Intel delays 45nm QC...

[aevm]

My opinion would be that there was an alternative reason for the delay other than 'competition' Honestly, if Intel was able to push the 45nm out now they would, first and foremost they will be getting more proc. per wafer, making more money for them. Remember we are talking about companies with bean counters here, regardless if you like AMD or Intel.

My thoughts exactly. There's got to be some other reason here. Maybe they found their own bugs that need to be fixed first.

 

[OlSkoolChopper]

Ya like I said its probably a combinition of the erata plus a "what the hell" attitude: Clear the shelves since the compitition isn't going anywhere fast.

 

[yipsl]

DAMMIT! I was looking forward to getting a Q9450...

Looking further ahead, this could spell bad news for Nehalem as well. If AMD can't get into a competitive position soon, Nehalem can easily be pushed out to 2009 for the very same reasons...

Well, if all the Intel fanboys get their wish, and AMD dies, then we'll be back to Netburst days in regards Intel business practices. They might even cancel their own fusion plans and go head to head with Nvidia in the discrete GPU market.

I can't understand people buying Intel right now. Sure, they have the "best" but X2 is still quite good and, though Phenom won't be decent before 45nm (AMD should have followed Intel's lead and avoided 65nm native quad core), buying Intel now is like asking for trouble a few years down the line.

Is it all about increments of 10 fps in a particular FPS or 40 seconds in video encoding? I can see the time differences in 3DS Max or Divx making a difference in the workstation market, but at home? All in all, I've bought more Intel CPUs since 1993 than AMD, but only when Intel had the working product. I chose Pentium over K5, but I went K62 because it was a decent CPU at a budget price. I went P4 Northwood over Athlon XP because of heat issues, but I avoided Prescott and Smithfield Netburst and went Athlon X2 instead.

Now, I don't see the real world difference that everyone claims is there, at least not enough of a difference that matters to me. So, I buy AMD. Plus, I'm a bit old fashioned, I prefer pins on the CPU and not the motherboard, and I find AMD stock coolers decent enough at stock speeds (I don't overclock). Now, AMD has the benefit of ATI chipsets and GPU's, and I've only bought an Nvidia chipset and GPU combo once as a barebones offer.

I do wish that AMD had planned to release a hybrid Crossfire in March that actually benefited from power savings, and that worked with more cards than the upcoming 3400 series. As is, I'll just have to buy a couple of 3870's instead and wait to see how fusion pans out by mid 2009.

Xbit Labs on fusion:

http://www.xbitlabs.com/news/cpu/display/20071216231717_AMD_Claims_First_Swift_Fusion_Processor_Due_in_Second_Half_2009.html

“The first APU platform is code-named Swift. It gives you the choice of technologies for high-confidence volume production ramp. We want to re-use as much [IP] as possible to accelerate our quality [qualification] and time to market. So, we have an AMD Stars CPU core, the graphics core that is based on the present high-end discrete GPU core and leverages the North Bridge that is presently found in Griffin, the CPU of the Puma platform. It will be our second 45nm generation product, so the maturity of the [production technology] will be proven. It is done on the current SOI design rules, which is the process that we know how to build on very well,” Mr. Rivas explained.

Initially the company indicated that Fusion processors “are expected in late 2008/early 2009”, and the company anticipated to use them within all of the chipmaker’s “priority computing categories”, including laptops, desktops, workstations and servers, as well as in “consumer electronics and solutions tailored for the unique needs of emerging markets”. A little later the company said that the first-generation of Fusion chips will be aimed at laptops and that production will start in early 2009. This time AMD claims that the actual chips will reach the market only in the second half of 2009, which may mean that the product will only be launched commercially in Q4 2009. Still, the company said that it is minimizing all the risks hopes to really deliver the product on time.

“By optimizing the choice of IP blocks we have less risks and faster time to market in the second half of 2009,” claimed , executive vice president of computing solutions group at AMD.

IMHO, Intel won't do a darned thing unless the market forces them to, and the market right now is AMD/ATI CPU's, chipsets and coming fusion. Nvidia won't be a factor until Intel brings out their discrete GPU's. Expect some Intel dirty tricks where they lock out Nvidia and ATI discrete GPU's on Intel chipsets once they have a full line of discrete GPU's up and running.

So, buy Intel if you want, but don't kvetch too much about their delays due to AMD's failures. AMD has vision and tries for the improbable. They get flamed for sitting pretty during X2 and not putting enough into R&D, but the flamers seldom had a bad thing to say about Intel during the Netburst days.

no matter how you fry this bacon up it leaves a nasty taste in your mouth...and just yesterday i said what a bad time it was for a build. apperantly timing just got worse. merry christmas everyone!!!!

