I trust him as much as anyone working at The Inquirer or Fudzilla. I wouldn't call them "industry insiders" either.
"Guessers", maybe.
The AMD Execution Thread [2007 - 2017]
- Thread starter overclocked_enthusiasm
- Start date
-
- Tags
- amd
- Status
I trust him as much as anyone working at The Inquirer or Fudzilla. I wouldn't call them "industry insiders" either.
"Guessers", maybe.
I trust him as much as anyone working at The Inquirer or Fudzilla. I wouldn't call them "industry insiders" either.
"Guessers", maybe.
They might have more reasons to have a "source" than some random basement dweller, even if those are just FUDs from industry.
I write more than 100 posts a day at work.
Perhaps the rumor mill can indicate how it is AMD's x86 license can continue if it is acquired.
Actual legal documentation indicates that at least since 2001 there's been a provision that the acquisition of either party will end AMD's rights to x86.
Unless that has been renegotiated, and I see no reason for Intel to budge on this issue, AMD is worth as much after being acquired as it would after bankruptcy (which also ends the license, by the way).
Actual legal documentation indicates that at least since 2001 there's been a provision that the acquisition of either party will end AMD's rights to x86.
Unless that has been renegotiated, and I see no reason for Intel to budge on this issue, AMD is worth as much after being acquired as it would after bankruptcy (which also ends the license, by the way).
That was a mistype.
The term was "other party", as in whichever company violates the clause will lose the rights.
However, the company that doesn't violate the agreement is unaffected.
If Intel were to be bought up (somehow) it would lose the right to use AMD patents and instructions.
However, AMD would still enjoy access to Intel's patents.
In the more likely case AMD is bought or is forced to restructure, it loses x86 rights and Intel can copy AMD tech at will.
The term was "other party", as in whichever company violates the clause will lose the rights.
However, the company that doesn't violate the agreement is unaffected.
If Intel were to be bought up (somehow) it would lose the right to use AMD patents and instructions.
However, AMD would still enjoy access to Intel's patents.
In the more likely case AMD is bought or is forced to restructure, it loses x86 rights and Intel can copy AMD tech at will.
Farewell to all the lost posts
In attempt to pick up the thread I will ask this:
In light of Intel's extremely good quarter, what impact are we to expect on AMD's stock price? Will it be another "rising tide lifts all boats" situation where the street will infer good fortunes for the entire sector from Intel's results or will it be seen more as a zero-sum situation, in which Intel is winning the tug of war?
In attempt to pick up the thread I will ask this:
In light of Intel's extremely good quarter, what impact are we to expect on AMD's stock price? Will it be another "rising tide lifts all boats" situation where the street will infer good fortunes for the entire sector from Intel's results or will it be seen more as a zero-sum situation, in which Intel is winning the tug of war?
AMD is up slightly in after hours, if that's what you want to know. Based on the Intel seekingalpha.com transcript though:
- Intel's ASPs are slightly up in desktop, slightly down in notebooks and probably servers. Flat overall.
- Much of the margin benefit was from higher revenue (better fab usage).
- " At the same time that happened, we walked from a lot of low end business on the desktop and the notebook at prices that we just didn't think made sense to us."
Flat ASPs vs AMD's is hard to judge, because every time Intel cuts prices, they have a new faster SKU but AMD doesn't. It's not a big effect, but I suspect it's there. However, overall, it's still good for AMD: Intel doesn't seem to want to do a full-blown price war anymore.
I would tend to agree with the average analyst estimate of a $350M net loss for Q3. However, I would forecast it on slightly higher revenue than $1.52B. If they only manage <=$1.52B, then I would expect a >$400M net loss, personally. So we'll see, but I'm rather bullish on revenue.
As for how all of this affects the stock; it probably won't. As long as the results are not catastrophic (i.e. weaker growth than seasonality), then the ONLY thing that probably matters is Q4 guidance, imo. And I'm NOT going to try predicting that one, heh! Analysts' estimates seem fairly aggressive there really, but not out-of-this-world, so we'll see.
- Intel's ASPs are slightly up in desktop, slightly down in notebooks and probably servers. Flat overall.
