Conroe does 4.6Ghz already!!

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Viditor

Diamond Member
Oct 25, 1999
3,290
0
0
Originally posted by: zephyrprime
Originally posted by: pctwo
maybe it's time to buy INTC.

how long do you think til we see this reflected in stock prices?
I've noticed that wallstreet is very behind the curve for these sorts oof things. It probably won't be realized that Intc is on the ascent until ~Q406/Q107 in my opinion. For the next quarter or two, intel will have to get rid of smithfield inventory at low prices and they won't have enough conroes to fill demand. It's estimated that 20% of their shipments by end of year will be conroe which means that msot will still be P-Ds. And they're not selling conroe for outrageous prices either.

Stock may get a bump when Conroe is released bue to good publicity.

Technology isn't why people buy stocks, profits are...
You know all of those real nice low prices that Intel has announced? Guess what happens to profits when you do that... ;)
 

Lonyo

Lifer
Aug 10, 2002
21,938
6
81
Originally posted by: lopri
Originally posted by: welco
I just hope that this performance applies to everyday applications too. I mean, sure it's pretty cool to be able to a 10 second 1M SuperPi run, but I need to see performance this significant of an increase with games, video editing, and encoding to push me to buy one.

He also ran 32M and 3DMarks, too. I just skimped through the thread and 32M was around 12mins and 3DMark01 score was around 65K with a single GTX. Indeed I was almost in shock for a few secs and couldn't find a word to express myself. Scary stuff indeed. This is a gigantic leap in the ASIC manufacturing and will probably the most significant event in CPU history since the birth of A64.

Like Anand implied quite a few times in his article, Intel will eventually have on-die memory controller and things will get even uglier. Imagine, even now, Intel cut the L2 cache of E6700 to half (still 2MB) and integrate the memory controller on die..

I'm guessing 2~3 secs 1M WR with SSE4 optimized SP and dual/quad-core Conroe. :D
According to Sciencemark the latency to the main memory is the same for Conroe as it is for the AM2/939 AMD64's, about 45ns, compared to well over 70 for the Prescott's.
They're already level in terms of mem latency, so with an on die controller god knows what would happen, the world might implode!

 

Henny

Senior member
Nov 22, 2001
674
0
0
Originally posted by: Viditor
Originally posted by: zephyrprime
Originally posted by: pctwo
maybe it's time to buy INTC.

how long do you think til we see this reflected in stock prices?
I've noticed that wallstreet is very behind the curve for these sorts oof things. It probably won't be realized that Intc is on the ascent until ~Q406/Q107 in my opinion. For the next quarter or two, intel will have to get rid of smithfield inventory at low prices and they won't have enough conroes to fill demand. It's estimated that 20% of their shipments by end of year will be conroe which means that msot will still be P-Ds. And they're not selling conroe for outrageous prices either.

Stock may get a bump when Conroe is released bue to good publicity.

Technology isn't why people buy stocks, profits are...
You know all of those real nice low prices that Intel has announced? Guess what happens to profits when you do that... ;)

They probably go up. Remember that their goal is to fill fabs and fabs are almost like a fixed cost business. Whether they're 50% loaded or 100% loaded costs are approximately the same since the bulk of expenses are people and depreciation.

You also need to remember that Conroe is already on 300mm/65nm technology.

 

Viditor

Diamond Member
Oct 25, 1999
3,290
0
0
Originally posted by: Henny
Originally posted by: Viditor
Originally posted by: zephyrprime
Originally posted by: pctwo
maybe it's time to buy INTC.

how long do you think til we see this reflected in stock prices?
I've noticed that wallstreet is very behind the curve for these sorts oof things. It probably won't be realized that Intc is on the ascent until ~Q406/Q107 in my opinion. For the next quarter or two, intel will have to get rid of smithfield inventory at low prices and they won't have enough conroes to fill demand. It's estimated that 20% of their shipments by end of year will be conroe which means that msot will still be P-Ds. And they're not selling conroe for outrageous prices either.

Stock may get a bump when Conroe is released bue to good publicity.

Technology isn't why people buy stocks, profits are...
You know all of those real nice low prices that Intel has announced? Guess what happens to profits when you do that... ;)

They probably go up. Remember that their goal is to fill fabs and fabs are almost like a fixed cost business. Whether they're 50% loaded or 100% loaded costs are approximately the same since the bulk of expenses are people and depreciation.

