What will happen in 2007?
What will happen in 2007? Good question. Very good question. So here are my predictions for 2007:
- We will see the starting downfall of general purpose CPU architectures
When you look into the development of the architectures, it´s obvious that the one-size-fits all CPU will come to and end. Even for home user. Look at gaming. Graphic Processing Units with the compute power are quite normal. Hardware accelerated game physics starts to get ground. And within a few years we will see hardware accelerated game AI. Then you need a main processor only for loading the game and managing the memory. But for this tasks, webbrowsing and managing your bank account you don´t need a 4 GHz Quad Core. The same for server architectures. We will see the final breakthrough of throughput optimized architectures within this year. Many people suggest, that HPC will be owned by Cell in the coming years. My opinion: Don´t underestimate the compute power of an Niagara II with 8 FPUs. And don´t overestimate the usability of Cell outside of visualisation.
- 2007 - the year Linux got a real open source competition
I know, there were the BSDs, often technical more sophisticated, but with a chronic undervalutation and a missing large scale infrastructure for buying support when needed. So Linux had never a real competition. This changed already last year. And this will change even more in the next year. Although Linux will be the dominant variant for the years coming. But Opensolaris will be a strong competitor. Linux kept us innovating the last few years. Now we keep them innovating for the years to come. Futhermore i predict more problems like the issues and troubles around the Microsoft/Novell deal. Linux is at the brink to full commercialization. The effects of this to the developer community of linux will be interesting.
- Apple vs. Microsoft
Only one short comment: The world of the offices will be a Windows stronghold. No doubts. But it will be XP and not Vista. The market share for Apple and Mac OS in the SOHO segment will surge in the next year, with an vast acceleration of the trend when 10.5 is available.
From all i know i´m sure that we will reach our targets. We will at least reach the 4%. I think, we have only one danger in front of us: We must be very careful, that all the reorganisations we´ve done in the name of sustained profitability, doesn´t bite us in our back at the end. We are different, we think different. We need the capability to be different. We are not like Dell. Dell will be better at being Dell anytime soon (And we are not like HP. HP is better at office supplies). We have diffent strengths. But at CEC2006 i´ve got the strong impression, that the top management of SMI had the same thoughts. And to make a bold statement at the end of my predictions: I think we will be at 20$ end of 2007.