- Much of the margin benefit was from higher revenue (better fab usage).
- " At the same time that happened, we walked from a lot of low end business on the desktop and the notebook at prices that we just didn't think made sense to us."
Flat ASPs vs AMD's is hard to judge, because every time Intel cuts prices, they have a new faster SKU but AMD doesn't. It's not a big effect, but I suspect it's there. However, overall, it's still good for AMD: Intel doesn't seem to want to do a full-blown price war anymore.
I would tend to agree with the average analyst estimate of a $350M net loss for Q3. However, I would forecast it on slightly higher revenue than $1.52B. If they only manage <=$1.52B, then I would expect a >$400M net loss, personally. So we'll see, but I'm rather bullish on revenue.
As for how all of this affects the stock; it probably won't. As long as the results are not catastrophic (i.e. weaker growth than seasonality), then the ONLY thing that probably matters is Q4 guidance, imo. And I'm NOT going to try predicting that one, heh!
Analysts' estimates seem fairly aggressive there really, but not out-of-this-world, so we'll see.AMD's product introductions are in a sort of a holding pattern.
Any Barcelona not at the bottom clock rung is lost to the retail channel, and the higher clock SKUs are either being sucked up by a few large HPC installations, or being gobbled up by Tier 1 companies.
That being said, there is little in the way of availability for such systems even from AMD biggies like IBM and Sun.
The extended time for verification might play a role, but the dearth of information has me puzzled.
If most of the processors sold have gone to big HPC installations, the news is still bad for AMD.
The number of chips per customer is on the order of tens of thousands of chips.
For a chip that should be selling in the millions, that is a few zeroes short of volume production.
The yields and bins for these chips look from all directions to be poor.
AMD's released a new speed grade of A64 X2's, and these Brisbanes will still lag behind the 90nm chips.
The dual-core Kuma and its variants won't be seen until 2Q 2008, and the projected clock speeds are rather middling. The first Kumas six months from now still might not outclock the top 90nm Athlon 64 FXs we've had for almost as long.
AMD's priorities seem to be on the server market, enough so that it appears that extending Brisbane's lifetime way past its prime is worth whatever die area they can spare for the quad-cores.
Let's hope AMD has found a way to make Brisbane nearly for free or these last Brisbanes are all inventory. Can they be when they are a newer stepping?
Intel's pushing down dual-core Core2 chips to the 50-100 dollar range soon.
I think the magic number for A64 break even was ~50-60 dollars a chip.
Any Barcelona not at the bottom clock rung is lost to the retail channel, and the higher clock SKUs are either being sucked up by a few large HPC installations, or being gobbled up by Tier 1 companies.
That being said, there is little in the way of availability for such systems even from AMD biggies like IBM and Sun.
The extended time for verification might play a role, but the dearth of information has me puzzled.
If most of the processors sold have gone to big HPC installations, the news is still bad for AMD.
The number of chips per customer is on the order of tens of thousands of chips.
For a chip that should be selling in the millions, that is a few zeroes short of volume production.
The yields and bins for these chips look from all directions to be poor.
AMD's released a new speed grade of A64 X2's, and these Brisbanes will still lag behind the 90nm chips.
The dual-core Kuma and its variants won't be seen until 2Q 2008, and the projected clock speeds are rather middling. The first Kumas six months from now still might not outclock the top 90nm Athlon 64 FXs we've had for almost as long.
AMD's priorities seem to be on the server market, enough so that it appears that extending Brisbane's lifetime way past its prime is worth whatever die area they can spare for the quad-cores.
Let's hope AMD has found a way to make Brisbane nearly for free or these last Brisbanes are all inventory. Can they be when they are a newer stepping?
Intel's pushing down dual-core Core2 chips to the 50-100 dollar range soon.
I think the magic number for A64 break even was ~50-60 dollars a chip.
Maybe it's an indication of how pessimistic I am about their fundamentals after the total lack of bottom line improvement in 2Q. . . .but I'd be quite happy with a $350M bottom line loss, so long as it didn't include a lot of "one-time" revenue from sales of assets and such. Not that it would be fabulous on its own, but first you have to get the trend line going back in the right direction and at least it would do that (which 2Q did not).
