Memristor Computers

The memristor is non-volatile memory computer component that could revolutionise computing.

It has the potential to revolutionised computing in a number of ways.

Memristors can store data without needing constant power, so they will reduce power need and increase battery life. And being non-volatile can massively reduce system boot/power on/resume/app loading times.

Memristors should be smaller so more memory and being lower power should run cooler allowing for memory stacking and therefore much larger memory.

But the real fun begins when Memristors are integrated with transistors to create flexible programmable logic units, think field programmable gateway flexibility but built into standard chips.

Then why not combine the logic and memory into single chips to get around the memory bandwidth bottleneck problems of current system.

So we could have flexible programmable memory and logic chips where the processing can be done within the memory where the data is stored.

Can you imagine the level of power next generation memristor-powered gaming systems could have?

Then we will just need to figure out how to program adaptive/programmable massively multi-data-computing systems.

 

Ah, yes. The memristor. Coming the same year as Linux on the desktop.

How does this compare to the current progress with graphene? Or is this comparing apples and oranges?

 

Actually, Intel and Micron is coming out with something similar called 3D XPoint.

https://en.wikipedia.org/wiki/3D_XPoint

Intel usually doesn't announce things until they know they're going to manufacture the thing. 3D Xpoint looks cool. It's fast, and it's persistent regardless of whether the memory gets power or not. It's supposed to be slower than RAM but faster than NAND-type flash memory.

 

Graphene is just a fabrication material technology. The memristor is a 'new' electrical component that fundamentally changes how computer memory and even logic can work.

 

It'll probably be like the Pony Express and the telegram, where quantum computing will come out almost immediately afterward and kill it.

FPGA's are so much cooler though.

 

Pony express lasted forever compared to HD-DVDs and Zip drives.

On the other hand quantum computing may end up in the same area as cold fusion. Perfectly sound in theory. Impossible to build.

 

You should've said "same year as GNU kernel".

Linux for desktop has been available for 7 years or so, as long as you can catch a wild sysadmin and trick him into maintaining the damn thing for all the eternity.

Been using KDE4 years ago. Experience was superior to windows platform. However, some of the bundled software had a bit unfinished feeling about it, and you were consantly hitting situation where you need very specific windows tool but WINE can't launch this one yet. Things could've improved since then.

In my experience, ye mythical average user doesn't know and doesn't care what OS is running on the computer, as long as he/she knows which "picture" to click with mouse.

 

I assume you aren't aware of the meme (at this point) that is "the year of the Linux desktop" that has been predicted for at least a decade?

 

Very few have hurd of that though.

Linux has had some success on the desktop but the meme refers to it being the predominant platform.

 

Ain't much of a meme if I'm not aware of it.

Please excuse me while I laugh hysterically at this idea.

Spoiler

That won't be happening, unless someone big decides to back the whole idea. And that someone big will just make a their own fork while ignoring the rest of the linux world.

 

It's a common phrase especially among news articles. There are typically a few every year. Like this one from 2014.

http://www.infoworld.com/article/2844225/will-2015-finally-be-the-year-of-the-linux-desktop.html

Or this one discussing Windows 10.

http://www.everydaylinuxuser.com/2015/09/has-windows-10-killed-linux-on-desktop.html

 

Because I don't monitor those kinds of resources, it need to be really big deal (as in: generate wave of noise all over the web) for me to notice it. As I said: not much of a meme. Just some guy under the rock secretly wishing for a miracle, once again.

 

This. Linux is a techies wet dream. As long as it stays techie focused it won't ever become mainstream. And if it's not techie focused, is it still Linux?

I understand the appeal of the idea of desktop Linux, especially if you've ever worked in support. In theory if everyone was smart enough to use Linux they would also be smart enough to turn the computer on. Or in our case find the make an MMO button on their own.

 

Year of the Linux Desktop has been predicted since before Windows 95 but it's just a programmer's toy really. It, FreeBSD, and all those are really good for study and use in university computer science curricula and for cheaply housing obscure government and business amalgamated data in databases.

 

Yes. At least so long as it maintains the same degree of freedom.

 

A Zip what?

click
click
click
click
...

 

That drive you keep a pair of tweezers around for because people keep trying to put 3.5-inch floppies into them.

 

But like most things developed by programmers for programmers, it's loaded with features without any regard for how it's actually supposed to be used. I've been working on a new install of openSUSE and running across many lovely quirks with networking (yay for broadcom wireless chipsets) that only make sense in a systems that's expected to have dozens of computers networked through it.