http://www.bloomberg.com/news/2014-10-16/amd-sales-may-miss-estimates-as-intel-gains-pc-orders.html
(...)
Illustrating what AMD is up against, Intel took in 94.7 percent of the revenue in the PC processor market in the second quarter, while AMD garnered 5.2 percent, according to market researcher IDC. Intel’s chips powered 84 percent of desktop units shipped and 88 percent of laptops. By all those measures, AMD lost ground to its larger rival.
My first reaction when I saw the miss and the dreadful 4th quarter earning was sucks for AMD. But the conference call to me slowly paints a rosier picture.
We've met our goal to double our commercial client design wins from last year, and are pleased with the initial progress we are making to build a richer mix in our PC business. New commercial client offerings from Dell, HP, Lenovo have started ramping, resulting in approximately a 50% increase in our commercial APU shipments from the second quarter.
We also improved our notebook APU mix in the quarter, as our Kaveri processors ramped in mobile design wins and our higher-end mobile processor unit shipments increased nearly 50% from the second quarter. Mobile discrete GPU unit shipments also increased from the second quarter, as new design wins entered production.
I'm starting to get a better picture of how AMD lead by Read and Lisa's approach has been. Priorities are Business market first, follow by mobile, then consumer APUs. Embedded will be hitting 20 nm in 2015 while consumer CPU will still be in 28 nm. Zen will give server market priority first so it sucks a bit if you've been an AMD diehard fan and been waiting for a FX refresh.
http://wccftech.com/amds-generation...-debut-2016-dense-server-market-top-priority/
So while I think overall they might have rapidly eroded their desktop shares in the consumer space, their long term strategy is to target the more stable business market and they're willing to reallocate resource, let the consumer desktop market bleed some more, and sacrifice 3rd and 4th quarter earning to make it happen.
I think the 3rd Quarter miss is for two reason.
1. The gradual irrelevance of their FX line which really makes them an almost none existent player in the enthusiast desktop market at this point. Yes I know Kaveri is awesome if you happen to want to integrated graphics or have something that requires HSA, but I don't think that small market is picking up rapidly enough to make up for the gigantic void left by the neglect in their FX line. Remember, once upon a time high performance desktop part was AMD's highest source of revenue.
2. Bitcoin used video card market. They pretty much admitted to this in the conference call without directly admitting to it. It's a double edge sword I suppose. The Bitcoin/Litecoin market lite up their 3rd and 4th quarter sale last year and give investors false hope that their turnaround was going better than expected.
Now as for the 4th quarter lousy forecast, here's what I suspect.
1. No answer to Maxwell. While Nvidia invested heavily to give their 28 nm product one last performance boost, AMD just kind let it go with minor improvements to appease the business market, aka, Apple with the m295X. Now this might be a good thing for AMD if it means they are in better position to release a 20nm gpu ahead of Nvidia in 2015.
2. Last quarter of no high performance CPU left by the FX void and one quarter before Carrizo debuts.
So if my assumption are correct on why Quarter 3 and 4 are bad, it's possible that 2015 is shaping up to be the first real quarter of success from their 3 years of turnaround plan. Here are the reasons.
1. Embedded market seems to be gaining traction for AMD. Nvidia and Intel has been in this market for awhile now and it's a very stable source of revenue. More importantly for AMD, it allows them to very cheaply reuse IP and that's probably the whole reason behind project Skybridge. Lisa mention her goal is 1-2 semi custom wins/year. If they can repeat what they just announced this year again in 2015, they should be in really good position business-wise. On top of the existing contract being 3 years 1 billion, Lisa did mention what she liked about them is that they are stickies. Once you're in it's long term. My interpretation to that is they would be in excellent position to renew that contract another 3 years once it's up.
2. Graphics should pick up too. The popular rumor is that Pirate Island will be 1st quarter of 2015 in 20nm and HBM. Nvidia even if they manage to release maxwell in 20 nm also, should have no response to HBM until 2016. I've even heard rumor that AMD plans on paper launching Pirate Island December of 2014.
3. Carrizo. It just needs to be good enough for them to gain some market share in the business sector. At 28 nm if they can offer the advertised performance numbers, they should be easily able to gain shares in the business market. Everyone seems to think lighter, more powerful, more energy efficient is key to the business market. That's not the case. Cost is a huge factor and not every business is going to opt for the premium of the 14nm Intel part. A 15 watt cpu laptop that can load jpeg like an aka47 should get them some attention.
So despite the bad earning call and the horrible 4th quarter forecast, I don't view it as a bad thing because it really does seem like AMD is setting themselves up for success down the road and we're really beginning to see tangible success at the end of their tunnel.
Now here is the BIG I don't like and this is huge. If AMD really was in a great position for 2015, why the big layoff and a very targeted one at that too at their consumer CPU and GPU division. This leads me to believe that either 1, AMD does not believe Carrizo will bring them any business market improvement over Kaveri or 2, AMD believes 20nm Pirate Island will slip or Nvidia will have an equally good or better answer to match or top or 3, the complete gutting of all resources none engineering related of their FX line since it's not due for a refresh for at least another year.
I hope for AMD's sake the reason is 3, the gutting of their FX line staff because if it's either Carrizo not being as good as advertise or Pirate Island not being able to give AMD an advantage, 2015 will shape up to be a pretty lousy year also.