LebGeeks

A community for technology geeks in Lebanon.

You are not logged in.

#1 January 11 2010

J4D
Member

thoughts about how to advance computers from a hardware point of view

I had a thought the other day about having a non volatile memory portion on the motherboard ,about 2 GB of nonvolatile memory allocated for the OS . that way the computer should boot wickedly fast . and the full ram would be used for other operations . plus would remove the need for a hard drive to be present at all times . that way allowing for new and advanced , slim , fast new platforms :)

Offline

#2 January 11 2010

pikatore
Member

Re: thoughts about how to advance computers from a hardware point of view

I think that's a good idea. Though if I'm correct, isn't SS technology still in it's early stages, regarding the lifetime of a SSD?

Offline

#3 January 11 2010

J4D
Member

Re: thoughts about how to advance computers from a hardware point of view

a few versions of  SSD have hit the market , which means the number of read/write cycles now is quite feasible .

Offline

#4 January 11 2010

arithma
Member

Re: thoughts about how to advance computers from a hardware point of view

Intel SSDs are quite good, but still not pretty mainstream. Last checked, 120GB could cost up to $500

Offline

#5 January 11 2010

rolf
Member

Re: thoughts about how to advance computers from a hardware point of view

I want to put rendering functions like html rendering from software to chipstets. Move more, more, more to the chipset.
Also unify ram and ssd storage.

Last edited by rolf (January 11 2010)

Offline

#6 January 12 2010

arithma
Member

Re: thoughts about how to advance computers from a hardware point of view

According to Intel, the biggest upgrade for performance you can do since conroe (core 2 duo) is SSD.
Moving things to hardware only means evolution will be a lot slower and we won't see such a big gain. Going back to chipsets is like going back to 80s. Sound mixing is now done on the CPU rather than in hardware. It's more flexible.

Offline

#7 January 12 2010

Xsever
Member

Re: thoughts about how to advance computers from a hardware point of view

Arithma is right. Although CPUs have been growing in speed and complexity, hard drives remain the slowest part in a PC. SSDs seem to be the next thing, but it is going to take time before they become mainstream.

Western Digital's Velociraptor Hard Drives are still an option without breaking the bank.

Offline

#8 January 12 2010

rolf
Member

Re: thoughts about how to advance computers from a hardware point of view

arithma wrote:

Moving things to hardware only means evolution will be a lot slower and we won't see such a big gain.

I think sound mixing moved from the hardware to the cpu because its cheaper this way.
But I still think real evolution comes from the hardware. Software evolution might be nothing but bug fixing and constantly improved icing on the cake.

Last edited by rolf (January 12 2010)

Offline

#9 January 12 2010

arithma
Member

Re: thoughts about how to advance computers from a hardware point of view

*slowly moves away*

Offline

#10 January 13 2010

nablaux
Member

Re: thoughts about how to advance computers from a hardware point of view

rolf wrote:

I think sound mixing moved from the hardware to the cpu because its cheaper this way.
But I still think real evolution comes from the hardware. Software evolution might be nothing but bug fixing and constantly improved icing on the cake.

I think this is a bit too extreme rolf. hardware and software are two sides of the same coin. In one hand , you need good, reliable hardware to be able to run software and you need more powerful hardware to run more complex software.
However, coming from a software background, it is not only the advances of hardware that made computers today as fast, reliable etc... as they are. The whole science of algorithmics and designing algorithms have contributed a huge deal to this advancement. Just a small example, finding the algorithm that solves the following problem "Given a set of points in the plane, find the closest pair", has evolved the computation time from O(n^2) to O(nlogn) and that is speaking in 2D. Now, this actually means in practice that given millions of point like in a GIS system processing went down from a few days to a few seconds. THAT IS SOFTWARE.
But still I agree how many instructions a hardware can process etc... has made things faster and feasible. As for getting hardware solutions for problems, it is indeed too expensive.
Software also contributes in bringing technology to the masses, like for example if GPS was a pure hardware solution, no one could have afforded buying one but since it is mostly software algorithms it really costs nothing to implement, upgrade etc... and it becomes cheap.

Anyways, I hope you get my point.

Offline

Board footer