I think any total loss exclusive of one-time revenues that started with a "2" would get a huge sigh of relief out of me.
I'm actually steeling myself for something starting with a "4" and hoping it won't be.
I think any total loss exclusive of one-time revenues that started with a "2" would get a huge sigh of relief out of me.
I'm actually steeling myself for something starting with a "4" and hoping it won't be.
Well, what do we have for one-time income?
There's the Spansion stake.
Prior to the ATI acquisition, someone pegged it as being roughly $650 million.
That was in 2006, and the price is about half now.
Since it is in competition with Intel's flash business (which by the way is doing well right now), that leaves AMD's options there at possibly ~$300 million dollars.
There's the sale of old 200mm equipment, which had some quoted prices of $80 million dollars.
Those two alone would make up for one 3Q.
Outside of that, what else is there?
There's real estate, most notably the New York fab property if things are dire.
There may be some plots in Texas, and possibly some stuff ATI had.
That value of that might be good for a quarter.
Those are all one-time revenue sources.
Too bad for AMD that its one-time expenses keep coming back.
I'm looking forward to seeing how much AMD charges to "one-time" costs from the acquisition.
There's the Spansion stake.
Prior to the ATI acquisition, someone pegged it as being roughly $650 million.
That was in 2006, and the price is about half now.
Since it is in competition with Intel's flash business (which by the way is doing well right now), that leaves AMD's options there at possibly ~$300 million dollars.
There's the sale of old 200mm equipment, which had some quoted prices of $80 million dollars.
Those two alone would make up for one 3Q.
Outside of that, what else is there?
There's real estate, most notably the New York fab property if things are dire.
There may be some plots in Texas, and possibly some stuff ATI had.
That value of that might be good for a quarter.
Those are all one-time revenue sources.
Too bad for AMD that its one-time expenses keep coming back.
I'm looking forward to seeing how much AMD charges to "one-time" costs from the acquisition.
Absolutely. Me too. I'm guessing they've got that on a trend line that they are running it out over. So I'm expecting something in the ~$45M range this quarter. We know they've been doing some layoffs and such as well, and those will probably show up as one time costs too.
I think you both might be right with the Ati costs still dribbling out, though this will be 9 months of reporting saying costs are still there if it is true.
I am going to go for a loss starting with a 3, probably the next number 5-9 range. I'm optimistic on their sales being slightly better than I previously thought.
I'm no financial whizz kid though so it is probably way off
I am going to go for a loss starting with a 3, probably the next number 5-9 range. I'm optimistic on their sales being slightly better than I previously thought.
I'm no financial whizz kid though so it is probably way off
How much in total does the ATI one off costs now add up to ?
Surely that FAB in New York is now down the pan.
They need to get K10 polished up pronto and sticking out good speeds with good yields if they are going to reverse the trend. Then they need to get Shanghai out the door with same.
Otherwise Intels tick tock will be counting down the seconds to midnight for them I fear.
Surely that FAB in New York is now down the pan.
They need to get K10 polished up pronto and sticking out good speeds with good yields if they are going to reverse the trend. Then they need to get Shanghai out the door with same.
Otherwise Intels tick tock will be counting down the seconds to midnight for them I fear.
No, the ATI related charges was $76M. The $120M included other stuff too.
However, worse than that --which is practically flat from last quarter-- is that they predicted the same for 4Q. So, without better visibility into what is really causing that charge, it is hard to say how much of a net drag/plus ex-ATI really is.
But anyway you look at it Graphics and CE made major improvements 3Q vs 2Q.
From my perspective these numbers are classic "Not as good as I hoped; not as bad as I feared".
However, worse than that --which is practically flat from last quarter-- is that they predicted the same for 4Q. So, without better visibility into what is really causing that charge, it is hard to say how much of a net drag/plus ex-ATI really is.
But anyway you look at it Graphics and CE made major improvements 3Q vs 2Q.
From my perspective these numbers are classic "Not as good as I hoped; not as bad as I feared".
- Status
