Oracle has rolled out a new high-end Sparc-based cluster server, a 16-core T3 version that seems to close off any debate on whether the company is serious about the hardware business. In fact, our rumor mill and survey data show that Oracle may be ramping up for a major effort in 2011. Software is the only place in tech where you can really build differentiation quickly and hope to sustain it for long enough to capitalize on your success. We’ve already seen software taking a larger role in networking, and what we’re seeing with Oracle is a software company exploiting the engagement that differentiation can bring to move itself into becoming a mainstream, full-service, IT player.
The Oracle blitz would have a significant competitive impact. Both IBM and HP, the incumbent giants, are relative lightweights in software relative to HP. IBM lacks lacks any strong connection to the networking space. Oracle’s server strategy focuses on database networking via appliances and through Infiniband, which is an alternative to the much-touted Ethernet-based data center networks. That runs in opposition to the data center network strategies of not only IBM and HP, but also of Cicso, who is potentially impacted by a big move by Oracle into servers and data center networks for a bunch of reasons.
Then there’s SAP and of course Microsoft. SAP unveiled a new real-time data analytics appliance, showing it’s going to shift more in the hardware direction, and there are rumors that it would be going even further in that direction, even as there are rumors that HP wants to do some serious software deals in 2011. Microsoft might be left as the only relatively pure software play, except that there are also rumors Microsoft might be looking at the appliance game, not only for database products but also for collaboration and other middleware elements. “Cloud-in-a-box” Azure-inspired technology is already sold by Microsoft partners, but Microsoft realizes that many of its hardware partners may end up being competitors if the appliance business really takes off. Still, Microsoft get into any sort of hardware is a big risk to its current partnerships, and unless they see real stress cracks they’ll probably go slow.
There’s a curious fast/slow balance going on with regulations relating to the Internet. The FCC’s December 21st meeting will be critical in setting out the details of what looks like a complete FCC reversal of perspective. Genachowski’s stance on neutrality has historically been “pro-Internet”, favoring OTT players over the ISPs/telcos. His statement regarding the forthcoming order is almost the opposite, admitting explicitly to the value of usage pricing and implicitly supporting multi-tiered services and “special” non-Internet-but-IP services as well. Further, the order appears to have little chance of being upheld if it’s appealed, raising the question of whether there were private discussions among the players that secured a promise not to appeal. But I’m also hearing that the vote on the 21st isn’t a sure thing at this point and the whole deal could fall apart at the last minute.
Other issues relating to lawful intercept and privacy are demonstrating that the Internet isn’t the PSTN. The problem is that in the current IPv4-dominated online model, we don’t have permanent addresses for service access points or devices, and the transient nature of IP addresses makes it much harder to identify a specific online user to either apply intercept warrants or to record their tracking preferences. Add to that the fact that many of the potential trackers are outside the reach of US law and you have some ugly potential. We’re actually building a pretty good argument for mandating an Internet shift to IPv6, but it’s not at all clear that there’s anybody who could make the mandate stick other than the US government, and that agency is apparently unable to agree on even fundamental points of taxing and budgeting.
A final point comes to us courtesy of Pew’s research, focusing on what people know about the world around them. Not much, it turns out. What’s frightening about the results isn’t even the somewhat-expected fact that people know nothing about government, geography, history, or current events (outside of sports and gossip). They also know little about technology. Only about a quarter know that Android is Google’s operating system, for example. Here we sit amidst the greatest collection of readily accessible knowledge of all time, and we’re getting stupider. That has significant consequences for political policy, but also for our ability to manage our own technology decisions. I’ve noted a similar effect in our enterprise surveys. This is important because it’s telling us that the buyers don’t understand what they need to. A solution might be to educate them, but not only don’t vendors really want to do that, the buyers themselves don’t want it. “I don’t want decision support; I want to be told what to do!” is the typical survey response. Buyers want a trusted agent, a company they believe in and will bet on to lead them through technology issues they believe they’ll never fully grasp. It may be that becoming such an agent will be the critical path to success in the coming year.