Last week, leading GPU developer Nvidia announced that it planned to start licensing its Kepler GPU technology to other chip designers for use in tablets, smartphones and other new form-factors. The Kepler architecture is currently part of Nvidia’s GeForce 600 GPU, with plans for its incorporation into the company’s next-gen Tegra 5 ARM-based SoC which is expected to be released in 2014. The company’s latest SoC is the Tegra 4, which was introduced earlier this year and has yet to be designed into any major devices. The previous generation, Tegra 3, has so far been incorporated into the Google Nexus 7 and Microsoft Surface RT tablets, and into the HTC One X, none of which are market leading devices.

This is quite an interesting development, another example of ‘unbundling’ driven by market forces. As mobile form-factors continue to evolve (from the various tablet and smartphone models that have been introduced, many of them using Android and ARM), it becomes more and more difficult to have a single SoC that can fit into all these devices. Device manufacturers with the largest device sales volumes – Apple and Samsung – are developing their own SoCs, and, like Nvidia has probably realized from Nexus 7 and Surface RT, the rest of the market may be too fragmented from a distribution standpoint. So, as competition with Qualcomm’s Snapdragon, Intel’s Atom and ARM’s Mali GPU intensifies, Nvidia, with its smaller war chest, has taken the smart step of going the licensing way. This allows it to focus on its strongest asset – the design prowess of its development team – while enabling contract manufacturers to undercut the GPU SoC market on the back of Nvidia’s technology leadership.

So what’s the end game for a company like Nvidia assuming it goes further along the licensing path? With ARM around, it might be difficult for the company to build a stand-alone licensing empire over the long run, so the long-term outcome should be an acquisition by one of the big players: Apple, Samsung, Intel, Qualcomm or ARM. But until then, we might have Nvidia, the company that is working to distill itself to its essence, the company that is using its work-force for what it is best at. Like a start-up, perhaps?

Continue Reading ...

GPS-based navigation app startup Waze was acquired last week by Google for a price just north of $1 billion. Founded in 2007 and headquartered in Israel, Waze had previously raised approximately $67 million in venture capital from investors based in Israel and the Silicon Valley. The startup has almost 50 million users that contribute information on driving conditions, gas prices and other commute-related activities across more than 40 countries. Google beat out rival Facebook in the bidding war for Waze, and in doing so brings into its fold one of its strongest current rivals in mapping.

Quite a bit has already been written about entrepreneurs from Israel, the Start-up Nation, and Waze is the latest example of the impact that a nation of less than 8 million people continues to have at the frontiers of technology and business. From the impact that ‘learning on the job’ at IDF has on the work-culture and technical knowhow of its citizens, to the understanding in the populace that success stems from the ability to sell into large markets (whether United States or Europe or Asia), Israel continues to be the one location outside the United States where the VC model might be most applicable. A startup is about selling products and services that solve customer problems, and in today’s connected world, this can be all done online by teams of technically talented doers and domain-knowledgeable marketers and salesmen. Through decades of hands-on experiences, the Start-up Nation seems to have mastered this process. So, while it might just be a rumor that part of what clinched the Waze deal for Google was their willingness to let Waze stay in Israel, that rumor wouldn’t be difficult to believe if it were true.

Congratulations to Waze for becoming a huge example for entrepreneurs to follow!

Continue Reading ...

Last week brought news of a partnership between IBM and 10gen, whereby the two companies would work together to achieve interoperability between IBM’s DB2 relational DBMS and 10gen’s MongoDB NoSQL database. As part of the partnership, IBM will push MongoDB as a core NoSQL database for enterprises building web and mobile apps. At present, DB2 is second in market share in the RDBMS market behind Oracle, whereas MongoDB is the leader in the nascent NoSQL market, apparently owning almost half of the market. Both players see significant competition in their respective segments – especially MongoDB, competing against DynamoDB, Couchbase and so on in a rapidly evolving market.

This feels quite a bit like an alliance of the needy, and a potentially effective one at that. The emergence of NoSQL has a lot to do with the high availability requirements of mobile and web-based collaboration. In that context, IBM gains some good street cred by partnering with MongoDB as the promise of unlocking insights from semi-structured and unstructured data increasingly drives enterprises towards NoSQL. On the other side, with competitive solutions such as HBase for true big data applications around enterprises, MongoDB could use a channel like IBM to penetrate the broader market.

VCs have funded multiple NoSQL approaches over the past five years, so the realities of monetization are probably starting to set in now. In that context, IBM’s announcement should set in motion a wave of consolidation here. While the days of one-size-fits-all databases might be over, the reality also is that the NoSQL folks might not be able to scale without a larger mother-ship. So then, it’s time to get ready for a few exits, let’s stay tuned…

Continue Reading ...

When it comes to smartphones and tablets, the OS market is quite concentrated, with Android and iOS together owning a 92.3% market share for smartphones, and a 96.1% share for tablets. Well, Mozilla and Foxconn feel it is time to fragment it a little. As of June 3^rd, the two companies announced that Foxconn will be introducing devices from smartphones to televisions and large display boards, all based on the Firefox OS. Foxconn and Mozilla plan to develop reference designs, but do not plan to introduce devices with their own brands. Foxconn represents Mozilla’s 19^th Firefox OS alliance partner (the previous 18 all being mobile operators, with Sprint being the only major American carrier in the list).

Good strategic move for both Foxconn and Mozilla. As the top outsourced manufacturing shop for mobile devices, it is in the Foxconn’s best interest to fragment the mobile ecosystem to the extent they can. As for Firefox’s browser-as-an-OS approach, this might be their best bet for relevance in the post-PC era. The bigger question though is whether the world needs a third mobile OS, and on a couple lines of thought, the answer may be a no. Chrome and Safari will both be technologically better equipped if the world chooses the mobile web approach over installed apps, from the knowledge base of their respective mother ships. As for the next five billion smartphone users, cheap android phones are already available for $50, and that’s even after discounting the fact that Mozilla’s brand recognition might be low amongst this next wave of adopters.

Will Foxconn’s reference designs bring enough long-tail device manufacturers to make Firefox OS relevant? Will Mozilla’s 18 carrier partners consider introducing their own smart device brands? Will there be enough firepower around Firefox OS to help Mozilla move beyond just playing third fiddle, when other well-funded operating systems such as Windows 8, Blackberry and Symbian have failed to do so? Well, the only things that are clear from last week are that mobility is huge, and that another player from the PC ecosystem wants in on a piece of the pie…

Continue Reading ...