You'd be surprised how many people read these forums and invest partially on information they glean from here posted by posters like you, especially since this is a thread about AMD's financials.
The WSA is a disaster for AMD
WSA financial obligation is known from day one, nothing unknown here.- massive unknown financial obligation
- cause of excessive inventory levels and write offs
- substitute products and architectures that aren't selling for those that are (Brazos)
- product cancellations (Wichita)
- designing large die size unnecessarily. (bulldozer)
- excessive R&D on big core playing catch up instead of investing on small core
- paying to cancel an exclusive with Gloflo even though Gloflo doesn't have the required process on time
The latest for 2014 I suspect will be failure to secure leading edge manufacturing capacity from TSMC well because a) they have to meet their WSA and thus cannot send business to TSMC and b) why should TSMC support AMD when as soon as Global Foundries is ready (probably 1-2 years later and about 3-4 years behind Intel), AMD will move the business over?
Last but not least, whenever I see a ad hominem attack, I know said attacker is basically admitting the other party is right as their only recourse is personal attacks.
Yeap, thats why they had the best financial year since 2011
.
Sooner or later it will breach $4 again, but possibly drop below $3.
That account for minimal amounts. Remember (GPU) mining is almost exclusively north america only.
WSA financial obligation is known from day one, nothing unknown here.
2013 was the best AMD financial year since 2011. People should start to acknowledge that and stop trolling with every chance they get. AMD is in the right direction, they have completed the first two faces of restructuring and have continue delivering their products on time. They are more than ever ready to operate with low margins and that will help them in the near future.
They have a long way ahead of them but they are doing the best they can under those difficult circumstances.
Eh, no. What is an North American phenomenom, however, is price gouging. There weren't any shortages due to mining in NA, just greedy retailers and etailers. You have plenty of miners in Europe, at least in Germany, Sweden, Netherlands etc.
Do you have any facts to back up your claims surrounding that statement or are you just pulling stuff out of thin air?
The fact an obligation exists is know.
The size of the obligation is unknown, because it depends on how much product AMD can actually make at GF and then sell to customers, that is what he is referring to.
Bulldozer continues to be an unbelievable albatross around AMD's neck, as all the subsequent Bulldozer variants have terrible single core performance.
And now within that small niche that Kaveri operates in(i.e. want more GPU oomph than Intel provides, but less than a discrete card), they have a bit more than 9 months to try and exploit this the best they can, before Broadwell arrives to spoil their party.
It also can't be a good thing that Intel is looking to bring Atom variants to the low end of the desktop market either, as this has been a relative stronghold for AMD.
And what record sales were these?That didnt stop it from having record sales of High-End A8-10 APUs in Q3 and Q4 2013.
Broadwell Desktop SKUs will only be High-End highly expensive K series CPUs at the $220+ segment. Broadwell Desktop will not affect Kaveri sales in 2014.
Those ATOMS were really crap and crap ATOMS has been what AMD has been facing in recent times.Intel already have ATOM based BGA only in the desktop from the first ATOM release, nothing new here. Whats new is the socket Kabini release in Q1 2014. That will have an impact in Entry and Low end Desktop segment.
It is good for AMD that they picked up the Console wins, but they are losing desktop marketshare at an alarming rate.
Bulldozer continues to be an unbelievable albatross around AMD's neck, as all the subsequent Bulldozer variants have terrible single core performance.
AMD just can't have their post Bulldozer CPU architecture arrive quickly enough.
Kaveri has clearly turned out to be much less than people were hoping for 24 months ago, 12 months ago, even 6 months ago.
And now within that small niche that Kaveri operates in(i.e. want more GPU oomph than Intel provides, but less than a discrete card), they have a bit more than 9 months to try and exploit this the best they can, before Broadwell arrives to spoil their party.
It also can't be a good thing that Intel is looking to bring Atom variants to the low end of the desktop market either, as this has been a relative stronghold for AMD.
All the single thread perf bashing is bs.
The single threadded cpu perf is an intel song used to protect their monopoly. The idea that performange should be improved here is technically and cost vise idiotic and harming consumers.
