No desktop K10s in 2007?

[Wombat2]

http://www.theinquirer.net/default.aspx?article=40347

Makes the rumoured Barcelona delay seem more likely.

 

[cryogenic]

I really hope Intel won't push back Penryn due to lack of competition or something ... I really want a quad core Penryn when AMD will finaly launch desktop K10 sometime next year and prices will drop (hopefully AMD will be able to trigger some price fall)

 

[croc]

Not that I'd trust Charlie, he's always been a bit optimistic when it comes to AMD :lol:

I think anandtech's take was fairly realistic. Not good, but realistic.

 

[Wombat2]

I really hope Intel won't push back Penryn due to lack of competition or something ... I really want a quad core Penryn when AMD will finaly launch desktop K10 sometime next year and prices will drop (hopefully AMD will be able to trigger some price fall)

Agreed. Have no doubt that Intel will postpone price drops and new models if AMD continues to stumble. Nothing we can do about that though ... AMD has to deliver.

 

[djgandy]

I really hope Intel won't push back Penryn due to lack of competition or something ... I really want a quad core Penryn when AMD will finaly launch desktop K10 sometime next year and prices will drop (hopefully AMD will be able to trigger some price fall)

I read something recently that they are bringing it forward because "the market wants it".

 

[Ycon]

AMD slows down the industry...
... and they slow ME down! I want my st00pid notebook as soon as possible and Penryn would be worth the wait if its not too long, but now >_>

 

[the_vorlon]

I really hope Intel won't push back Penryn due to lack of competition or something ... I really want a quad core Penryn when AMD will finaly launch desktop K10 sometime next year and prices will drop (hopefully AMD will be able to trigger some price fall)

Agreed. Have no doubt that Intel will postpone price drops and new models if AMD continues to stumble. Nothing we can do about that though ... AMD has to deliver.

Pure speculatiion here on both our parts... but.....

Intel has sacraficed substantial $$$ in this price war in terms of forgone profit. IF (large if) Barcelona is indeed late and Penryn is early, why not release it early?

The idea of a price war it to kill, or at very least injure, your enemy. If the enemy is down, not only would you kick them when they are down, you would, if anything, kick them harder.

Remember too, the 45 nano Penryn will be cheaper to make than the 65 nano Conroe, so I expect it's full speed ahead on Penryn for Intel - I expect a November, maybe October, launch actually....

 

[Wombat2]

Yes you have a good point on the 45nm stuff being cheaper to make. However, my point on the price cuts still stands.

Bottom line: Intel (AMD and the neighbourhood kids lemonade stand) will always act in their financial best interests.

Further, on the desktop, I agree its almost inevitable now that Penryn will beat K10 to the desktop.

 

[chunkymonster]

I'm beginning to believe that Barcelona delays from Q3 into Q4 2007 and possible Phenom delays into Q1 2008 just really don't matter anymore. Granted it makes for great debate, discussion, and fanboy fodder here in the forums. But as a consumer, what's a few more months of hurry up and wait? And as an investor, I expected AMD to bleed profits and market share after the $5B buyout of ATI, as well as a minimum of a 2 year recovery time for the new company to gel and capitalize on their synergies.

So, aside from being slightly disappointed that there's little to no new AMD hardware to play with, like Cryogenic said...

I really hope Intel won't push back Penryn due to lack of competition or something...

I too, hope I can look forward to at least one native quad core product.

We'll see...hurry up and wait for it... :lol: 8O

 

[BaronMatrix]

I really hope Intel won't push back Penryn due to lack of competition or something ... I really want a quad core Penryn when AMD will finaly launch desktop K10 sometime next year and prices will drop (hopefully AMD will be able to trigger some price fall)

Agreed. Have no doubt that Intel will postpone price drops and new models if AMD continues to stumble. Nothing we can do about that though ... AMD has to deliver.


At this point there are too man y conflicting stories. AMD is saying they will get Barcelona out this summer. Rumors say "yields are bad and they didn't start a new rev in time."

Certain stories say that it's a major problem, some say it's easily correctable. Between the Inq and Fuad there have been 10 stories that say totally opposite things. Add in the stuff from Computex and I could scream. I will still take AMDs word until it doesn't happen. I mean I have never been in their Fabs and have no idea where Barcelona is in terms of shipping revisions.

Fortunately the server parts have the lowest (max) clocks and multiple reports have said they will clock higher than the Computex speed.

 

[BaronMatrix]

Yes you have a good point on the 45nm stuff being cheaper to make. However, my point on the price cuts still stands.

Bottom line: Intel (AMD and the neighbourhood kids lemonade stand) will always act in their financial best interests.

Further, on the desktop, I agree its almost inevitable now that Penryn will beat K10 to the desktop.

