Does the Hewlett Packard saga hold any significance for the mobile space?

The recent Hewlett Packard saga with Autonomy has roots stretching back to 1999 (for an in-depth briefing paper on HP fortunes to date, with its main product segments and analysis of future possible directions, plus a podcast, please click here).

With the latest developments in the Autonomy case, HP’s tale now seems to have become even more bewildering, with accusations of impropriety, calling in of the relevant financial authorities, challenging open letters to the HP board, while Moody’s cuts HP’s debt rating a point, to the lower-medium grade range.

Does this matter? For the mobile and broadcast industries, with their focus on spectrum availability and new broadband advances, such happenings seem to be entirely disconnected, taking place in another world, despite the talk over the last 20 years of the ‘convergence’ of the two sectors with the computing industry.

But, stepping back, such events are ever more important as the move to media content over the top (OTT) for web access with mobile broadband tends to emphasise the role of the computer suppliers in the mobile and future broadcast industries, with impacts really at two levels.

• Servers – notably blade servers for dense rack mounting and a web farm pre-requisite
• Specific software applications in fields that are important to mobile, such as the Autonomy enterprise search engine, cloud toolkits and database utilities, especially ‘big data’ mining, storage and management
• Data centre engineering, particularly for a low carbon footprint.

In general, the future role of the systems suppliers in the mobile industry will become even more important for the back office, transaction, digital ingest and media distribution, as rich web services, streaming and downloads power the OTT world of mobile broadband, while m-commerce takes off.

To power all this new media stuff and expanded business models based on tablets and smartphones, the spectrum must be used far more effectively, especially to compensate for LTE’s hunger for bandwidth. Reframing of traditional allocations, sharing and offloading using alternative protocols are the future (see PolicyTracker’s latest research report on the second didgital dividend, Strategic implications of the Second Digital Dividend). 

Thus the second role of the computing, software and semi-conductor industry comes into play.  The consumer depends on ever high computing power, using energy efficient technology, for those smartphones and tablets and also for base stations of any kind, be they LTE, or WiFi or whatever, with their non-stop operating systems.

 Low power microprocessors form the underlying platform for these devices – usually some variant of an ARM design for a system-on-chip (SoC) from a range of suppliers such as Samsung, MediaTek, NVIDIA, TI, or Qualcomm, with others, such as Apple, which designs its own versions. It also includes a graphics processor, for instance from a specialist such as Imagination Technologies. The SoC works with an air interface chipset, often combined to form one chipset. The next round of radio interface chipsets are coming from the likes of Qualcomm, Seqans Communications, Altair Semiconductor, MediaTek and soon Nvidia and others. Many of the next generation of sub-$200 Chinese smartphones and tablets are based on the MediaTek chipsets for Android.

In consequence, the leverage on the mobile industry of the descendants of the original computer and semiconductor industry is liable to expand as greater signal processing capability is now demanded.

This is not just for smartphone graphics and other sophisticated features but for future spectrum management regimes, especially for sharing spectrum.

 Sharing will increasingly depend on intelligent front ends, with a software defined radio, to produce a cognitive radio which can adapt to ambient signal conditions, and/or with use of geolocation databases of signal uses and strengths. Future modes of use will all require higher processing power in the mobile device and access points, but with lower energy demand.  This will support multiple protocols, including new licence-exempt politeness procedures, white space protocols plus far sharper filtering, directional working and adaptive signal selection.  Such technology will also creep into the future internet TV and set top box markets. 

 The new spectrum management regimes will enable the explosion in media and web services over radio bearers, both mobile and fixed, which harbour the underlying profit motive behind new spectrum management. However the door to harnessing all that new spectrum requires ever greater DSP power.

Thus both effects tend to increase the leverage of the software, semiconductor and systems providers over the mobile industry and increasingly over the broadcast industry as digital terrestrial broadcast is replaced by IPTV for nonlinear and linear TV and audio.

 Very simply, raw spectrum bandwidth is being progressively replaced by computing power in both the handset and the networking data centres. New network architectures for media distribution may also drive distributed intelligence in the mobile networks, both RAN and core, towards the edge and closer to the consumer.

 Why the destiny of HP matters to operators in the spectrum domain is more to do with the first role, of data centre infrastructure able to exploit the services that efficient spectrum management enables.

But in the past, major systems suppliers such as HP also designed and produced processors that eventually turned up in servers and laptops. They might have possibly even been used in future handsets and tablets if low power versions had been developed.

 Moreover these ICT suppliers will also be key players for the next step for mobile’s union with the internet – the Internet of Things (IoT). That will demand lots of ‘big data’ processing and well as billions of ultra low power devices in the field, data routers and concentrators for sensor networks.

 These are all challenges that the ICT suppliers relish, especially as spectrum demands bite harder, requiring more sophisticated signal processing – and then even more computing and storage power to gather the data seeds – and then a lot more to process their flowering into information, knowledge – and eventually even wisdom (?) (after TS Eliot).