nie celkomV Samsungu mame 8 jadier s vykon 1- 30% =0,7 naosobku jadra v Aceri (test je jednojadrovy)
tym padom teoretikcy mame vykon
8*0,7=5,6 jadra, takych ako su v AceriV Aceri mame 4 jadraTeda cisto technicky Samsung je5,6/4=1,35 kart rychlejsi ako Acer.Ak fungovalo vykonnejsie jadro v Samsungu (ma 4 ks vykonnejsi a 4 ks menej vykonnych jadier) tak je to menej krat, ak vykonnejsdie je to viac krat.Lenze na toto potrebujeme paralene programovanie a to vie ani nie 1% program,atorovNov 27 2007, 12:28INTEL RECKONS barely one per cent of software programmers are prepared to face the challenge of parallel programming, which the hardware giant (unsurprisingly) reckons is the future of development.
http://www.theinquirer.net/inquirer/news/1026585/pro... ...
a preto takych programov je malo - tiez okolo 1 %Software needs meaty cores, not thin, stringy ARMs, says Intel
26 Feb 2014
“The world has a big issue around vectorisation and parallelisation of code,” Graylish said. “99% of code isn't written that way.” Graylish also feels “defining a workload that can run in 1000 cores is hard.”
http://www.theregister.co.uk/2014/02/26/software_needs... ...
Ak nie je program napisany paralenen tak bude Samsung pomalsi, ak je napsiany paralene, bude rychelsji ako Acer.
Vyhodou je,ze jedna z osobnosti vyvoja jadar Linux (a teda aj ChormeOS) o tom pise knihu
"Is parallel programming hard?", first edition
[Posted March 11, 2014 by corbet]Paul McKenney has announced that the first edition of his 500-page book Is Parallel Programming Hard, And, If So, What Can You Do About It? is available in electronic form; a printed version will follow soon. The entire book is available under the CC-BY-SA 3.0 license.
https://lwn.net/Articles/590231/...