Not really in this case because the initial outlay for 4 Fabs is a heckuva lot to recoup. They are fortunate though to be able to do that. AMD can't do it so they have to say for example limit 45nm to server parts for at least 6 months.

The fortunate thing is that they are only doing a shrink of Barcelona and adding cache. If Montreal (MCM) can come out in time for Nehalem it will definitely be a fight.

Anyway, as far as Phenom, I reserve judgment. I mean FX was due first and is a low volume/high margin part. I can see it happening especially if ATi sales start adding to the coffer.

 

[elbert]

The artical is only about the FX barcys and only the S2, meaning 1207, are delayed till Q1 08. The S1, AM2+, FX by the artical are set for NOV or Dec so thread name is incorrect.

 

[the_vorlon]

Bottom line: Intel (AMD and the neighbourhood kids lemonade stand) will always act in their financial best interests.

.

I actually have a contrarian theory re pricing..... I think all these Intel price cuts are actually going to increase the average selling price..

I know that sounds crazy, but hear my argument....

When you go to the movie theatre, the pricing structure always looks something like this at the concession stand..

Tiny Pop - $3.00
Medium Pop - $3.50
Vast Huge Gigantic Vat-O-Soda - $ 4.00

Now bulk soft drinks are (almost) free when you buy in bulk, so the movie theatre actually makes HUGE margin on that $1.00 bump to the Vat-O-Soda.. It's just about all profit....

Let's apply this to Computer chips..

An E6850 and a E4300 basically cost Intel the same amount of money to make, they come off the same fabs, same masks (more or less) same R & D, etc.

If the E4300 was $100 and the E6850 was $1000, the vast, vast majority of buyers would buy the E4300.

To make the math easy, lets say at a $1000 / $100 cost structure, the ratio was 19 to 1.. revenue on 20 chips would be...

1 x $1000 = > $1000
19 x $100 => $1900

Total revenue => $2900 => average selling price => $145

Let's say that they now cut the price of an E6850 to $300... (remember an E6850 cost the same to make as an E4300)

Lets say suddenly 50% of people bought E6850s...

10 X $300 => $3000
10 x $100 => $1000

$4000 / 20 chips => $200 ASP (versus $145 previously - a 38% increase!)

Even if only 5 (25%) people upgraded from the E4300 to the E6850..

5 x $300 => $1500
15 x 100 => $1500

3000 / 20 => $150 per chip... a $5 ASP boost from the $1000/100 scenario...

In the real world there are many chips and prices, so the scenario is vastly more complex, but I really do think by making super speed chips cheaper, Intel may actually boost it's ASP.

I build a few systems on the side to make a few extra $$$, and I know last year I sold very few "high end chips" most folks wouldn't fork over another $500 to pick up 20% in speed.

I know today I sell almost all x2-5600s and E6600 or better chips because the price bump versus the performance gain is a pretty good ratio...

An x2-3800 to an X2-5600 (maybe a 35% speed boost) for another $150 is a no brainer for most folks.. same going from an E4300 to an E6600..

People want a better product and if the higher speed grades are actually worth it, they will find the cash...

 

[iterations]

Intel will not postpone the introduction of 45nm products. They will pull it in as far as they can, dictated by fab capacity. Why? Because they have talented engineers who have really delivered, getting 45nm spot on. They are building the fab capacity now and since the initial R&D for 45nm is done a paid for, it is more profitable to use it.

Intel will end up with better products, better energy efficiency, better yield per unit area of silicon, and therefore a higher gross margin.

This industry is an interesting one, the relentless march forward in process nodes actually makes the company more efficient, as long as they are able to absorb the huge R&D costs of the *next* node. As long as 45nm is working (and damn, it sure is) then it is always better to use that capacity.

 

[yipsl]

AMD slows down the industry...
... and they slow ME down! I want my st00pid notebook as soon as possible and Penryn would be worth the wait if its not too long, but now >_>

Don't blame AMD. If it weren't for AMD, then Intel would still be doing Netburst. Between competition from the Athlon X2 series and the ingenuity of the Israeli Intel design team, the suits who brought you Prescott and hoped for 10 gigahertz single core CPUs were finally convinced that Intel's only future was by going back a generation to upgrade decent Intel designs.

I still blame those Intel suits. Now, they don't want a vibrant marketplace, they want layoffs on one hand and price cuts based on the cash they got from all the dudes who bought Netburst Dells for their businesses. If they destroyed AMD, then we'd be back to the usual Intel lack of innovation and vision.

Our other PC is a P4 Northwood 2.8, so I'm no AMD fanboy. I bought Windsor now and am waiting for Barcelona and Agena. Even if it's just below penryn in speed, I'd rather keep the smaller, more responsive company in business than to help bring about a monopoly on the desktop.

Engineers seek innovation, management seeks profits but often decide on stagnation and marketing tries to convince us it's the next best thing.