Yes, indeed. We have two AMD systems that I was going to upgrade, but it didn't pan out because of three reasons:

1: The X3870 is virtually unavailable (I wanted the 1 gig version anyways).

2: Hybrid Crossfire won't be out in the U.S. before March, and the first incarnation won't have the power saving features, and won't work with any cards above the X3400 series anyways. I had hoped that, while it gave a 60% improvement with entry level cards, it would give both power savings and a 10% improvement with the midrange GPUs.

3. Phenom is a bust at 65nm. So, I'll just upgrade my AM2 boards with 2.9 gigahertz 65watt Brisbanes first quarter of 2008 instead. I'll wait for fusion at 45nm instead of going Phenom this spring.

I'll still try to find the X3870's for our two systems after the one gig version arrives. If worse comes to worse, I can always get the regular 512 meg version or, as an absolute last resort, two X3850's with 512 megs.

Makes you think that both companies wanted to get rid of old inventory for the holidays. Add Nvidia to the list too. I don't know how many people went for the older 8800GTS when the 8800GT wasn't available. Ideally, old tech should be marked down once new GPU's hit the market in force, but the release of ATI and Nvidia's new cards are sort of papier-mache.

 

[yipsl]

Sorry about the double post, I'm correcting it now.

 

[archibael]

My opinion would be that there was an alternative reason for the delay other than 'competition' Honestly, if Intel was able to push the 45nm out now they would, first and foremost they will be getting more proc. per wafer, making more money for them. Remember we are talking about companies with bean counters here, regardless if you like AMD or Intel.

The "bean counters" comment is spot on, but you missed the big one here: the more volume Intel gets out of 65nm, the more they get to run on equipment and tooling that's already bought and paid for, and half-depreciated. While 45nm saves money in the long run, ramping 45nm to high-volume manufacturing at multiple sites worldwide costs a substantial amount of startup cash. If they can postpone a significant amount of this startup cost because their current product slate is already more than competitive at the older process node (which it is), they can essentially bank the cash they would have spent on the ramp for a couple of quarters and make their bottom line look even better.

Yorkfield may indeed have some minor errata, but Intel would microcode it away without flinching if they had to for market reasons. AMD simply has not given them a market reason to rush things and spend the extra money.

 

[ro3dog]

"3. Phenom is a bust at 65nm. So, I'll just upgrade my AM2 boards with 2.9 gigahertz 65watt Brisbanes first quarter of 2008 instead. I'll wait for fusion at 45nm instead of going Phenom this spring."From yipsl
The thing is most people who have the Phenom do like it ,with the black edition it will look even better .Intel has a problem with the cpu thats a fact which has been out for about a week now.I t'hink the X48 are being held for this reason.Also dont for get there's a war going on btwn Intel and Nvidia

 

[gpippas]

Why would Intel be having problems with there 45nm cpu's when they have already got there highest clocked version working fine. All they need to do is lower and lock the multiplier.

 

[cnumartyr]

Why would Intel be having problems with there 45nm cpu's when they have already got there highest clocked version working fine. All they need to do is lower and lock the multiplier.

Same thing could be said about AMD. AMD's bug isn't a huge deal to most desktop users is it? It hasn't really caused a ton of problems in the desktop realm.

Intel's problem is a rumor and no official statement has been released from them regarding it. Maybe that's how AMD should've played things.

 

[gpippas]

Same thing could be said about AMD. AMD's bug isn't a huge deal to most desktop users is it? It hasn't really caused a ton of problems in the desktop realm.

Intel's problem is a rumor and no official statement has been released from them regarding it. Maybe that's how AMD should've played things.

I agree the errata in Phenoms has been blown out of proportion. In the desktop it hasn't caused any problems. As for hiding it I don't they could have done. Not with Opterons being identical. I think in the server space where reliability is key it would have been found out or they would have had to disclose it. It wouldn't take long for people to start questioning as to whether the errata is in all K10 cpu's.

The clocks however is different. Although AMD is blaming the errata for the clock speeds I don't think this is entirely the case. Intel have shown how high they can clock there 45nm cpu's with room to spare for overclocking. AMD on the other hand has shown us how low they can clock Phenom.LOL

 

[someguy7]

All the extreme editions are are the highest binned of the penryn line. Intel didnt have a problem getting the qx9650 out i dont see how they would have problems with the lower binned cpus. AMD with is working the other way around. They release a lower clocked cpu and work there way up to higher clocks.

 

[homerdog]

Intel is correct in stating that they have no pressing need to release new processors at this point, but this still sucks for us. I hope (but doubt) that the B3 Phenom steppings will be all that AMD has claimed them to be, and then some. We need competition so bad, and all the while AMD is down to $7.68 a share.

