Moore’s Law, multiple cores – shaping tomorrow’s architecture

Everyone is aware that CPUs are no longer a single core only capable of handling a single thread of activity. Intel and AMD both have produced multi core architectures. Intel’s latest being around the Duo Core, Quad Core and more V8! AMD are also in that arena.

While these architectures are also available to server platforms, there is a major market for these multi-core architectures on the laptops, workstation and PCs that will be the default type of machine to be found in an office or home. With such power and capacity being available locally what impact will that have of the office and enterprise architecture as well as home user and public service consumer?

  • Will it result in a further bloating of a PC’s operating system?
    • Yes, almost definitely – a consumer market will want to see a more visual rich and impressive GUI.
    • But will that be all?
  • What about the Thin client architecture – what will it’s future be?
    • If that depends on the necessity that the central processing power and provision of services that lie with a centralised server then – yes.
    • But what if the distributed clients now have the power and capacity to do the visual rendering as well as efficiently manage any requirement to complete the background computation work. Why send the data to a central core, to await it’s processing and then receive an answer?
  • Hang on what about storage, backup and general business continuity?
    • Yes that will need to be provided, but lets not bloat the hardware if we don’t need to.
  • Will that apply to all application types?
    • No not altogether, I think the main beneficiaries will the productivity and collaboration suites. Where content is generated mainly via HIDs rather other computers, where there is a need to maximise the use of technology in reducing or making up for virtual or distant relationships and allow communication, collaboration and relationship to happen as best as possible. Certainly there will be major enterprise applications like DBMS, and systems supporting ERP or CRM systems which will probably not see a big shift in their architecture styles, which are reliant on the server to provide the computational power behind the application and also attaching to the large storage needed to house the application.

Will the Role of the Server Change?

Quite possibly. The server will need to remain the central core of service provision, where the latest version of data is held, backed up and stored. However, if the clients have the capacity to do the work of the application locally then the traffic between the server and client may well be reduced to the synchronisation of the data.

Peer to Peer application architectures will certainly look to exploit the increases power and resources available in the generic workstation. This may reduce a server’s role to that of a persistently available peer, to allow for asynchronous and synchronous communication for a variety of clients that utilise the application. This may also develop to the point where all instances of the applications are full versions with complete sets of the backend and graphical data sets and also the application executables.

Impacting on our interface

This boost to local resources may also have a major impact in way we interact with operating systems. Until recently any major graphical work required a separate GPU on the motherboard, to release the CPU from this processor intensive work, which otherwise could produce major lags in the application and OS performance. But with multi-core architectures we could see the end of the dedicated GPU, was it becomes absorbed into the main system bus but on dedicated cores Why not use a core on the CPU to the do work, probably rather like the Cell processor introduced to the world via the PS3. This could release the flat 2D nature in which we current navigate through applications, into a more spatially 3D rendered environment, which I see Vista is starting to do with “aero”. Graham and Steve have been debating on the future of the file server see, here, here ,here and here; perhaps the file server won’t go away in the near future but perhaps how we access and relate to file storage will change.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s