Intel Tick Tock slows down: Sandy Bridge slips into 2011?

[cbn]

http://www.brightsideofnews.com/new...-slows-down-sandy-bridge-slips-into-2011.aspx

"Lack of competition hurts the market - AMD's continuous delays of their next-gen CPU architecture combined with Intel's ill-fated effort on Larrabee resulted in a slowdown of the Tick-Tock cadence."

"At the top, the i7 980X will wipe the floor with anything AMD has right now, or any stepping they may come out with, in 2010. And at the low end, the dual core highly overclockable Westmeres can hold their own against entry level quad core AMDs."

EDIT: This is surprising to me. Wasn't AMD hoping to start production of Bulldozer in 1H 2010 according to one source?

 

[Voo]

EDIT: This is surprising to me. Wasn't AMD hoping to start production of Bulldozer in 1H 2010 according to one source?

Wasn't that just one site who couldn't even write off correctly? Should be 2011 not 2010

 

[cbn]

Wasn't that just one site who couldn't even write off correctly? Should be 2011 not 2010

I think you are right. As I recall there was some confusion on the production date.

If AMD starts production in 2011, I'm sure Intel would rather shift its resources/money into other niches like SSD and GPU. This slowdown of CPU development probably makes a lot of sense for them.

 

[lothar]

Sandy bridge has always been assumed to be 2011.
When websites reported in October that Intel delayed USB3 implementation until 2011, that means Sandy Bridge won't be out until then. Hint, hint.
Common sense FTW!!!

That's why everyone has been recommending buy your Quad core processors now!
I upgraded from my P4 system...I can't wait another 16 months.

 

[cbn]

Maybe if Intel delays Sandy Bridge enough we will end up seeing 32nm quads on LGA1156?

 

[cbn]

That's why everyone has been recommending buy your Quad core processors now!

I am not really sure "quad core" is a must have at the moment. Quite honestly the speed of my hard drive is what I notice most. Big difference loading up CS:S (small 2004 game occupying only 1GB total space) compared to newer games (which can occupy 10GB to 20GB).

It seems CPU implementation is progressing slower than other PC techs such as SSD and Video cards.

I upgraded from my P4 system...I can't wait another 16 months.

I hear you. I was using a Coppermine PIII before I built my Core 2 duo system (which was my very first computer build).

 

[Idontcare]

Node cadence isn't done for the sake of node cadence, it is done for the sake of gross margin enhancement...be it for cost reduction reasons, ASP enhancement reasons, or combination of both to mitigate the effects of ASP erosion over time.

If the info is true then the news here isn't that node cadence is slowing but rather that Intel's decision makers are doing exactly what we previously speculated (but had no proof) they would do at some point in time when the node timeline disparity between them and AMD grew to be about 1 year (give or take) and that was that they would intentionally slow down the release timeline to better maximize their margins.

We just didn't know whether the magic number for the timeline disparity was going to be 1 yr or 2 yrs or 3 yrs. It is now looking to be closer to 1yr than 2yrs. I can't begin to convey the magnitude of the relevance of this information when it comes to stock price valuation estimates for a forward looking R&D expenditure perspective. (again assuming the info is true)

The data speaks to the manner of decision making that is going on behind closed doors which we'd otherwise never gain access too. It is one thing to generalize this and say "yeah we knew a slowdown was coming, this only makes common sense" versus being able to leverage the quantification that this info contains into investment and business decision directions.

From a "pace of technology innovation" POV for the consumer it would absolutely be fair and accurate to generalize this as "Intel is resting on their laurels"...as resting on one's laurels is one way to enhance gross margins.

It is also a way to inadvertently allow your competition to regroup and possibly upstage you at a later date...but that is Intel's next CEO's problem (and institutional shareholders, and career employees), not a concern for the current CEO (nor day-trader shareholders or golden-parachute execs).

 

[aigomorla]

intel just doesnt need to be fast anymore.

They aren't losing, they arent even close to losing, nor are they close to someone catching up to them.

I told IDC this earily this year, that 32nm outlooks are grim because of lack of competition.

And lastly, i heard from my sponsor, the FTC investigation has halted shipment of sample processors to qualifying vendors even.

So intel is being a total ASS in this situation.
But i guess thats what we get for letting them get so big.

 

[Nemesis 1]

Man I don't get this , This is old news , Over a year ago we had same thread were intel had announced a cadence slowdown for 32nm. and SandyB.