Unfortunately this song is repeated at AT constantly and keep us all from getting faster cpu perf because software then dont have to evolve.
Oxide have shown an game engine can be build that scales perfectly even on dx and mantle have shown this can be done easily and at low cost. The prior understanding we had of it have been proved wrong. People dont seem to get that. More cores is the future and so is hsa like architechtures.
And what record sales were these?
Those ATOMS were really crap and crap ATOMS has been what AMD has been facing in recent times.
The BayTrail ATOMS will be good enough to give AMD grief in the low end desktop market.
No it is absolutely on the money.
When affordable Dual and Quads, with and without HT are available, then single thread performance is KING.
Wow, who would have ever thought that Anandtech was holding back the entire industry like that.
No, single threaded performance matters most, because for so many applications and problems to be solved, there is a limited amount of multi-threading that is going to be happening for the forseeable future.
The future will over time come to utilise more cores, but even then, there will always be a high value placed on single core performance.
But right now and for the next few years on the desktop, it is not MOAR CORES we are lacking, but rather the performance of each core.
The market makes the final decision, and obviously has chosen efficiency and CPU performance.
Not to mention that the same argument applies to the gpu portion of amds apus. More performance than 90 % of the users need, but too bandwidth limited for gpu intensive tasks like gaming. The market makes the final decision, and obviously has chosen efficiency and CPU performance.
But if CPU perf was still an issue , why would intel promote
things like this :
You're right, a company should never promote any products other than their highest end.But if CPU perf was still an issue , why would intel promote
things like this :
But if CPU perf was still an issue , why would intel promote
things like this :
Seems that intel dont buy the ST argument they use
for marketing reasons but in practice they know that
most of their CPU offering is overkill in respect of the
average people daily usages.
So I read AMD's 2013 SEC annual report filing yesterday, here it is in .pdf format:
http://services.corporate-ir.net/SE...hbmNlZE1pY3JvRGV2aWNlc18xMEtfMjAxMzAyMjEucGRm
The 2014 filing will show up sometime in the next couple of months, I'm sure.
According to that filing, the WSA agreement expires in 2022. It has been amended several times, which is noted in the document.
The filing notes the obligations AMD had to Global Foundries (see page 60). The final payment due was $250 in Q1 2014, which AMD has already paid off this month.
It should also be noted that the PS4 and XBox One chips are being manufactured at TSMC.
I did note that GlobalFoundries' website mentions that they have over 150 customers currently, including most of the top 20 semiconductor companies. I also noted in another article that GlobalFoundries accounts for 3.3% of total 200mm equivalent wafer production capacity, behind Intel's 6.5%.
As of December 2013:
200mm-Equiv Wafers per month x1000
Samsung 12.6% (1,867)
TSMC 10.0% (1,475)
Micron 9.3% (1,380)
Toshiba/SanDisk 8% (1,177)
SK Hynix 7% (1,035)
Intel 6.5% (961)
ST 3.7% (551)
UMC 3.5% (520)
GloFo 3.3% (482)
TI 3% (441)
The top ten above account for 66.8% of total wafer capacity as of Dec 2013.
TSMC, UMC and GlobalFoundries, as pure-play foundries, have held about 80% of the pure-play market since 2010. Also, Intel's wafer capacity is only about 2/3rds of TSMC at this point.
http://www.icinsights.com/news/bull...ron-Top-List-Of-IC-Industry-Capacity-Leaders/
Global Foundries continues to grow, so IMHO from the Fab employee's perspective it looks like AMD made the right choice. As for whether it was a good move for AMD, well the more interesting question is, would AMD even be a company today if they HAD NOT done the deal?
The last question is easy. No.
But the agreement was from the start very unclear and not market and consumer oriented.
2022 is so far in the future its like forewer.
Btw: the game is at 300mm wafers today.
Show me some highend products from gf besides amd?
ULV mobile chips are actually good ST performers. They have low base clocks but turbo VERY aggressively in ST tasks.
Well, not the Celerons through i3s. Which is why the Kabini A6s can match those low voltage i3s in multi-threaded performance.