Skybuck's Parallel Static Huffman Decoding Algorithm

Hello,

Today/Tonight I had an idea how to fund my Algorithm skills ;) :)

It's always funny to see how some programmers believe that Huffman cannot be decoded in parallel.

I believe it is possible to decode (static) huffman in parallel.

However as I wrote in the past I want a million dollars for it ! ;)

Perhaps nobody is going to pay 1 million dollars all by him/herself so instead I am going to give crowd funding a try ! ;)

I just created a project for it and here you can find it and donate if you are curious:

formatting link

Tomorrow I will also add a new project for my Cuda/Delphi Framework so that hopefully a million Delphi programmers can start using Cuda with ease ! ;)

The Cuda/Delphi framework is super real.

The algorithm is inside my head and in a draft so it will need further work and ofcourse implementations will be made ! ;)

Read the project description for more information about the algorithm ! ;)

Do tell everybody about it too ! ;)

Bye, Skybuck.

Reply to
Skybuck Flying
Loading thread data ...

..and you are going to do the programming for that on your super-reliable (NOT) raft of dead computers?

Reply to
Robert Baer

" ..and you are going to do the programming for that on your super-reliable (NOT) raft of dead computers? "

Yes, my latest configuration seems to work somewhat reliably, thanks to automatic shutdown in case of overheat scenerios.

The drivers or windows 7 do seem to have same boot problems. But I put the system to sleep, which prevents a long boot time.

I'm thinking of buying a new computer but I see some issues:

  1. The haswell processor runs on a slightly slower clock frequency ?!?

  1. The haswell processor has an embedded gpu, eating away transister space that could have been allocated to more L1, L2, L3 cache and/or additional cores.

I am likely to buy an additional graphics card from NVIDIA because I do like CUDA somewhat.

  1. NVIDIA's graphics cards run really hot and there heatsinks with the fans on it seem to get stuffed fast... even GT 520 with passive heatsink might still overheat system.

  1. Though CPU heatsink also gets full of dust.

So far some solutions have been to put the system in auto-shutdown-when-idle to prevent dust from being sucked in unnecessarily... and keeping appartment a bit more clean ;) :)

Therefore my desire for a dustless computer is high on my list as I already mentioned a while ago.

I do expect such systems to become available in the year 2016... at least I hope so ;)

My AMD X2 3800+ combined with the GT520 is starting to show it's age... sometimes websites open a bit slowly. I think it's CPU related, bandwidth with 6 MB/sec should be adequate.

Some websites are just fokked with lots of ads everywhere... but still ;)

  1. AMD's octo-core processor has only 16 KB of level1 cache per core, a very big disappointment. This will make individual cores perform very poorly such is my expectation.

  1. NVIDIA's volta will probably get delayed and might not arrive in the year

2016, but 2017... also NVIDIA might introduce a CPU+GPU design... not sure if and when that will come out ;)

I think new technology/new fabrication plants will be necessary to take computing to the next level. I suspect Intel has little competition right now at least when it comes to top-performing-cpu. However the trend seems to be towards energy efficienty for obvious reasons ;) Now intel battling it out with ARM at mobile front ;)

I am very curious to what will happen when energy/heat limit are reached at mobile front... maybe maybe front innovations will lead to innovations on desktop/server front.

So hopefully mobile will have a good effect on desktop in future.

Personally I do expect that as long as new technology is not invested in, that existing chip manufacturers will try to put chips into everything, watches, glasses, fridges, everything ;) :)

A last despirate attempt to produce and sell chips ;)

I am also not convinced that CUDA/Parallel processors are the way of the future... perhaps they simply run to hot... and are still work-inefficient... time will tell ;)

Bye, Skybuck :)

Reply to
Skybuck Flying

Have you considered solving the problem at the input end? Feed the computer an adequate supply of dust-free air? If you filter the air and then use it to pressurize the PC chassis (e.g. blow it in through a vent opening in the side), and maintain positive pressure in the case (blowing in air faster than the exhaust fans remove the warmed air) then there will be no path for dust to enter the case.

Replaceable air filters are inexpensive. You could select any of a whole bunch of types - an automobile air-filter cartridge, or a home forced-air furnace filter would be a good choice. Even a standard filter will catch a lot of dust; a HEPA filter would be even better; a standard filter (to pre-filter the coarse stuff) followed by a more-expensive HEPA filter would be best yet. Some furnace filters are "washable" - take 'em out, vaccum 'em off, and then rinse under running water to remove most of the remaining dust.

Build the filter chassis out of cardboard or wood. To get adequate airflow you'd probably want to use a large-diameter muffin fan... start with a big one, and run it on reduced voltage to cut down on the noise level if necessary.

Reply to
David Platt

I have also been fighting this problem for a long time. I tried sealing the case with caulk, but the fans sucked in air from the front panel where the CDROM burner sits. You can't seal that area.

I found the ideal solution is to upgrade to a Sandy Bridge I3 2120 cpu and place the entire computer in a 121 litre clear polyethylene garbage bag from Walmart. I used a woman's hair clip to seal the end of the garbage bag around the power cord and other wires coming out the back end. The clip looks something like this but is much shorter:

"

formatting link
"

I find these clips are great for collecting bundles of wires and cables that clutter up your desk and workspace. In this case the warm air exausts from the power supply, works its way down past the cool cables, and re-enters the case at the bottom where all the adapter blanks have been removed. I leave the case side panel off for additional cooling.

The new cpu runs about 230% faster than the old one, so the computer loads programs almost instantly and runs very fast. The cpu is idling most of the time so the power consumption is very low. The power supply and case are barely warm to the touch. I did an OCR on a 50 megabyte pdf file which kept the cpu 100% utilized for a long time, maybe half an hour. I couldn't detect any heat rise when it was finished. Hopefully my next motherboard will have a working interface to lm-sensors so I can actually measure the temperature when running.

JK

Reply to
John K

Wrapping a PC in a plastic bag somehow sound dangerous to me.

I can vaguely remember that air blowing against plastic could cause static electricity build up.

Or perhaps tiny little particles flying around.

If ya think about it it makes sense...

Tiny little electrons picked up along the way ;)

Bye, skybuck.

Reply to
Skybuck Flying

No problems. The bag is in contact with the case which is grounded via the third prong on the power plug. No tiny particles are generated by the fans or anything else.

The only problem is heat dissipation. This is a problem with pre-Sandy Bridge cpus, but the Intel Core i3 2120 cpu has very low power dissipation when idling. It requires a LGA 1155 socket.

However, most LGA 1155 motherboards also include the UEFI bios which I believe is used by Microsoft to try to kill Linux and to remove XP from service:

formatting link

There really is no need for UEFI on Linux. If a virus tried to write in the boot area, it would destroy grub, the boot manager. You then have to reinstall Linux.

This gives a pretty clear and fail-safe detection of boot viruses.

JK

Reply to
John K

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.