At that time I assumed intel was having problems with 32nm process. But knowing thats not the case and the fact that 32nm 2 core with IGP is ass kicking . I see maybe Intel just blew it . These 2 core 32nm. are going to be the cpu to have all sectors. Sandy released as just in time for AMDs BD would be a sound stratigic move by intel , refining there steppings in the mean time. Intel may have goofed with the 2 core model time will tell. But man you guys are going to love these things, If They work with hydra chip and ATI discret graphics it will be a totally ass kicking gamer.

So backing sandy off makes good sense here as AMDs next big chip is do around same time. Hell Intel probabably has 2 differant designs ready to market.

 

[ilkhan]

If the info is true then the news here isn't that node cadence is slowing but rather that Intel's decision makers are doing exactly what we previously speculated (but had no proof) they would do at some point in time when the node timeline disparity between them and AMD grew to be about 1 year (give or take) and that was that they would intentionally slow down the release timeline to better maximize their margins.

Umm... Sandy Bridge isn't a shrink, its a new micro-arch. Unless you're referring specifically to the consumer quads, in which case it is a shrink.

 

[Idontcare]

Umm... Sandy Bridge isn't a shrink, its a new micro-arch. Unless you're referring specifically to the consumer quads, in which case it is a shrink.

I'm also referring to the present lack of westmere (32nm) in 2009...

from the OP's link:

Intel's Tick-Tock cadence lasted from Conroe to Nehalem... then it fell apart. Westmere is a year late, Sandy Bridge isn't on the product roadmaps for 2010. Don't ask about the Haswell, CPU+LRB part.

 

[aigomorla]

hey idc i'll be extremely lucky if i see 22nm next year.

Let me put it to you that way. I think you and a few others know what i mean by that comment also.

 

[ilkhan]

I'm also referring to the present lack of westmere (32nm) in 2009...

from the OP's link:
[MG]http://www.brightsideofnews.com/Data/2009_12_21/Intel-Sandy-Bridge-slips-to-2011-Tick-tock-with-a-slower-clock/INTC_TickTockMiss_675.jpg[/IMG]
Intel's Tick-Tock cadence lasted from Conroe to Nehalem... then it fell apart. Westmere is a year late, Sandy Bridge isn't on the product roadmaps for 2010. Don't ask about the Haswell, CPU+LRB part.

True enough.

 

[cbn]

From a "pace of technology innovation" POV for the consumer it would absolutely be fair and accurate to generalize this as "Intel is resting on their laurels"...as resting on one's laurels is one way to enhance gross margins.

We (as consumers) need AMD to catch up with their process technology.

But how will this happen if AMD is only competitive in the low profit margin value sector?

I can buy a Athon II 240/MSI mobo combo for $50 when it is on sale at Fry's. However, the MS Operating system required to run it is $110 (just for a OEM license). This is why I am hoping something good happens with Google OS. If that takes off it could decrease the total cost of PC ownership significantly.

Low cost OS plus low cost CPU/mainboards just makes so much sense. In fact, I am really hoping these become affordable enough to penetrate even impoverished markets (out selling even gaming consoles).

 

[cbn]

At that time I assumed intel was having problems with 32nm process. But knowing thats not the case and the fact that 32nm 2 core with IGP is ass kicking . I see maybe Intel just blew it . These 2 core 32nm. are going to be the cpu to have all sectors. Sandy released as just in time for AMDs BD would be a sound stratigic move by intel , refining there steppings in the mean time. Intel may have goofed with the 2 core model time will tell. But man you guys are going to love these things, If They work with hydra chip and ATI discret graphics it will be a totally ass kicking gamer.

I think these 32nm dual cores with HT could be surprisingly good (particularly the higher binned ones like Core i5 661).

P.S. I didn't even think about combining IGP with Lucid hydra. Thanks for the good idea.

 

[TuxDave]

I'm also referring to the present lack of westmere (32nm) in 2009...

from the OP's link:

Intel's Tick-Tock cadence lasted from Conroe to Nehalem... then it fell apart. Westmere is a year late, Sandy Bridge isn't on the product roadmaps for 2010. Don't ask about the Haswell, CPU+LRB part.

Is that how the public sees it? Westmere being a year late? Man...

 

[edplayer]

99.5% of the public doesn't know what a Westmere is


Only the nerds do

 

[aigomorla]

u guys honestly.. if u want faster gen tech, go on the xeon platform for your next cpu.

