AMD priorities

Very good questions/thoughts. I suspect that the one gain that AMD could realize in showing early benches is that it could delay future pruchases of Intel processors by those willing to wait. Much like Intel did with revealing benches ~6 months before it launched Core2 and attempting to achieve the same result. The one caveat here is that, as you imply, there has to really be something credible to brag about, otherwise when the product comes out, credibility is lost.

Call Hardline much?

You'll get it if you look it up, or watch late night tv.

Well, lets see:
-Because K10, AMDs only hope to outperform Intel in the near future, is going to be 4 quarters behind in C2D in market introduction.
-Because K10 is going to be released 2 quarters beyond the 'grace' period C2D needed to prove it wasnt another piece of crap like netburst was, thus setting up Intel to start taking back market share from AMD.
-To forestall the above mentioned market shifts to Intel until AMD can get K10 to market. In quantity.
-Because AMDs margins are so thin right now they must maintain volume sales to compensate.
-Because AMD doesnt have a lead in anything to help them hold on to those shares they so desperately need.
-Because Hector Ruiz doesnt want to lose his job
-Because AMD doesnt want to see its stock prices fall any lower

Shall I go on? There are lots of good reasons for AMD to lie about K10. Does that mean AMD is lying? No. But dont say they have no reason, because they do in fact have several.

You know, you are absolutely right about credibility. Credibility is worth something for a company. So riddle me this, if credibility is so important, why did AMD put out 'quad core for dummies' mocking Intel for its 'glued' die quad core, then release a 'glued' mobo quadcore which did not live up to the hype AMD generated for it? Why was AMD digging up obscure/meaningless metrics for K8 PR after the release of C2D? Why was AMD keeping its mouth shut tighter than a frogs behind when they couldnt supply either the channel or the OEMs last year? Where, exactly, is the credibility in that? And what about the credibility here? Why was the horde itself trying to build expectations for AM2 and 65nm beyond what AMD ever claimed? Did AMD ask them to? No. Did they help AMD? No. Its funny that your talking of credibility because the Horde has made crap up as they went along the past 7 months trying desperately to validate AMD in anyway they could, yet even the most hardcore Inteliot freely admits Netburst was crap. I bring this up because I can help but wonder how the Fanboys credibilty reflects on, or is a reflection of the credibility of their chosen company.

Its very funny that in one sentence you make a statement about the importance of credibility, yet in the next sentence, you are saying there in no gain to be had from early testing.

Test results validate performance claims.

Validated claims = Facts

Facts = Credibility

So, which is it hal, they need to be credible, or they dont need to be credible?

Through testing, K10 could gain the credibility you were just talking about. If it is as good as AMD claims, then why not test it? They could forstall market shifts to Intel with validated claims, but without the validation that testing provides, those claims are nothing more than supposition. When it comes to sales, a bird in hand is worth more that 2 in a bush. Without testing, K10 is 2 birds in a bush. Right now, C2D is in hand.

Yes they could, but guess what: Its already too late for that.
1) There is finally enough info out to negate any advantage secrecy could have brought them.
2) Intel has far more R&D funding, and, contrary to what Hector and the 'boyz' at AMD are claiming, I highly doubt Intel is "behind" AMD. Yes AMD got to 64B first. But Intel beat them to 65nm and will beat them to 45 nm. And by the time AMD is just (maybe) releasing its first k10 45nm products, Intel will have a new Uarch for its 45nm node.
Further, as far as Uarch 'secrets' go, I suggest you go read the Anand K10 article very closely....you may be suprised at what you find.

Which has exactly what to do with the point of your post? Do you beleive this is a revelation? Its not. Do you think there is anyone out there who cares about CPUs that hasnt already known this for 3 years? What are we supposed to be taking caution against? The "dangerous" Intel? Just what exactly is your point? That the "morally virtuous" AMD is above that sort of nonsence? If you believe that, you had better go check out the PR/marketing shenanigans AMD has been up to since C2D was released.