You also need to remember that Conroe is already on 300mm/65nm technology.

In this case they don't...Intel has already forecast that profit margins will be dropping.
Their goal is to try and get back (or at least stop) AMD's big markestshare takeup as well as avoid having to write off the Netburst CPUs...
The problem that analysts and investors have is that once you lower your profit margin, it's VERY difficult to get it back.

As to costs for Fabs, you are partially correct that the majority of the cost is a fixed cost. However, the non-fixed costs aren't inconsequential...things like wafers are quite expensive (the silicon wafer cost is about $30-$40 of every Conroe made).
And while Conroe is probably going to be a very nice chip, it's also going to be the lowest volume part Intel has this year.

Edit: BTW, Intel's pricing is exactly the "Price War" that some analysts predicted and is the reason both AMD and Intel were downgraded. I don't expect to see Intel's share price move up until next year, and then only if they can bring their profit margin up significantly. Even then, I don't think we will ever see Intel at the lofty share prices they once held...now that AMD has established itself into the corporate world, it will be impossible to "stuff the toothpaste back into the tube".

The other thing you have to look at is revenue marketshare (not volume marketshare). AMD has increased their revenue marketshare for 11 of the last 13 quarters, though their volume marketshare hasn't increased nearly as much. Remember that you have to sell 100 Celerons to equal the profit of 1 Opteron 8xx...
 

Viditor

Diamond Member
Oct 25, 1999
3,290
0
0
Originally posted by: Questar
Technology isn't why people buy stocks, profits are...

No, expectations of future price increases are why people buy stock.

Huh?
You need a better understanding of how the semiconductor business works...
Hoping to raise prices again after cutting them so drastically is managerial suicide.

1. If Intel cuts prices with the intention of raising them again anytime soon, then there is absolutely no incentive for business to invest in the platform (nobody buys all of their chips/systems right away, they need both a technological as well as cost stability in order to plan their own business costs going forward).

2. The fastest ASP (Average Sale Price) growth ever was from AMD...it took them 3 years to bring up the price of their chips to present levels (which are still lower than current Intel offerings on a performance basis), and many are STILL bitching about it! The only reason they could even do that was that Intel stumbled badly and there was basically very little competition. While Conroe appears to be a very nice move for Intel, even the most rabid Intel fan wouldn't be silly enough to think that AMD won't be bringing major competition within the next year (let alone 3 years)...
 

pcoffman

Member
Jan 15, 2006
117
0
0
Originally posted by: pctwo
maybe it's time to buy INTC
:thumbsup:

Price per earnings is very low now, which means it's cheap. It will go lower. Problem is, it's practically impossible to buy a stock at its very bottom.
how long do you think til we see this reflected in stock prices?
Intel's going to have at least another rough quarter. I'm talking the current calendar quarter (Q2). So I don't think the stock has hit bottom yet.

Core 2 Duo doesn't come out until next quarter. Intel should have healthy sales of its new desktop chip then, at healthy margins. I would think the stock would start to bounce back early next quarter. But these things are notoriously difficult to predict.

Where Intel is really hurting is in the x86 server space. The desktop is important too, but its a shrinking market. Hopefully Woodcrest, or Xeon 5100 series I think it's called, will help restore a competitive balance to this space, but it's going to take time for Intel to regain market share there.

Long term, I think the stock will do great. Intel has sure had a hard time entering the multicore era, though.
 

hardwareking

Senior member
May 19, 2006
618
0
0
where do they get the engineering samples of the core 2 duo(conroe and merom) chips anyway?
Do intel just give it to them?And what about the compatable mobos?The revisions which support conroe are rare in the retail channel!
 

TekDemon

Platinum Member
Mar 12, 2001
2,296
1
81
Originally posted by: pctwo
maybe it's time to buy INTC.

how long do you think til we see this reflected in stock prices?

It'll be a while because Intel is still stuck with tons of old inventory that they have to get rid of, and there's no place to get rid of it to. Especially once Conroe actually drops if Intel hasn't dumped it's inventories by then they'll basically have to write off anything else leftover.