As I side note, I guess that Intel would like to take this extra time that AMD has given them to further refine Penryn and squash every bug that they can. Remember, 45nm processors are cheaper to make than 65nm processors, so Intel has nothing to lose by switching over as soon as possible.

Side note #2: To MrsBytch, who stated that Penryn requires new chipsets, please take that BS somewhere else. Your FUD is not welcome here.

 

[metrazol]

It makes perfect sense for Intel to milk 65nm for everything it's worth as long as AMD can't keep up. Every month of 65nm production is another month of amortization on the 65nm equipment. The tax implications are no small factor in this decision.

 

[hcforde]

Basically there are three sides (actually 4)that have to be balanced. The marketing side, the technical side and the public perception side(the legal issues must be thought of "anti-trust"). Most of the post here have been leaning toward a single side. PR departments are caught in the middle and MOST times are NOT a part of the marketing side as a lot of people believe;especially in large organizations like INTEL. By backing down from their scheduled release the marketing side takes a back seat and the PR dept comes in for damage control while the legal side makes sure the company not only covers it rear, but looks good while doing so. (I was anticipating moving up to a Q9450 in Jan-but fortunately I bought a boxed Q6600 & MB on Black Friday from Frys for under $200 )

Intel needs AMD for ANTI-TRUST reasons. They can not blow them out of the water by overpowering their technology ability. It is a delicate balance they must maintain. I would be a suprise if they came to their rescue but not unheard of in this market. Rember a few years ago when Bil Gates loaned/gave APPLE a huge sum to remain competitive? That is the nature of competition/capitalism/avoidance of monopoly stature here in America with American based companies. Intel may or may not be having problems. The PR story is one on managed information. There are news releases already written to basically respond to any public reaction/senario that comes out of this situation. These guys think in terms of the end game not flying by the seat of their pants and giving knee-jerk responses. If I were in their good situation right now I would do the same thing while taking the time to perfect current and emerging technology. Their 45nm parts will probably have greater yields come release time(and maybe even a better stepping). Also do not forget that Intel is more than just a processor company; they are highly diversified, much more so than AMD. They can wait it out and still have abundant income from their other businesses.

The argument concerning their tier 1 MB partners is also valid a Good BIOS takes time. In the end if AMD goes down this way it will NOT be because INTEL pushed them out of business, but because AMD was not able to keep up technologically and this PR release is proff that Intel held back and did not pile on. They gave the competition time to breathe and prepare to be competitive. Smart move to keep them out of harms way legally.

Back to my Q6600. Glad I purchased it, now I have no qualms about putting it in my ASUS Maximum Formuls SE over the holidays it might be a long wait for the CPU that was intended for it.

 

[DaveElls]

I've been waiting to buy one of the new quads and this annoys me greatly as I would've just purchased a kentsfield a couple months ago if I knew that I'd have to wait even longer. I hope intel gives some sort of official statement on this soon so I know for sure if I should just get a Q6600 or not.

 

[OlSkoolChopper]

I'm exactly inthe same boat, DaveElls. I'm just figurng out a way to get mess around with my prehistoric PentD 915 to make it through to Nehilem!

 

[aevm]

Bill Gates made a fortune for Microsoft shareholders when he bought that piece of Apple.

He also saved Corel some years ago. You guys may not have heard much about that, but where I live Corel was the second largest IT employer at the time and it was front page news. Very weird too, considering that Corel's CEO at the time was declaring final war on both Word and Windows at the time.

I wonder if Intel could/should help AMD. Would that make any sense, or just raise even more anti-trust issues?

 

[jimmysmitty]

Why Intel is delaying the Yorkfeild we really don't know. Hell for all we know it could truly be because AMD has no competition and Intel wants to release Yorkfeild when they get the B3 Phenoms out to have a good set to compete against.

Or it could be Intel does want to work out the errata. We will probably never know. Either way Intel can afford the time right now but AMD can't afford anything they are doing as they don't have the superior product like the A64 was.

 

[MarkG]

The "bean counters" comment is spot on, but you missed the big one here: the more volume Intel gets out of 65nm, the more they get to run on equipment and tooling that's already bought and paid for, and half-depreciated. While 45nm saves money in the long run, ramping 45nm to high-volume manufacturing at multiple sites worldwide costs a substantial amount of startup cash.

Indeed. No company wants to waste a lot of money releasing a new product which obsoletes their old ones, if there's no reason to do so... it's not just bad for profits, but potentially harms relationships with OEMs if they're stuck with old chips which are no longer so valuable.

Since Intel's current chips can already compete with Phenom, they have precisely zero reason to rush new ones to the market. But you can be pretty sure they will be out to decisively beat AMD as soon as Phenom is properly released.