Well, in that case I guess you should run out and buy yourself a fleet of fords/gms/chryslers/Sony PS3's/retail box Vista Ultimate/Netburst's, TV Wonder Knives, George Foreman Grills, body wash, etc etc etc, because all the companies that made those said "yes, our product is good".
Nothing is good until it has been proven good. By your rationale, Netburst should have been great because it looked great on paper, and Intel said it was great. Too bad for Intel the reality was that Netburst was crap. Interestingly, according to the first actual indepth look that has been provided for K10, it looks great on paper too. Better hope it doesnt turn out to be AMDs Netburst.

You have no idea just how wrong you are. That is exactly what AMD needs to do. In case you didnt keep up with the times, THG, Anand, HotHW, HWzone etc etc etc and their respective forums have sold more AMD products than AMDs marketing/PR division has. Hell, the people who frequent this board right now, Vern, Tanker, Ches, Clue, Exit, myself just to name a few, have probably sold more AMD DTCPUs over the past 4 years than AMD marketing has. Its called 'word of mouth'. It comes from providing a better product to people who know enough to actually know that it is (or rather was) better. And those people providing advice or products to people who dont know better. If you build a better mouse trap, the world will beat a path to your door. They did for K8. Now they are doing it for C2D. If the mousetrap that is K10 proves to be better, people will buy it. If not, people wont buy it and AMD doesnt have the money to pull the marketing coo Intel did with Crapburst. And guess what else...if K10 doesnt perform, the only review site/forum selling it will be AMDzone.

AMD, or the horde, or the horde accolytes, or the horde wannabe's saying Barcelona is "a beast" or will outperform C2D, or RoXoRz doesnt mean jackshit. You dont win a race by showing up to the track and saying..."well, when I get my car done its gonna do the quarter in 11.12"...You actually have to bring the car to the track and run it. Right now, K10 isnt winning any races. Its not even in qaulifying yet. It doesnt even have tires.

AMD does need to send out ES's and they do need to be tested if AMD wants people to hold off from buying Intels until K10 arrives. They need to have their producted tested so their sales force (us, not the morons at circuit city, CompUSA, Office Depot etc) know what it does. They also need to have their product tested to prove to the stock holders that yes, they actually do have a competative product. In fact, there is only 1 reason AMD should not be publically testing: K10 ES's arent performing up to spec/expectations.

Well theres a revelation

Hal,

You are in the wrong forum. This forum is about CPUs. Not commercial morality. The majority of people here want facts.. not hearsay. The forum is about performance, not unicorn wishes. Value, not manufacturer loyalty.

If it was any of those things, it would be called the AMD CPU forum, or the Intel CPU forum, or the Manufacturing Morality forum.

You keep trying to promote AMD and denegrate Intel, and you keep getting more and more subtle about it, but no matter how subtle its wording, the bias of a biased message cant be masked. Heres a thought..why not give your fingers a rest and let AMD promote itself through its products and performance. Right now K10 is not a product. Its a bunch of articles at Anandtech and other sites, and thats all it is until it hits the shelves. When it gets here, it gets here, not before. If you cant do that, then you should probably be hanging out at AMDzone.

Carnal Funnel what????

Carnal Funnel what????

Can you do that legally in a french bakery?

I have a story anyone want to hear it?

Bring it bro

Tenderizing the prime rib? 8)

Steak sauce?

Once upon a time, a long time ago, in a land far far away, I was walking through a bakery after hours with a buddy who worked there. . It was the first time I had been there, so I didnt know the layout, or about the employees. The one emlpoyee on at that hour was the janitor. He was a special needs fellow in his 30s. As I rounded a stack of proofing trays, I was presented with the site of the fellow pleasuring himself on some....&*^&...
I was about a half step in front of my buddy when I saw this. I froze in my tracks, wide eyed and unconsiously threw my arm out to stop my buddy, hitting him rather hard in the chest. My buddy responded to my action with a "WTF", scaring the janitor off in the process.

"holy #@!$#!@! man, that guy was just &^*&^*&!! off"

To which my buddy blythly replied "so what, he does that all the time"

Now you know where the filling in the donuts comes from.