So...their balance sheets might still be pretty ugly for a while. Plus there's always the chance that Conroe is not going to have enough volume, or there's going to be some kind of problem with the chips, or AMD actually comes up with a competitive answer to Conroe, etc.

And since people price stocks based on risk, there's still plenty of risk for quite a while.

But I think INTC stock should be doing better sometime in 2007, assuming nothing goes wrong that is. Of course, even while Intel is reclaiming the desktop their server space is pretty much being eaten up by AMD so while they're winning back lost territory they're still losing old territory all the while.

Still, while profits probably won't get back to where they used to be, the stock is probably going to do a lot better than it is now if Intel manages to reclaim a significant amount of it's old territory-because even if profits are being hit at least they'll be set for growth in the future.

I'm going to track INTC though...thinking about buying about $2000 worth of it in the coming months. Unless of course AMD announces something truly fantastic.

Actually I'm almost hoping AMD announces a super clocked FX chip or some similar stunt. It'll probably ding Intel's stock prices, while not actually posing much of a threat if AMD is just cherry picking since there just won't be the volume they need.

GM and Intel are on their way back baby! (yes I bought GM stock, but unfortunately I bought it right after it first crashed-so it kept going lower and lower while I grimaced. Finally in the black after this last week though)

Heh I like this quote from one of the XS threads from a guy with an A64 clocked at 2.6Ghz:
"It makes me want to throw my pc through the window :suicides:"
lol...maybe it is time to buy Intel stock ;)
 

TekDemon

Platinum Member
Mar 12, 2001
2,296
1
81
On a side note, this is getting to be like drag racing...
DUDE YOU OWE ME A 10 SECOND CPU!
 

imported_Questar

Senior member
Aug 12, 2004
235
0
0
Originally posted by: Viditor
Originally posted by: Questar
Technology isn't why people buy stocks, profits are...

No, expectations of future price increases are why people buy stock.

Huh?
You need a better understanding of how the semiconductor business works...
Hoping to raise prices again after cutting them so drastically is managerial suicide.

1. If Intel cuts prices with the intention of raising them again anytime soon, then there is absolutely no incentive for business to invest in the platform (nobody buys all of their chips/systems right away, they need both a technological as well as cost stability in order to plan their own business costs going forward).

2. The fastest ASP (Average Sale Price) growth ever was from AMD...it took them 3 years to bring up the price of their chips to present levels (which are still lower than current Intel offerings on a performance basis), and many are STILL bitching about it! The only reason they could even do that was that Intel stumbled badly and there was basically very little competition. While Conroe appears to be a very nice move for Intel, even the most rabid Intel fan wouldn't be silly enough to think that AMD won't be bringing major competition within the next year (let alone 3 years)...


Dude, I was talking about stock price increases.
 

skooma

Senior member
Apr 13, 2006
635
28
91
Originally posted by: dexvx
Originally posted by: thestain
Also.. what is up with the front side bus.. how did these overclockers get it up so fast?

And.. is there any stability at any of these levels.. ie.. will it really work?

1) They hard mod the FSB. This is well beyond the level of enthusiasm for most AT'ers. XS is called xtreme for a reason, and they go to extreme levels to achieve high overclocks and pushing hardware to its limits.

2) No, its not for long term stability. But relatively speaking, it also gives you a good idea of ceilings for said CPU's. The highest Athlon FX-60's they got was also around the low 4Ghz's.
Exactly.

ell, I'll be happy with a conroe @ 3.2 :) That thing will scream compared to my x2 @ 2700.
 

Looney

Lifer
Jun 13, 2000
21,938
5
0
What's taking them so long to release it anyways? Fabrication success rate? I just gave my Athlon 2500XP to my brother and am running my server on an Athlon 1600 now. I so want to upgrade to Conroe!
 

SexyK

Golden Member
Jul 30, 2001
1,343
4
76
Originally posted by: Looney
What's taking them so long to release it anyways? Fabrication success rate? I just gave my Athlon 2500XP to my brother and am running my server on an Athlon 1600 now. I so want to upgrade to Conroe!

They are probably waiting to ensure adequate supply to satisfy what is looking like a huge demand at launch time. Shouldn't be having any problems with yields at this point since they've been fabbing at 65nm for quite some time now.