L;
:EDDDDEEEEEEG EEEEt
.EEEEEEEEEEDj .tLLt EEEE
EEEEE EEEEEEE EEEEEEEEEEE EEEE
EEEE ;EEEEEEEEEE EEEEEEEEE EE .. LEEEE iEEE; EEEE
EEEE EEE; :EEE. EE EEEEEEEEEL EEEE EEE EEE,
iEEEE EEE EEE EE EEEE EEE EEE EEE EEE
EEEEE EEf EEEEEL EED EE EEE EEE EEED DEE EEE
EEEEG EE E,EE EEt:;;;ttfGEE EEE EE EEEf GEE EEE
EEEE ;EE EE EE EEEEEEEEEEEEE EE. EE DEEE EEE EE.
EEEEEE E EE; EEG EE EEi EE EEE EE: EE
LEEEE EE ,EE EEG EE EEE ,EE ,EEE EEE GE
EEEE EEEEEEE EEE EE EEEE EE EEEE .EEEj
EEEE EE EEE EEEEEEEED EEEEEED GEEEE;.iEEEE ,EEEEEEEEt DGj ,
EEEE .Ei EE EEEEEEEEE LLLGLLL iEEEEEEE DEE L G
EEEE EEEEEE EEE DEE
tEEEE . DEEEEEG
EEEEEEEEEEEE ,,,,,,,
>>11 All the compound data objects we have used so far were constructed ultimately from numbers. In this section we extend the representational capability of our language by introducing the ability to work with arbitrary symbols as data.
How to enjoy OpenCL in 14 steps.
1.buy tons of expensive books about GPGPU programming. They're all outdated but they'll work.
2.buy 1000$ graphics card.
3.Overclock it.
4.Copy algorithms from the internet, slightly modify them.
5.design your own algorithms.
6.create your own OpenCL library of algorithms. Open Source it.
7.Write a project using your library. Open Source it.
8.Wait so alot of people use your OpenCL library.
9....5 years...
10.Graphics cards advance to support newer standards, newer hardware, making your library obsolete.
11.Your 1000$ graphics card is now worth 10$ at best.
12.Your algorithms now rely on legacy features and half of them are deprecated.
13.You cry yourself to sleep and swear off GPGPU programming for good.
14.People write really awesome GPGPU code and you get jealous. Proceed to step 1.