Which kind of sucks from my viewpoint, as I'm not intending to replace my P4 system until Nehalem.

Meanwhile, let the AMD fanboys bitch about maybe, possible, rumored Intel bugs so they can pretend AMD don't have very real Phenom bugs.

 

[LukeBird]

Meanwhile, let the AMD fanboys bitch about maybe, possible, rumored Intel bugs so they can pretend AMD don't have very real Phenom bugs.

I would point out there has been no "AMD Fanboy bitching" on this thread.
It was actually myself that made the point I thought Intel were selling bullshit...
There was some agreement, but we're all welcome to our own opinion!
I have made my point clearly, with argument from what I think the reasoning is and i guess I could be regarded as a "fanboy" does that make my opinion less valuable?
Anyhoo back to the discussion...
I personally don't think the anti-competition laws are much thought about there... If Intel could release crushing CPUs (and lets face it, they are high-end and unlikely to be bought by people who buy uni-boxs, where AMD still has a very strong showing with X2's) then I can see no reason why they wouldn't...
I can't imagine them being released at a price point that would drag people en masse from buying the (and I'll even admit this! ) very competitively priced Q6600.
It's economic madness to let a product that is ready for release, sit there gathering dust as it becomes obsolete!

 

[MarkG]

I would point out there has been no "AMD Fanboy bitching" on this thread.

I guess you missed all the 'LOL! Intel are just trying to hide their 45nm bugs!' posts.

It's economic madness to let a product that is ready for release, sit there gathering dust as it becomes obsolete!

There's a reason why you're not running Intel. Why exactly are chips which will easily beat the best of AMD's _unreleased_ chips 'becoming obsolete'?

The only thing that will make them obsolete is Nehalem, unless AMD produce something particularly impressive in the meantime. But if AMD do, then Intel will just release Nehalem and it's back to square one.

As I said, they have precisely zero reason to release now, when the only competition is their own chips. Only lunatics obsolete their own products for no reason, when they're currently the best in the market.

 

[wolverinero79]

This makes sense and doesn't make sense. Intel has a dedicated, fully ramped 45nm factory (or should at this point - F32). I don't know how many 65nm fabs they have, but it's almost definitely more than 1. If they convert all product SKUs over to 45nm, they don't have enough factories up to meet the current demand. So they're slowly converting/building their fleet and the one current fab is either just trying to release ultra extreme edition CPUs, or is stock piling the normal CPUs so that Intel has enough chips on hand not to see shortages when they release all SKUs. This is finite though as holding extra inventory has issues (and maybe older 130/90/65nm parts still have some inventory that Intel would like to blow out).

This announcement, I guess, doesn't mean Intel will be halting any plans of moving forward (since having their whole line doing 45nm brings Intel cost wins as they're cheaper to make). It just means that they'll wait a little bit before releasing the chips they're currently producing. If AMD is far enough behind, Intel may be able to just end of life tons of SKUs at one particular time and start selling all the 45nm parts. *shrug*

The other issue is that this isn't official. Intel would certainly never say we're delaying our chips because AMD sucks - even if it's the truth (which it is - AMD's way behind). It would be bad PR to say "hey, our competitor is so far behind, even if customers want new and improved functionality/power/performance, we'll just sit and wait".

 

[DaveElls]

http://www.xbitlabs.com/news/cpu/display/20071219103937_Intel_Set_to_Postpone_Introduction_of_New_Quad_Core_Microprocessors.html

xbitlabs is reporting that it's due to a defect now, as well as lack of competition. I'd really like to see an official statement by Intel so I can figure out if I should buy now or not.

 

[gpippas]

I guess you missed all the 'LOL! Intel are just trying to hide their 45nm bugs!' posts.


I agree I don't think there has been any fanboy bitching in this thread. Its one of the most civilised threads I seen on Tom's in a while. I would like you to quote where you have seen these 'LOL! Intel are just trying to hide their 45nm bugs!' posts in this thread.

 

[OlSkoolChopper]

Yah. Your right. This has been one of the most civilized and bitch-less threads around. So...

Hector Ruiz and thereby AMD is single handedly responsble for global warming, bird flu, mad cow, tainted cat food, crime in the streets, the price of gasoline, pimples, chicks who call you impotent in front of your friends, and guys who sell you car stereos on Ebay and ship you a box of rocks. Everyone ever asociated with AMD must be tortured to death on prime time televsion.

I hope I have restoerd the balance.

 

[gpippas]

Just read the xbitslabs article. They must have been reading my posts.LOL. I said the exact same thing near the beginning of this thread.

Even though they say there is a defect I can't see that it would be a major issue otherwise the QX9650 wouldn't have been released.