs) will be able to use the power of all four cores. Intel's multi-core processors show up as 4 logical cores, meaning that programs have to be multi-threaded to harness the full power of the four cores. What this means in the end is that Intel quad core processors with their superior clock rates will be better for multi-tasking, however AMD quad cores will be able to dedicate huge amounts of processing power to single threaded applications (like video games). So it becomes a question of what you are using your computer for.
-Audrais

"What many people fail to realize here is the true benefit of the AMD quad cores is that they perform as one logical core, this means that single threaded applications (around 90% of programs) will be able to use the power of all four cores"

Single threaded applications will use 1 core and the other 3 will sit idle... the one nice design feature of the K10 is the split power planes (assuming you buy a newer board that has this compatibility), so the 3 cores that are unused can go to a lower power idle state.

What you described above (taking a single threaded program and using 4 cores to process it) was speculated ~1 year ago and given the ridiculous name "reverse hyperthreading". While not theoretically impossible, this tech doesn't exist commercially yet and certainly is not part of K10. A single threaded program on a K10 quad core will run as fast on a K10 dual core (or a K10 single core if they make one)

If Intel was so good how come AMD have had the lead for 4 years? Because AMD did their homework and came up with something better. Intel is not in a rosy position..ask all the people they laid off. Neither is Intel off the hook with core improvements..that quad core processor they have is not a native quad core, and neither does it have a low latency, onboard memory controller.

William,

Spare us the 'native quad core' marketing spew.
C2D trounces AMD's designs without being 'native quad core' designs or having onboard memory controllers...

...the fat lady is hoarse.

Firstly, what AMD did have a better CPU than Intel, I never denied that (Well sort of, if you don’t take into account that they had the Pentium M, from where Core and Core 2 Duo is derived, which is quite on par clock for clock with AMD, although maybe not a desktop CPU). Why don’t you mention how many years Intel was on top since x86 came out? People always seem just to remember failures, isn’t it?

To the people that Intel then lays off and re-hires and your statement: “Intel is not in a rosy position”: It seems to be an acceptable practice industry wide, including IBM, AMD etc. (Although I do not agree with this practice). That’s how they keep themselves lean. As for their rosy profit, in the period from 1996 to 2006 they’ve made AT LEAST $1.3 Billion with an average of $6.1 Billion per year (See: http://www.tgdaily.com/content/view/33192/118/); how that not rosy?

We all know that the quad Code 2 is not a native quad and have a integrated memory controller, Intel weren’t due for an architecture upgrade, AMD was. Why should Intel through their resources behind that now? They don’t need to especially if Barcelona can be matched and beaten by what they currently have (Which seems to be the case with preliminary benchmarks AMD released, see: http://blogs.zdnet.com/Ou/?p=510 , but we’ll have to wait and see as I’ve said in my first post. Tom’s also had a similar story, but I can’t find the article)

Intel needs to use their resources of Nehalem, a NATIVE 8 CORE (When will AMD release that?) with a new version of hyper threading giving 16 threads and your integrated memory controller in the second half of 2008 (See: http://www.tgdaily.com/content/view/31408/135/). Care to qualify your statement “Neither is Intel off the hook with core improvements”?

Again, Barcelona can be beaten by hook or by crook with manufacturing superiority (More non-native cores) or clock scaling (Shown by how overclock-able the Core 2 is already) until Nehalem comes out, which is around the corner. AMD isn’t properly ramped up to 65nm, needed for Barcelona, yet (Expect low volumes) and may only be when Nehalem comes out (Bad news for them). Also, AMD cannot really afford retooling for 45nm. Care to comment or deny these statements (Which I was making)?

I posted a reponse about over clocking and don't see it here...

What happened?

[The post is on hold. As a reminder, we do not release any posts that we believe go beyond a certain comfort level in terms of respecting other readers. -ed]

I thought Nehalem was native 4 core? scaleable to 8 meaning you can put two native 4 cores like they have two 2 core quads now.

I wouldn´t say Nehalem is behind the corner since it is still a year before Nehalem is launched. Year and a 1Q if you are waiting okto cpu. a lot can happen in that time.

I think most people are more intrested in quad cores than dual.Also there are many people who know hz means nothing. Now they think more core, more performence. While intel beeing only one to deliver quad might affect amd:s market share. I also think amd 6000 is good cpu atleast in short term. In the long run no one can truly know. Games might start to use 4 cores like alan aware desiners have said that their game will use 4 cores. So it seems intel chose better path to not rush "native" quad. This is the miss step that will cost amd a lot, but like they say you can´t always win.