True story.

I see it this way. It was essential for Intel to get out their benches early. The big engine that couldnt, couldnt hold the server market, couldnt hold the desktop segment, time and again tried to beat out k8 with high energy consuming, high revving (if you will) processors. AMD on the other hand has been gaining ground in the server/desktop market, they were bested by C2D and threatened in both markets, but not for a long period of time. Like many have said before Barcelona is important, essential for AMD. To lie about it or to over embellish would really harm AMD (No.2,trying harder) Being second they HAVE to produce. I also wonder what would happen if one: say theyve had Barcelona done for quite some time (tho they cant produce huge amounts being that the 65nm process change is brand new to them ) and two : bringing out those benches early on, not having the fabs, nor the experience with 65nm. I would say at least here as well as in the business sector (server) that those people would be crying foul. Paper launch, nothing but air yadda yadda and the fanbois from either side would be having a hay day just my 2 c

I agree entirely. Two can play those games.

Personally, I think the reason both Dell and Intel have lost a lot of (repeat) customers is actually all the Intel advertising for P4, and many (not most!, but many) people finally figuring out that they had been fooled by advertising regarding the P4.

These are reasonable --- however, to contrast what you are saying...

Intel put up benches 6 months before release, there is even video of them being run and most all major HW journalist were allowed to run the benches ... yet most assumed Intel was lying ....

What does AMD have to gain putting up benches --- lost customers. They are losing server share, the most important of the three market segments.... they have everything to gain by freezing purchasing decisions, they should show benches.

However, you argument that they have nothing to gain by lying is a good one because if it fails to live up to the claim, the credibility is completely shot..... however, the method by which they are making the claims is subject to interpretation.... SPEC_FP is a targeted bench but not necessarily representiative of real work loads, but this is where they are claiming superiority.... not surprising, even K8's FPU is still strong, C2D only beats it because the FPU is wider. Barcelona will certainly trump C2D in this arena.

Yeah, I remember back when the first benches with the Core stuff was being conducted. I didn't bother to read a lot about it. It become clear the new arch would be good soon enough (over the next few months) (and long before it was widely understood it seems). I remember the objections to the testing, that specifically it was controlled, only on the benches Intel chose. Vaguely, if not confusing with another instance, for a while there were also complaints that reviewers had to make do with watching the intel demo, and could not get their hands on chips to do their own tests. There was concern both about what was in the box, and whether the benches were carefully choosen. All that jazz.

Now, I'm not much concerned about the AMD benching situation, just as I wasn't much concerned with the Intel situation back at that time.

They need to get better info out to curb the loss of momentum.

That could be. I've learned, over time, to take stock market oriented analysis with many grains of salt. Usually, what you get is some trader or analyst basically talking their book. Trying to convince others to join them on the short or long side of a stock, for profit. If they can get enough people to go their way, it moves the stock price in their favor.

So keeping in mind that idea may be one-sided (without info that balances the overall picture), it is still a concern, and no doubt AMD had that press meeting just exactly to work on that.

btw, I enjoy reading your posts

I actually thought it was a very good post.

You've got to be kidding me. Workstation? Don't you remember anything AMD was spouting about 4x4? It was targeted at gamers and enthusiasts, just look at the old press releases and such. Like this one, on the AMD website.


SUNNYVALE, CALIF. -- June 26, 2006 --AMD (NYSE: AMD) announced today broad software developer support for its upcoming four-core, multi-socket enthusiast platform, codenamed 4x4. Leading software developers including Bioware, Cakewalk, Crytek, Havok, Irrational Games, Midway, NERO and Sony have been working on software applications and games to take advantage of multi-core processors and benefit from AMD’s Direct Connect Architecture.

“We’re looking at a decidedly multi-core processor, multi-threaded application future for the PC industry,” said Bob Brewer, corporate vice president, Desktop Business, AMD. “AMD has planned this enthusiast platform so that software developers can design outstanding multi-threaded applications for consumers who demand the very best immersive computing experience.”