You will pay a small premium... but for 32nm Quad about 6months to 1 year early, i think that might be worth it.

 

[edplayer]

I'm not caring that much about new cpus these days

What I want is chipsets with integrated USB3 and SATA 3 (integrated by AMD or Intel) and next generation SSD drives.

 

[A554SS1N]

I think I *might* go for the Xeon route if it has much benefit other than being simply 32nm. No point getting it if it has no clock-speed/performance advantage.

 

[IntelUser2000]

Correction: The ORIGINAL Westmere schedule was high-end part in Q1(probably same as now, March) and dual core mainstream versions in 1-2 months later. Auburndale and Havendale parts were supposed to hold out for dual cores until then.

Well, Auburndale/Havendale got delayed and Intel decided to pull its 32nm dual core parts forward by a few months.

Sandy was always late-late 2010 to 2011. The point of Sandy Bridge I think was to close the differences in release dates for desktop/server/laptop. I think we'll see quad core Sandy Bridge first, then high-end+Xeon parts, similar to Westmere now.

 

[Idontcare]

Correction: The ORIGINAL Westmere schedule was high-end part in Q1(probably same as now, March) and dual core mainstream versions in 1-2 months later. Auburndale and Havendale parts were supposed to hold out for dual cores until then.

Well, Auburndale/Havendale got delayed and Intel decided to pull its 32nm dual core parts forward by a few months.

Sandy was always late-late 2010 to 2011. The point of Sandy Bridge I think was to close the differences in release dates for desktop/server/laptop. I think we'll see quad core Sandy Bridge first, then high-end+Xeon parts, similar to Westmere now.

Not sure what is original, what is official, what is delayed, and what is unofficial anymore...but I am just going by the rather splashy "32nm is coming bitches" media push Intel made back in Feb and am letting Intel set my expectations for what I was to see in Q4'09:

Tick-Tock: U R Doin it Right

Let’s check the stats; Conroe in July 2006, Penryn in October 2007, Nehalem in November 2008. That’s a tock, tick, and another tock, each about a year apart. Note that the cadence does appear to be slipping a bit, but we’ll see exactly when in 2009 we get Westmere before making any accusations.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3513&p=4http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3513&p=2

Enter the 32nm Lineup

Instead of Havendale in Q4, we’ll get Clarkdale and Arrandale. These are both dual-core, quad-thread processors, and both have on-package graphics. The CPU cores will be built on Intel’s 32nm process and in fact, they will be the first Westmere CPUs shipping into the market.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3513&p=5

We owe Intel a huge thanks for being so forthcoming with its roadmaps. It’s going to be a good couple of years for performance.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3513&p=7

 

[JFAMD]

Maybe if Intel delays Sandy Bridge enough we will end up seeing 32nm quads on LGA1156?

This type of statement comes up all the time, both on our end and our competitors' ends.

Just throwing something into the roadmap to fill a gap is a lot more complicated than you would think. The only time this really works is if a.) it is a derivative design (i.e. making a quad out out a six-core) and b.) there is enough slack in resources and scheduling to allow this to go in without disrupting the flow.

Most likely, in today's business environment, resources are scheduled tight based on the recent downturn, so the ability to just add something might not be realistic. Plus the test/validation resources are the typical bottlenecks once you get the design out. Reactive designs take a lot of time. Nothing is easy in this business. If you do see quads, then those products were probably already on the roadmap but not necessarily public (i.e. they were contingent products). In my mind that is probably a long shot.

 

[IntelUser2000]

Not sure what is original, what is official, what is delayed, and what is unofficial anymore...but I am just going by the rather splashy "32nm is coming bitches" media push Intel made back in Feb and am letting Intel set my expectations for what I was to see in Q4'09:

http://www.dvhardware.net/article33060.html

This 32nm die-shrink of Auburndale/Havendale was originally slated for the back-to-school season in Q3 2010, but it has been brought forward to Q1 2010.

I think if Auburndale/Havendale didn't get canned we'd have seen the dual core parts along with Lynnfield. But as always, they struggle with the graphics part, which is probably why it was delayed. On the plus side, really good things are showing for the 32nm parts.

 

[magreen]

not sure what is original, what is official, what is delayed, and what is unofficial anymore...but i am just going by the rather splashy "32nm is coming bitches" media push intel made back in feb and am letting intel set my expectations for what i was to see in q4'09:

rofl !