Intel just bought Infineon's wireless unit: http://www.infineon.com/cms/en/corporate/press/news/releases/2010/INFXX201008-069.html (although the transaction is only expected to close in Q1 2011)
This is obviously huge news. Infineon has the industry's most successful 'slim baseband' and probably the best cellular RF engineering team in the world (followed by Icera's excellent ex-Sirific team and probably Broadcom's). They've been fairly aggressive on process with their 40nm XMM 626 baseband and 65nm UE2 RF chip. They're working on 32nm RF macros for single-chip integration. They've got industry leading 65nm single-chip solutions for the ultra-low-cost 2G segment.
The relevance of this to Atom is obvious, but if it was just for that I'd consider the acquisition a gigantic waste of money (slim basebands are just fine for several more years and integration will increase risk). What's more interesting is whether Intel will try to aggressively bundle Infineon's HSPA+ baseband with their WiFi chipsets and, more generally, their entire notebook processor portfolio. This could accelerate the growth of 3G/4G modules and hurt USB sticks.
What's interesting is that Qualcomm once tried to penetrate the module market directly with Gobi and learned a valuable lesson: OEMs and customers care about some of the features they added there, but they couldn't handle the design flow on their own and had to rely on partners such as Sierra Wireless and Option anyway. Cutting the middle man just didn't work and I doubt it would work for Intel either. Qualcomm is a very strong incumbent in all the markets Intel is aiming at and it's going to be an interesting fight to follow.
Another company that will be affected is Icera. They've had a lot of success in USB sticks lately thanks to higher performance and lower cost, but AFAIK haven't achieved as much traction in embedded modules. They could fail to penetrate that market if both Qualcomm and Intel fight hard for it with both proprietary software and bundling. On the other hand, they could really benefit from this acquisition in the smartphone space where customers may be wary of combining an Intel baseband with an application processor from Texas Instruments, NVIDIA, Samsung, or Marvell.
This is obviously huge news. Infineon has the industry's most successful 'slim baseband' and probably the best cellular RF engineering team in the world (followed by Icera's excellent ex-Sirific team and probably Broadcom's). They've been fairly aggressive on process with their 40nm XMM 626 baseband and 65nm UE2 RF chip. They're working on 32nm RF macros for single-chip integration. They've got industry leading 65nm single-chip solutions for the ultra-low-cost 2G segment.
The relevance of this to Atom is obvious, but if it was just for that I'd consider the acquisition a gigantic waste of money (slim basebands are just fine for several more years and integration will increase risk). What's more interesting is whether Intel will try to aggressively bundle Infineon's HSPA+ baseband with their WiFi chipsets and, more generally, their entire notebook processor portfolio. This could accelerate the growth of 3G/4G modules and hurt USB sticks.
What's interesting is that Qualcomm once tried to penetrate the module market directly with Gobi and learned a valuable lesson: OEMs and customers care about some of the features they added there, but they couldn't handle the design flow on their own and had to rely on partners such as Sierra Wireless and Option anyway. Cutting the middle man just didn't work and I doubt it would work for Intel either. Qualcomm is a very strong incumbent in all the markets Intel is aiming at and it's going to be an interesting fight to follow.
Another company that will be affected is Icera. They've had a lot of success in USB sticks lately thanks to higher performance and lower cost, but AFAIK haven't achieved as much traction in embedded modules. They could fail to penetrate that market if both Qualcomm and Intel fight hard for it with both proprietary software and bundling. On the other hand, they could really benefit from this acquisition in the smartphone space where customers may be wary of combining an Intel baseband with an application processor from Texas Instruments, NVIDIA, Samsung, or Marvell.