@xpeedX because you increase the road 2 cars can go over the road so you double the speed. but if you increase the single treat performance you could increase the clock = speed or you could increase the instructions/clock look at it as using a car that has 6 seats instead of 4. and HT is just getting all the cars full instead of leaving one guy whit one car. turbo is like increasing the of the roads when there are a low amount of cars at the road.
Man, I know this was in 2011, but jesus you'd think the camera quality would be better coming from a magazine company. Nonetheless, it's nice to see Amdahl's Law verbally explained like this.
@xpeedX i think you forgot that it is (somthing)+p/n so the n can only make p/n zero and the equation that is left whit like 1000core would be 1/(1-p). so you can only speed up than by making more of the core parallel or increasing the single thread performance. look at it as driving form A to B but you got a speed limit of 120km/h (and you can't go faster) the distance between it is 240km so it takes 2 hours. no matter what you do. but you can get more cars trough the road by laying more roads
I've found a Bitwise method for multiplying 2 numbers that does not require addition steps. My method can be done sequentially 2 bits at a time but it also can be done in parallel with each compute cell working on 2bits of each number due to the fact that in each step of the process you don't require any sequential information passing thus each of the 2 bit processes can be done independently of ever other 2 bit process. Although It seems as if I can achieve multiplication in a bit level parallelisable way I have still yet to achieve the same ability for addition. I suppose in my case the sequential part would be the 3 steps required to perform the multiplication. It would be wired after completion of my work I was to find that I had to use multiplication steps in order to achieve bitwise parallelization of addition functionality.
Hi Douglas, I am too bored to read all that text about this law and verbally it sound simple and easy.....Very nice explanation, thanks for that...and kind of personal comment...you kind of look like "al pacino" and sound like him too...you know that ?
great video!! but i have a question regarding the division of labor.so i have a code of which the 60% can be handled with parallel computing.that 60% is actualy 20% with K proccesors and 40% with N. K + N adds to a total of 5.The division of labor should be K = 3 and N = 2 right?
This happened to me rendering videos! Waaaaaaah :'( At least I know why I only got a three minute gain after adding another processor. Also explains why my cores only run at 30% rendering DVD's. Glad I found this video. Lucky the power isn't wasted tho, Unreal Engine still gives the cores a serious thrashing.
Amazing introduction explained with wisdom and enthusiasm. Thank you!
I like that part. boo-hoo it's doesn't!
saul goodman
so glad someone else said this haha, scrolled down for this alone
omg
Gustafson's law (weak scaling) might be interesting to discuss, as some people see it as more relevant than Amdahl's Law.
Video not particularly focused but it was helpful and you have an excellent accent. ☻
Thank you for the upload.
@xpeedX because you increase the road 2 cars can go over the road so you double the speed. but if you increase the single treat performance you could increase the clock = speed or you could increase the instructions/clock look at it as using a car that has 6 seats instead of 4. and HT is just getting all the cars full instead of leaving one guy whit one car. turbo is like increasing the of the roads when there are a low amount of cars at the road.
Man, I know this was in 2011, but jesus you'd think the camera quality would be better coming from a magazine company.
Nonetheless, it's nice to see Amdahl's Law verbally explained like this.
@xpeedX i think you forgot that it is (somthing)+p/n so the n can only make p/n zero and the equation that is left whit like 1000core would be 1/(1-p). so you can only speed up than by making more of the core parallel or increasing the single thread performance.
look at it as driving form A to B but you got a speed limit of 120km/h (and you can't go faster) the distance between it is 240km so it takes 2 hours. no matter what you do. but you can get more cars trough the road by laying more roads
Watch it in 720p or in 144p.. u get the same quality.. *_DAMN_*
2011, 2021 ARE AWESOME
I've found a Bitwise method for multiplying 2 numbers that does not require addition steps. My method can be done sequentially 2 bits at a time but it also can be done in parallel with each compute cell working on 2bits of each number due to the fact that in each step of the process you don't require any sequential information passing thus each of the 2 bit processes can be done independently of ever other 2 bit process. Although It seems as if I can achieve multiplication in a bit level parallelisable way I have still yet to achieve the same ability for addition. I suppose in my case the sequential part would be the 3 steps required to perform the multiplication. It would be wired after completion of my work I was to find that I had to use multiplication steps in order to achieve bitwise parallelization of addition functionality.
Sry I have to write this.
As a practitioner uploading a 360p video as hd is just fail :D
So simple and understandable. Thanks!
You really remind me of Saul Goodman from Breaking Bad
This comment does not get enough recognition. So true!
8 years later and you are so right, I scrolled down to find any comment about that
Hi Douglas, I am too bored to read all that text about this law and verbally it sound simple and easy.....Very nice explanation, thanks for that...and kind of personal comment...you kind of look like "al pacino" and sound like him too...you know that ?
great video!! but i have a question regarding the division of labor.so i have a code of which the 60% can be handled with parallel computing.that 60% is actualy 20% with K proccesors and 40% with N. K + N adds to a total of 5.The division of labor should be K = 3 and N = 2 right?
Is he Computer Bob Odenkirk?
Amazing explanation thanks so much!
Brilliant video, really helpful, thank you!
omg, that's 10 years ago.
This happened to me rendering videos! Waaaaaaah :'( At least I know why I only got a three minute gain after adding another processor. Also explains why my cores only run at 30% rendering DVD's. Glad I found this video. Lucky the power isn't wasted tho, Unreal Engine still gives the cores a serious thrashing.
nice video 2011 - seen by me at 2022
his voice is like rick from rick and morty and he looks similar
Love the shirt explanation lol
Good stuff thank you professor.
thank you sir.. I got the concept very clearly now... :) It was really helpful...
what is the end of N?
That black shirt makes you look chiseled!
hi, my name is frigider and i AM mathematically challenged.
Awesome video thank you!
Great explanation. Thank you
simple and clear. thank you
He cuts to the brass tacks.
Great video
I guess parallel computing is awesome for chess, right?
Thank you very much, it really helped :-)
the graph seems wrong , cose for 0 processor it is taking 0 time !
The graph is the amount each core speeds the program up not time taken.
I love that shirt!
Turn on English(automatic captions)
yes just turn it off
Thanks
Guy kinda sounds like Jordan Peterson
nice shirt
i skipped the mathematically challenged part :)
So thin I could barely see you
low quality video :(
good download lecture
Thank you!
al pacino, are you kidding me
Amdahl's MAGGOT food now.
Great explanation, thank you so much for this video
Thank you!