Icera another option for TI
Sunday, May 18, 2008
In my Texas Instruments (TXN) valuation series, I pointed out that the company did not have a strong 3G product strategy. I subsequently also discussed in detail the synergies that existed between InterDigital and TI. I had also noted that if not InterDigital, TI could consider Icera also as a viable option to buttress its 3G roadmap. Let us take a quick look at this possibility.
Bristol-based Icera Semiconductors has one of the better performing HSDPA products in the market today. The company’s soft modems have been gaining traction in the data-cards market. The company recently announced a merger with Sirific Wireless. Sirific’s RF front-end and digital RF expertise is expected to complement Icera’s Livanto soft modems. Icera also announced a mobile phone design win recently. Combined with Sirific, Icera has raised over $200mn till date. The company is also looking to raise another round on the strength of its recently announced design wins and the broader market traction it has been receiving of late. More on Icera can be found here.
Texas Instruments, as I have been lamenting once too often in this blog, does not have its own 3G-baseband roadmap. The ‘foundry’ model worked for the company so far. But with Nokia and Ericsson sourcing from multiple vendors, its situation has become precarious. Besides, Nokia also moved its chipset engineering team to STM which after its recent merger with NXP has become a larger force to reckon with. These industry-events together with Broadcom's (BRCM) aggressive ‘3G on a chip’ campaign spell bad news for TI.
TI has to act soon. Pacts with InterDigital (IDCC) and/or Icera to bundle their baseband solutions with its application processors or even to offer a combined OMAPVox solution are possible. Or, if TI desires, it can attempt to acquire one of these companies. IDCC, with its strong IP is unlikely to be a cheap acquisition target. Besides, there is a chance that IDCC’s patents may only incrementally benefit TI in patent cross-licensing discussions given the latter's already strong IP position.
A good alternative then is to acquire a focused product-centric company such as Icera. With the scale of investments, Icera may be a cheaper option to pursue than IDCC. The company’s expertise in soft modems may complement TI’s own DSP expertise to create niche products for the future. With its support network, TI can leverage this unique technology to stay ahead of its competitors in product launch as well as performance.
Of course, the other question to ask is if either TI or Icera will consider its path. For TI, the RF expertise in Icera may not add much value. So, the UK-based company may have sounded more attractive to TI before the Sirific merger (though an additional $100mn is not much for the semiconductor giant.) Icera, on the other hand, is also positioning for an IPO in the 2010 timeframe and Sirific is seen as a strategic addition in this direction.
While Icera seems to have big plans, TI may also just resign to its dwindling 3G baseband business. With the convergence devices movement slowly heating up, the company also finds itself competing with more players on the applications processor front. So, if it does not buy or partner with someone now, it will have to think ahead, start on 4G and hope that it can recapture any lost market share at that time.
3 comments:
Arun,
Sent you an email..You could alternately use the contact me form that I put in recently. Feel free to contact me privately.
I think a 4x return would be reasonable for a semiconductor company. I have not valued Icera per se and should perhaps do that with my models. My gut is $1bn may be on the higher side of my valuation. On the other hand, if you read my IDCC+TI articles, the cost for TI will be much higher (~$3bn). So, Icera may still be a good deal for TI at around $1bn.
Vijay
Cheers!
Well, soon Icera will have $200B of VC funding, and Stan Boland managed to grab that much VC funding despite minimal revenue so far is because of what he achieved with Element14: a 32 times Return on Investmenr for VCs. Yes, 32; it was before the internet bubble burst, but you get the idea. First-round VCs are likely expecting 10x+ RoI at this point, which is why I think even $1B is on the low side of what they'd want.
However, if you're right and InterDigital might cost $3B, then Icera would indeed still be 2-3x cheaper - which does make them a very noteworthy strategic option, especially given they're as good performance-wise and very likely better die size/cost-wise. On the other hand, their custom logic approach means they might not be the best option for SoC integration because they lag behind process node-wise.
It's a complex debate certainly... :)
Posts
I agree with your points, but I think you're underestimating how much Icera management and VCs would want for the firm. They seem very confident they can achieve a $1.5B+ IPO; so I can't imagine them considering any proposal for less than $1B, and even then they would likely hesitate substantially because Stan Boland is one hell of an ambitious guy.
So being optimistic, TI might be able to buy Icera for $1B while they'd have to pay, I don't know, $1.5B for InterDigital. That's not that huge of a difference, really.
BTW, I've been Googling to try to find a way to contact you privately but came back empty-handed. I'd appreciate if you could just quickly send me an e-mail at arun AT beyond3d DOT com so I can take contact; there are a few industry happenings I'd like to discuss but I don't think I can do so publicly... :)