Recently there have been rumours that the next version of Windows will move the whole platform completely over to 128-bit architecture, dumping 32 and 64-bit compatibility completely.


This first came about because of a news story broken by Microsoft Kitchen a few days ago in which Robert Morgan, a senior research and development officer at Microsoft posted the following on his LinkedIn profile.

Experience

Senior Research & Development
Microsoft
Public Company; MSFT; Computer Software industry

January 2002 – Present (7 years 10 months)

Working in high security department for research and development involving strategic planning for medium and long-term projects. Research & Development projects including 128bit architecture compatibility with the Windows 8 kernel and Windows 9 project plan. Forming relationships with major partners: Intel, AMD, HP and IBM.

Since then the sweaty masses have, frankly let their imaginations run away with themselves so I thought we needed to look at the situation more objectively, fact by fact.



It’s true that AMD and Intel are working on 128-bit chips, this is something that’s been out there for a while now. But Windows 8 can’t possibly be 128-bit only. The reasons for this are as follows.




Windows 7 is still, sadly, saddled with 32-bit compatibility. This is despite the fact that almost every modern processor supports true 64-bit computing. This is purely to maintain compatibility with older software (primarily in Microsoft’s most important market, business). XP Mode, the virtualisation plug-in that allows you to run older software in a licensed copy of XP, goes some way towards addressing this and many people are hoping that 32-bit support will be dropped from Windows 8 completely. Sadly though, it can’t happen.

This is because in order for XP Mode to work you need hardware virtualisation support on the motherboard and nowhere near enough motherboards yet support this. By the time Windows 8 comes out it’s possible that all motherboards will, but Microsoft have to get started on Windows 8 code now, and they can’t guarantee this will happen.

It is for exactly the same reasons why Windows 8 can never be 128-bit only. By the time the technology is mature enough Windows 8 will be ready to launch but the cost of the chips will still be extremely expensive. They will, I predict, only be used in server farms for hard-core virtualisation tasks on servers.

It’s much more likely that Windows 8 Server will offer full support for 128-bit architecture and that the feature will then be ported to the business and consumer editions of Windows for Windows 9.


The move to 128-bit in itself isn’t such a big job, but it becomes a nightmare when you consider that you also have to maintain compatibility with 32-bit applications. I can’t foresee a situation where we can have an operating system that supports 32, 64 and 128-bit code simultaneously while still remaining stable. The only reason that Windows 8 Server could support it is because the 32-bit legacy support has already been dropped from Windows 7 Server.

For all the people who want to see 128-bit support in Windows 8, forget it. Frankly you wouldn’t get any benefit from it at all anyway. No software will be written to support it for years to come, full 64-bit support in software is only beginning to appear now, and it will offer the average user or gamer no additional benefits over the 64-bit architecture you already have.


Unless and until the whole computing experience we use in our daily lives demands a move to such technology, it will remain in the server farms of the future.