The AMD Execution Thread [2007 - 2017]
- Thread starter overclocked_enthusiasm
- Start date
-
- Tags
- amd
- Status
Exactly what AMD needed, yet another obscenely-high paid executive putting more drain on their finances.
http://www.xbitlabs.com/news/cpu/di...roprocessor_Still_Does_Not_Exist_Company.html
Of course if AMD keeps its usual practice of not using new tech for high-clocked CPUs then Shahghai won't be a threat anyway. I sure hope this won't drive prices sky-high.
I was expecting them to have a few Shanghai samples done some time ago. Assuming it takes a year from samples to launch it seems as in 2008 we'll see as many 45nm CPUs from AMD as we saw 65nm in 2006. Also Intel demonstrated working Nehalem in September or a whole quarter before first Shanghai samples so it should launch first making it even tougher for Shanghai.
Of course if AMD keeps its usual practice of not using new tech for high-clocked CPUs then Shahghai won't be a threat anyway. I sure hope this won't drive prices sky-high.
The Barcelona-Shanghai transition is less than that of Penryn-Nehalem, though a time of initial sample to release of less than a year would probably be a record for AMD.
Neither IBM nor AMD have released much in the way of detail for the 45nm high-performance process, and that's the one AMD would be working from.
AMD in particular likes to gradually (glacially) shift from the start of a process to the final optimized version. The information is so sparse, there is no real data to go on to expect a significant improvement in circuit performance, and 65nm for AMD was already a step backwards in some respects.
AMD's lack of information is usually indicative of a lack of something good to report.
45nm may not be anything to write home about for high-performance x86, and the early AMD process will not implement a lot of performance features.
The real danger here, if Shanghai regresses with respect to clocks and is only slightly more performant per clock than Barcelona, is that Nehalem will basically lap the field.
If the rumored specs for Beckton are correct, Nehalem's upper tier will exist in an entirely different league than AMD's best efforts.
Neither IBM nor AMD have released much in the way of detail for the 45nm high-performance process, and that's the one AMD would be working from.
AMD in particular likes to gradually (glacially) shift from the start of a process to the final optimized version. The information is so sparse, there is no real data to go on to expect a significant improvement in circuit performance, and 65nm for AMD was already a step backwards in some respects.
AMD's lack of information is usually indicative of a lack of something good to report.
45nm may not be anything to write home about for high-performance x86, and the early AMD process will not implement a lot of performance features.
The real danger here, if Shanghai regresses with respect to clocks and is only slightly more performant per clock than Barcelona, is that Nehalem will basically lap the field.
If the rumored specs for Beckton are correct, Nehalem's upper tier will exist in an entirely different league than AMD's best efforts.
Well if they stick to their roadmap Shanghai can't ship at the end of the year, or Montreal will slip into '09 at which point they may as well just declare bankruptcy.
So if we here about 45nm in Jan or Feb we know they are ok and if we do not hear about it in that period we can assume it did not boot. Goo of himto put this marker down for us.
I think after all's said and done AMD will either be noted as the worst management failure in the history of the PC industry, or the best. Right now it's leaning towards the former.
My understanding of that is they're going to write-down the entire goodwill. The point in doing that, rather than just a part of it, is that afterwards they will no longer have ~$80M of amortization related to ATI every quarter. That means their apparent losses afterwards will seem smaller.
Now, let's look at this differently: why do this now specifically? If they were going to achieve break-even this quarter, like they said they would, they certainly wouldn't want to make things look worse now in favour of future financial results. To me, this implies AMD won't beat guidance or analyst estimates in Q4.
However, that doesn't mean they're going to be much lower than guidance/estimates either. Who knows at this point...
Now, let's look at this differently: why do this now specifically? If they were going to achieve break-even this quarter, like they said they would, they certainly wouldn't want to make things look worse now in favour of future financial results. To me, this implies AMD won't beat guidance or analyst estimates in Q4.
However, that doesn't mean they're going to be much lower than guidance/estimates either. Who knows at this point...
There's one small thing not quoted above, but it's one that i believe will leave future consumers even more apprehensive and future competitive levels uncertain:
Could this mean they are giving up for good in the high-end GPU business due to AMD's inability to spend enough for R&D in the field ?
Could this mean they are giving up for good in the high-end GPU business due to AMD's inability to spend enough for R&D in the field ?
Well, I would support that, actually. Dead fish doesn't smell better by aging. Whatever it is, then get rid of it now.
I could turn it around tho too. It might be that by doing that they are hoping it puts them in position to be profitable in 1Q results.
If that was truly the case, they would write-off *part* of the goodwill. But it looks like they're getting rid of all of it.
But yeah, this *might* be good news for 1Q results too; hard to say for now, we'll see in due time, I guess.
Its odd why they would do it at this point though, because the ATi division is doing better from AMD's past quarter stand point, the write off will have less impact.
What constitutes the reasoning behind the goodwill that is present?
AMD paid more than the asset value of ATI, so the reasons for the extra payment would be things AMD hoped to realize from the ATI purchase.
Wouldn't some of those reasons be, among other things: fab ties, OEM relationships, bundling, future discrete products, Fusion, and those fuzzy "synergies"?
What among those has evaporated?
AMD paid more than the asset value of ATI, so the reasons for the extra payment would be things AMD hoped to realize from the ATI purchase.
Wouldn't some of those reasons be, among other things: fab ties, OEM relationships, bundling, future discrete products, Fusion, and those fuzzy "synergies"?
What among those has evaporated?
Yes, I agree. This was very common after the telecom bust. Acquire a company, write it off completely a year or two later. Have one quarter with an insanely large loss (think Nortel, JDSU, ...) to avoid dragging along depressing numbers as you go forward.
This is also better for the employees (especially the ones on top): large drop in stock price, write out new options or reprice existing ones, slow recovery after that is much more profitable than a stock that languishes for ages around the same depressed price.
- Status
