Linked by Thom Holwerda on Mon 7th Apr 2014 22:17 UTC
Microsoft

It looks like the Internet of Things could be the next big computing battleground, and Microsoft seems willing to sacrifice a few battles in order to win that war. Facebook is chasing virtual reality; Google wants home automation, smartwatches, and internet-connected glasses. More than 200 billion devices are likely to be connected to the internet by 2020, a huge example of the way the technology industry will shift and new battles will emerge. Satya Nadella believes the future isn't Windows desktops, Windows tablets, and Windows Phones. It's not Windows everywhere, it's Microsoft everywhere, offering software and services for every device - including an entire world of interconnected devices that have yet to be built.

The speed with which is doing this u-turn makes it quite clear that people within the company wanted to do this for a long, long time (otherwise it could not have been done this quickly), which implies that Ballmer may have simply held these changes back.

The elephant in the room here is that while people talk about Microsoft as if the company is down and out, it's still hugely profitable and has consistently been posting great financial results. It's just that Microsoft's money isn't coming from sexy products like smartphones and tablets, but from enterprise and backend stuff - stuff the technology press either can't write about, doesn't understand, or both. It's very similar to all those articles claiming Apple no longer innovates and disrupts, even though the company sent shockwaves through the microprocessor world.

In any case, it seems like Microsoft finally found the right direction in this new world.

Thread beginning with comment 586626
To view parent comment, click here.
To read all comments associated with this story, please click here.
oiaohm
Member since:
2009-05-30

"Just because Microsoft is still profitable now question is how far into the future.

Microsoft has massive government and military contracts that will keep the cash pouring in for a long long time.
"
Those are where the ice is cracking badly.
http://www.computing.co.uk/ctg/news/2337233/london-borough-to-roll-...

You claim Microsoft is doing fine is wrong. Its those massive contracts that are disappearing.

I am not claiming the year of the Linux Desktop at the moment this is more the year of the VDI in large enterprise. Result is less copies of Windows and MS Office required. VDI provided work force only need number of active copies.

Then you have other governments paying for extension to XP for another 12 months.

What exactly do you think the future looks like for the linux desktop?

What you will see in sometime end of this year start of next a battle of dominance. Wayland is convergence. Convergence in the most scary ways possible. There will be reduced application difference between a Linux desktop and Linux Phone. Sailfish OS phones are already using wayland. The work on wayland is bringing render nodes into existence on Linux.

Render nodes brings the means to share a gpu between many users. So now a multi-seat set-up with like 5 users to 1 PC can now have full 3d acceleration. Yes the 1 to 1 model of MS clients will not exist in the Linux world as the dominate option. Multi-seat could also include VDI.

Gets worse that copy of Windows you have running inside a virtual machine can be passed render node access. This brings a new form of scary. Windows XP not getting new drivers will this matter. The answer is no if all the alterations that are happening to Linux. Windows XP or Windows 7 inside a Linux hyper-visor separating them from the real hardware will be able to achieve 90 percent+ of native performance. Historically windows depends on the fact you cannot run it on newer hardware due to no more drivers so you are forced to OS update.

The changes in to the Linux desktop are massive.

Linux desktop after wayland deployment cannot really be compared to the time of the Linux desktop using X11. X11 never supported officially sending 3d rendering to many users sharing the 1 video card. Never support proper secuirty either.

Linux kernel lack of blue screen of death equal is also solved. No more ice-station wondering of the kernel has stopped or not.

X11 will at some point be nothing more than Legacy support on Linux.

I cannot tell you exactly what the Linux Desktop will look like in 12 months time. But I can tell you now that what I will allow will change the world of VDI.

Please remember chromebooks most can be converted to full Linux installs but cannot be converted to Windows. The candy for Linux Desktop growth has been spread.

Reply Parent Score: 2

WereCatf Member since:
2006-02-15

Render nodes brings the means to share a gpu between many users. So now a multi-seat set-up with like 5 users to 1 PC can now have full 3d acceleration.


I'm just skipping the rest of your rant and focus on this: how many people would actually have use for that? It's a fun gimmick, sure, but you're touting that as some sort of a major feature that will really shape up the landscape, but alas, home users won't have any use for that, very few -- if any -- companies will have use for that, and the people who actually need a powerful GPU for heavy lifting won't be sharing it with others.

Reply Parent Score: 3

oiaohm Member since:
2009-05-30

WereCatf companies are already having uses for it. Render node sharing allows far more control.

Really if you look at a case of blender movie production examples and other real worlds you find out something bad. When you need a really big GPU the reality is you will want to be able to take up any slack usage.

Multi seat allows light users to share their unused GPU time with heavy users.

There is more than 1 way to slice up the GPU using Render nodes.

Render Nodes have the name Nodes for a reason. A user on a node might in fact be using many Nodes at once. Each node being a individual system.

The means to share the GPU to more than 1 user opens up a stack of new options.

Basically Render Nodes work both ways. Means to break up gpu power between many users inside a single node/system as well as combined the power of many GPU's on many different nodes/systems into one super processing solution.

Interesting right. Light users can be stacked on reduced video cards as well so leaving more GPU processing space for heavy users.

Reply Parent Score: 2

TechGeek Member since:
2006-01-14

"Render nodes brings the means to share a gpu between many users. So now a multi-seat set-up with like 5 users to 1 PC can now have full 3d acceleration.


I'm just skipping the rest of your rant and focus on this: how many people would actually have use for that? It's a fun gimmick, sure, but you're touting that as some sort of a major feature that will really shape up the landscape, but alas, home users won't have any use for that, very few -- if any -- companies will have use for that, and the people who actually need a powerful GPU for heavy lifting won't be sharing it with others.
"


I will tell you who has need of it. Game companies who want to provide remote access to games. Companies like Amazon are working on technologies that will do the video processing on the server side allowing a low powered device on the front side to access it.

Reply Parent Score: 3