Multi-threaded games are among the first applications to see immediate advantages from using multi-socket, multi-threaded architectures like the upcoming 4x4 enthusiast platform and make exceptional demonstrations for new PC technologies.

“With Crysis, our flagship title, we are pushing the boundaries of game-physics, intelligent combat AI and cinematic visuals,” said Cevat Yerli, President and CEO, Crytek. “Crysis will be an experience culminated in next-generation hardware when utilized with a multi-core system like AMD’s highly anticipated 4x4 enthusiast platform.”

I believe the phrase to which you are referring is 'The best-laid plans of mice and men / Often go awry'.

Which is actually precisely what I was trying to say. Sure, 4x4 sounded like it had potential, but it didn't work out. That doesn't make AMD a bad company, just like Netburst doesn't make Intel a bad company. Well, not in itself. Pushing Netburst for so long is another debatable matter.

Exactly, I believe this is the whole crux of the matter. AMD can't get their K10 clocks up so what's the point of benchmarking? A 1.8 GHz Barcelona would perhaps keep up with a 2.1 GHz C2D but the faster C2Ds would smoke it. Sure K10 might be far superior to C2D in performance / watt, but nearly all of the hardware sites focus on raw performance. Benching a 1.8 GHz Barcelona ES now would show it losing to the faster C2Ds and AMD would be in even deeper water.

My guess is that AMD has a very good idea at this point how Barcelona is going to run. And they also have a very good idea what yields and availability will look like at launch time. In short they are following a calculated marketing and business strategy at this point. One that they hope will take consumers pleasantly by surprise.

It's obvious AMD is slowly rolling out the carpet with the benchies on this one. AMD has been very quiet the last few months until this week. But if they reveal too much too soon, the "surprise" will be ruined.

Maybe, but IMO they should be trying to ruin Intel's current Core 2 sales, not protect their surprise party.

If they do have ES chips that just can't run at fast enough clocks yet, they might be better off showing performance per clock on real benchmarks, and not disclose the clockspeeds at all. Just run a Core 2 and a K10 at the same clockspeed and show the relative performance in percentage. It wouldn't be as good as a legit benchmark, that's for sure, but it'd be better than nothing at all. And then when they clockspeeds finally are up to snuff, everyone would see that the old per-clock performance ratings were correct. But I'm just speculating here, it'd be cool if they did something like that.

Hal, in your time here you have already proved that that you wont read. Thats no revelation. Everytime someone counters your biased posts, you ignore it, so theres nothing new there.

How does knowing your "concerned at all with impressing you or proving anything to you" help anything? Do you think I care what you think? You have already demonstrated that your position is set. You are an AMD fanboy. Period. Nothing is going to sway that. You will continue to believe that AMD can do little wrong, is somehow morally superior and will prevail, and you will manufacture or interpret whatever you feel neccessary to support that view, so to debate with you is as pointless as trying to break stones with a water balloon.

Debunking the tripe youve been posting, well, thats another story. I care about the people who dont know whats going on, and may be incorrectly infuenced by your bias. You yourself are a lost cause, so dont kid yourself thinking that I somehow hope to convince you of anything. Everyone else, with the exception of a few other AMD/Intel fanboys, are not lost causes.

Au contrare mon ami. I like performance and value. Again, this is the CPU section. Not AMDzone, Intelzone, Fanboy zone or Jerry Falwells Moral Majority. Just like BaronMatrix, 9 inch and MMM, you assume that people who dont agree with your bias are pro Intel.

So once again, I have to state the blatently obvious.
The bias's in this forum (errant fanboys delusions aside) here include:
Performance
Value
Anti-BS
Anti-FUD

Once again, I must post that for someone who is Intel biased, I have owned an unusual number of AMD systems and shocking few Intels.
Once again, I must post, that a large number of people you are trying to BS, myself included, are actually fans/owners of AMD products. Once again, I must point out that in this forum, perfromance and value hold sway, not good intentions and wishes. For good intentions, wishes and moralizing, you should go to AMDzone.

