A decade from now ...

[Terry McCracken]

A decade from now, the typical workstation or high end consumer PC will have 16 cores at 8 GHz, 128 GB RAM, 32 TB disk storage, 256 GB removable optical storage, a meter size LCD with 10,240 by 6,400 resolution, and an always active gigabit link to the Internet.

A decade from now, how will your chess program be different from what it is today?

The HDD will be a thing of the past. Electronic and or Optical storage will replace it. CPU's will be quite different due to spintronics et al.

Welcome to Future Shock Regards...

[Aleks Peshkov]

From the above article:

The DRAMs also have relatively larger access times as compared to SRAMs. Further, the dynamic nature of the DRAMs accounts for their reduced performance as compared to SRAMs. However the simplified design structure and cost feasibility has made DRAMs the choice for main memory.

[Dann Corbit]

From the above article:

The DRAMs also have relatively larger access times as compared to SRAMs. Further, the dynamic nature of the DRAMs accounts for their reduced performance as compared to SRAMs. However the simplified design structure and cost feasibility has made DRAMs the choice for main memory.

True but the speed of SRAM is also not increasing as fast as the CPU increases. I suppose it would be possible to use gallium arsenide sram but then nobody could buy the computer.

[smcracraft]

A decade from now, the typical workstation or high end consumer PC will have 16 cores at 8 GHz, 128 GB RAM, 32 TB disk storage, 256 GB removable optical storage, a meter size LCD with 10,240 by 6,400 resolution, and an always active gigabit link to the Internet.

A decade from now, how will your chess program be different from what it is today?

I'm not sure that will be enough to run Windows properly in the future.

Cheers,
Tom.

Forget Windows. Get Ubuntu GNU/Linux, Debian, SuSE, etc.

Get into the platform of the future today.

And have access to your source code!

[Tommy]

Forget Windows. Get Ubuntu GNU/Linux, Debian, SuSE, etc.

Get into the platform of the future today.

And have access to your source code!

Linux is looking better everyday...

http://searchsecurity.techtarget.com/ne ... 95,00.html

"This stuff just takes a knife to a large part of the security mesh Microsoft built into Vista," Dai Zovi said. "If you think about the fact that .NET loads DLLs into the browser itself and then Microsoft assumes they're safe because they're .NET objects, you see that Microsoft didn't think about the idea that these could be used as stepping stones for other attacks. This is a real tour de force."


Cheers,
Tom.

[MikeB]

A decade from now, the typical workstation or high end consumer PC will have 16 cores at 8 GHz, 128 GB RAM, 32 TB disk storage, 256 GB removable optical storage, a meter size LCD with 10,240 by 6,400 resolution, and an always active gigabit link to the Internet.

A decade from now, how will your chess program be different from what it is today?

Rest in peace SJE, you were pretty damn close with your prediction.

[smatovic]

A decade from now, the typical workstation or high end consumer PC will have 16 cores at 8 GHz, 128 GB RAM, 32 TB disk storage, 256 GB removable optical storage, a meter size LCD with 10,240 by 6,400 resolution, and an always active gigabit link to the Internet.

A decade from now, how will your chess program be different from what it is today?

Rest in peace SJE, you were pretty damn close with your prediction.

Pretty close.

Considering that in 2008 we already had 4 core CPUs,
he probably took an slow down of Moore's Law into account,
therefore the 8 GHz.

The only thing he missed are SSDs and GPGPU.

--
Srdja

[jdart]

16 cores at 8Ghz? If Moore's law kept progressing at its historical rate, we would have much more than that in a decade. But there is some evidence now that CPUs are hitting a limit. Intel for example has had multiple delays bringing 10nm chips to market.

--Jon

[Dann Corbit]

16 cores at 8Ghz? If Moore's law kept progressing at its historical rate, we would have much more than that in a decade. But there is some evidence now that CPUs are hitting a limit. Intel for example has had multiple delays bringing 10nm chips to market.

--Jon

True, but Taiwan has 7 done and is working on 5.

On the other hand, at some point silicon will run out of steam.
That is when a new technology will be invented, like it always has in the past.
History shows for calculation, there are lots of new paradigms:
A. Simple mechanical devices (with gears like Pascal's device or with nodes like the abacus)
A.1 More complicated mechanical devices (proposed) like Babbage's machine with Ada's language but never actually implemented until much later
B. Relays (Hollerith machines)
C. Vacuum tubes
D. Transistors
E. Integrated circuits
F. Parallel integrated circuits
G. Something will be next...

[Joost Buijs]

Intel's 10nm process has approximately the same feature size as TSMC's 7nm, and TSMC has problems with yield at 7nm too. It's all marketing talk because the actual feature sizes are more like 40nm. Both companies are still using UV lithography (193nm) and its already quite a feat that they manage to create such small features with it.

Since AI and ML are booming at the moment its my guess that in the future we will see processors with lots of 'tensor cores' included. Another development could be inclusion of FPGA. https://www.anandtech.com/show/12773/in ... to-vendors