“I thought Nehalem was native 4 core? scaleable to 8” – All information that I could find is that 8 cores will be on a single die, therefore I’m certain it’s native (Clarification/Confirmation from Tom’s would be nice, but: http://www.anandtech.com/cpuch...i=2955&p=3)

“I wouldn´t say Nehalem is behind the corner since it is still a year before Nehalem is launched. Year and a 1Q if you are waiting okto cpu. a lot can happen in that time.” – Certainly a lot can happen, but again, argue the merits of what I said in my post (@ William - RE: FAO Nixo - Pt 3); it seems that Intel will be able to counter any attack by AMD (And possibly even beat) for the interim.

Lastly, GAMES ISN’T ALL! I wish that all these gamers would get themselves PS3’s, Xbox 360’ and Wii’s! Computers are not gaming consoles, why do you think gaming consoles exists? By far, most computers are used to do business/research; WORK. I don’t mind, actually encourage PC gaming over consoles, but it tick me off when gamers expect computers to be only designed with that in mind; even people making stupid statements of wanting to play Oblivion on a Sun UltraSparc T2. Grow up.

The processor architecture war is currently on Intel's side. However, AMD is the only reason you guys have good prodcuts from both cpu makers.
Support AMD even if they are a bit behind. AMD brings competition to the market and is making Intel to be more creative and honest.
If there was only Intel, we will not see such a brilliant and priced products.

Computers can do nearly everything
You can play almost all console games on a computer using controller adapters and images.
You dont need to buy console games anymore.
With the power you have with a great computer set up you can do anything. I can play SNES games on my computer while using a SNES controller and images of the games on my computer.
I also know of some friends of mine that play xbox games on their computer using the same kind of setup.
Sure you can't make the game any better on a computer than a console....oh wait.... you can. I can crank up the image quality and antialiasing on a desktop if you have a high end graphics card, making all of SNES games look much better and text much more easy to see.
Sure i like to play console games now and again but i know that i dont have to buy a console anymore to play those games.


and to say something about the subject Intel will lead the cpu war for a long time unless amd pulls something big out of their butts.
intel has the tick tock method that they supposedly are going to stick to, you just can't compete with that.
OH, and if was going to annihilate amd i would cut prices on core 2 duos and quad core processors (INTEL)so nearly everybody would buy intel and not AMD. LAST SENTENCE DIRECTED AT INTEL'S MARKETING DEPARTMENT. just a suggestion

Here was lil old me. sitting here thinking that when the dual cores 1st came out.. AMD was the 1st to bring out a speed demon which also had efficiency. Or was i wrong about using ALL P4's as toasters :S

Oh, and need we forget that it took Intel to bring out the core 2 duo to match AMD with there dual cores... which was how many years after the release of dual core cpu's !!

ATI/AMD's chips are cheap, powerfull, less power consuming, more efficient and did I mention that they cost way less? Come on. No gamer needs a QX series chip that costs well over a grand to do what most of a 200$ chip can do on two cores. Intel might have an edge in terms of cores, but they eat power, run hot, the memory controllers are still ancient, leading only good performance in office type tasks, and they are ridiculously high priced. I'd rather get a 200$ chip then a 350$ one with the same or worse specs, and put that 150$ to a new physics card.

it is pretty interesting to see how people dump on AMD so quickly. their chips that are available beat out the comparable intel chips at that price pt and they try to keep it that way. ok they don't have the fastest chip and that might be disappointing but the reality is most people are not buying the highest end chip.

"2Ghz? Pathetic. They are ten months behind Intel."

Judging by the architectural changes, a 2GHZ K10 will probably still be competitive against a 3GHZ K8

"2Ghz? Pathetic. They are ten months behind Intel."
what about if the barcelona with 2Ghz beat your might clovertown with 3Ghz?

AMD = Al-Most-Dead

intel was never as far behind AMD as AMD is behind intel now. Plus, AMD is in financial trouble and there's nothing on AMD's map that says it can compete.

Intel has some reserves it can employ to stay way ahead of AMD in the coming years. It can die-shrink faster, cheaper, and on a larger scale. It has healthy overhead for higher clocks. It can integrate a memory controller. When apps become more multi-threaded in the coming years, they can even have each core hyperthread. Not to mention some of the interesting things coming from left field like Photonics, in that they have invested tons of R&D money.

I'm quite surprised by Penryn's benchmark results. 5% average increase (maybe 10 after proper BIOS updates) seems too little. This means that if Agena edges out Conroe (and Opteron edges out current Xeons) its performance numbers will be right up there with Penryn. Now, Intel has higher yields so it will use that to its advantage but I'll put my money on the fact that AMD will have better performance/watt and attractive prices. So the good news is it's going to get competitive once again, the bad news is I doubt AMD will steal the performance crown. AMD, you better ramp up those 45nm Barcelonas fast!

Wow, lots of AMD performance complaints. It's a good thing Cray and other supercomputer builders look at the facts and not blogs.

Here's a link to the worlds fastest supercomputers.
http://www.top500.org/list/2007/06/100