Just like Baron, your perspective appears to be "If you dont agree with me, you're anti-AMD"

Very curious that you should say that after you made this comment:

I have read hal, and no matter how you couch it, it comes out the same, you prefer AMD, and you will besmirch Intel, however gently, to promote AMD.

I prefer AMD, but I do not besmirch Intel through fabrication and intentful ommision. I also aknowledge that they have created a superior product and it is that product along with AMDs improper responces that has placed AMD in their current prediciment. Not some 'evil' intent on Intels part. Even were Intel malevolent, it would not cahge the fact that their current Uarch outperfroms AMD's and that the trend in the value contest favors Intel (though I havent checked prices today)

Well of course you disagree. To agree would be akin to a defense attorney walking up to a judge and saying, "your honor, my client pleas not guilty. But he/she did it"

You are right that you are not going to sway Jack. He knows far to much. For that matter, you are correct in that you are not going to sway the majority of regulars. Theyve been around far to long, and you are by no means the first fanboy, either Intel or AMD, to come through. They know whats going on. But as I said (which you obviously didnt read)

Hal

Again, please spare us the attempts to shift focus. But you are right, I am obsessed, with anyone who would spread BS or FUD, so please dont feel singled out as you are only one of several.

Hal, please spare us the attempt at shifting the problem to someone else shoulders.

You wrote this, not me. Its your pattern, not mine. Its BS, not warfare. There is nothing to imagine, its printed for all to see plain and simple. Do you really believe, by implying that there is a personal perception problem, that it somehow changes what you wrote, how you wrote it, or what it says? If you truly believe that, then there is a perception problem, but its not mine.

Another of halls subtle anti Intel threads

Hals "your all Intel biased thread"

Hal, there is no projection here. This is the second time you've attempted your 'projection' ploy, and just as with the first time, the words are yours, not mine. There is no projection, only what youve written, how you've written it and the trail you've left behind you. I would think you would examine yourself as you clearly have a personal agenda, are acting to effect it, and are willing to circumvent fact while attempting to present the appearance of reason. Tell me, which term is it that describes those actions?

So far Barcelona is vapourware.

Who is running all the benchmarks? AMD? Not credible enough, since they are bound to be biased.

On a certain level, they are right to disclose certain information because, as you wrote, it's good for business, but remember that those that AMD's statements may affect aren't really the consumers. What I mean is by now AMD seems to be trying desperatly to hold investors trust on their future. Most investors aren't really tech savvy people, and they simply believe whatever numbers they are fed.

You ask what AMD could possibly gain by releasing Barcelona for early benchmarks. My answer is: to raise consumer's expectations on AMD's products.

Intel will hardly steer away from its 2007 roadmap if AMD releases something now. Probably Intel already has much more facts on AMD than we here could possibly know by now. Do not underestimate the power of industrial intelligence.

Back on track: if AMd releases Barc. for benchmarks now, even if they are planning on releasing it on Q3 or Q4, at least they could keep consumers interest on the product, and may hold off purchases and investments in equipment until Barc. is released, whenever that happens.

What is bad about all this is that we - the consumers and people who actually have influence on buying decisions - are being kept in the dark about it. If I were to invest $40k right, having the option to wait for a few months, the longest I'd wait is for Intel's proce drop, because I don't know long I'll have to wait for AMD, not even about their pricing strategy.

So, AMD will be losing that kinda of consumers that need to buy, but can wait, but will buy now because the future is uncertain.

With selective memory and an agenda to pursue, its possible to forget ones own history

Makes sense if you say that the opinions of some in this forum are representative of general opinion. I'm thinking though that average people are more likely to trust the statements from AMD itself. I don't think any *main primary* source, like PCWorld, or such, is likely to try to trash barcelona because of a lack of benchmarks. Meanwhile, average joes see news headlines about barcelona coming and the claims for it, and I bet that satisfies that need you are pointing to.

yes you could, but you shouldn't!

the concept is actually very important, and each and every person, me too, needs to remember that it's easy to label someone and begin to forget they are unique and not like anyone else.

Now...this chip fan needs to go to bed.

Have a good evening.