HP, at its annual analyst event last week, presented what could be a one-two punch that could have the company flipping the current networking, storage and server market on its ear and leaving it in the position IBM once was with the mainframe. With a lead so pronounced, it is virtually untouchable.
IBM has a similar opportunity in terms of spinning the market with a variety of technologies it has under development -- like neural networking -- but it is still over five years out, while the first punch from HP is within months and the second well within the five-year horizon.
But currently HP has spread itself so thin that it may succeed with product but still not reach its full potential due to under-resourcing.
Let me explain.
Currently, in its class (HP and Cisco lead the networking market; HP is larger than the next five competitors combined), HP is the leader in software-defined networking and second only to Cisco in networking in general. Software-defined networking has significant cost and agility advantages over traditional hardware and can dramatically lower costs, which is why Cisco appears to be dragging its feet with the technology.
Cisco revenues and profits are tied closely to a telecom-sourced ecosystem that predicates relatively expensive hardware (when compared to servers and storage) and substantial labor costs. As a result, it is reticent to push this envelope hard because it would prematurely sacrifice revenues it currently is getting, which is likely why HP is reporting an increasing number of design wins.
HP is taking this same concept and applying it to both storage and servers aggressively; it is hardly the only one creating software-defined servers. But in its class it is currently the most aggressive with both ARM and Intel Atom architectures to get there, suggesting it will be able to take a sharp cost advantage into the other two areas of this market as its Project Moonshot matures, at least for implementations that will take advantage of highly configurable, relatively large numbers of low-cost cores.
HP showed volume silicon for its Memristor technology, which will then take this software configurability down a potential magnitude of cost. Memristors are memory components that can use system memory for performance and flash memory for storage. They can also do processing, though that capability, so far, hasn’t proven as effective as more traditional processors. Prior estimates had this part in market by 2014 and most HP competitors thought this was overly aggressive. The fact that HP was able to showcase manufactured silicon suggests it may actually be on track for a 2014 event.
What I think most folks don’t get is that this is basically a one-two punch because HP is refining the software and tools using what it currently has and then plan to slipstream in the memristor technology once it is available. This would allow HP to have, by the end of 2015, a potential massive cost advantage with a relatively mature and tested ecosystem surrounding this as-yet untested technology and vastly cutting down the months and years it would typically take to mature and get the market to accept the resulting products.
However, standing against this opportunity is HP’s current ability to resource strategic efforts like this adequately. Changes like this require a massive effort to market the advantages and technology in order to get buyers to accept it. Currently, HP isn’t able to fund a decent campaign surrounding its potentially leading networking gear and those products are available today. This is because, unlike HP’s competitors who are tightly focused on the segments of the market they are most interested in, HP remains broadly spread from industrial printers all the way across to consumer technology -- most of which are equally starved for marketing and sales resources.
If HP can’t find the resources to properly fund the areas it is leading today, it is unlikely to come up with the massive additional funding to convince the market to accept the far more powerful, but far less known, technologies of tomorrow.
This has been a historic problem for HP; it has consistently been out-marketed by IBM, Apple, Samsung, Cisco and even Oracle. HP was even unable to retain its Apple-trained consumer marketing team and lost its first-place position in the PC market partially as a result.
It is rare a company has the potential to flip the market like HP has, and I’m not just talking storage, servers and networking either, because this would have a massive impact on the hardware and support costs for the cloud solutions that would most benefit from the related advancements. But HP’s historic problem has been an inability to leverage marketing and fully take advantage of the lead it has. For instance, it had e-commerce solutions that were well in advance of IBM’s a decade ago, but IBM out-marketed HP and eliminated that advantage.
HP will likely need two things to fully take advantage of the potential of its efforts: Fully fund them as strategic to the firm’s future and demonstrate a marketing skill in line with IBM and Apple. Otherwise, someone just like me in about a decade will be referring to this missed opportunity like I just did e-commerce and with even more regret because the potential here is far greater.