If you think about no interpolations system then of course CUDA wins…but CPU is millions times faster with interpolation system. For interpolation system you need genius code engine and only that can handle CPU. CUDA can only operate brute force system with very small code without reverse channel and is easy to write down.
The way it looks to me as that each processor type is gradually evolving, while adding features of other types. Some companies have built experimental neural processors. Those will probably take over most work that quantum computers are not suitable for. GPUs will probably go back to being hard wired specifically for fixed purpose pixel programming or each core having only all functions needed for graphics rendering.
Good job, that is what parallel processing is, doing thousands of actions all at once. But what people often forget is the extensive time and cost of the overhead to prepare for that one fascinating shot! Nevertheless parallel processing is still a powerful concept but the software community still have miles of development to make a meaningful use of true parallelism.
This implies that the GPU is superior to the CPU, but in reality, the CPU would be handling much larger tasks, one at a time. They both have their uses. I wanna see the AMD CPU/GPU hybrid.
Is GPU used for gaming?
My CPU’s temperatur stands 73°C on the app in my laptop but my GPU ‘s temperatur was UMA. What does UMA stand for?
Was my GPU not turn on?
A CPU connected to the 2.0 would do just as fast. It is true that GPUs can simultaneously compute multiple functions, but the speed of an individual computation is so fast that nobody would notice it, even on a scale of 1000×1000.
While I agree that GPU computing has massive potential, this video at best is a stunt for high-schoolers to all scream WOW while the camera captures their reaction.
Even commodity CPUs now have 4 hyper-threaded cores, the video emulates an 8-bit single core CPU with a clock rate of 1 Hz at best.
It would have been great if Mr 2.0 was able to pick up the right ink colors out of a pool load them into the parallel pipes (emulating CUDA cores) in order to match a target pattern and fire them on time.
this is absolutely not a good way of portraying the comparison. Why don’t they show where a CPU shines? my 1080 ti is supposedly super fast, but it still lags every few seconds. i7-7700k setup
Se habla de CPU, y GPU, si pero me trae a la mente el símil de SQL, y BIGDATA, parecerá loca mi comparación, pero es así, en SQL, se va armando el conjunto de datos, usando join, unio, y demás , en BIGDATA, es simplemente muéstrame el todo.
Loco pero NO..
Not a fair comparison. The GPU machine is fast, but it is also massive in size. The puny CPU machine is only about 1/20 the size. A true CPU machine would also be able to switch colors with its one nozzle and paint more than just blue.
Amazing blast !! The machine you guys built for this demo incredible !!!
Great resolution on that thing 🙂 I bet it would have been better, but there was a malfunction 1:22
In then the idea of minining came and along with it the most expensive graphics card generation ever.
If you think about no interpolations system then of course CUDA wins…but CPU is millions times faster with interpolation system. For interpolation system you need genius code engine and only that can handle CPU. CUDA can only operate brute force system with very small code without reverse channel and is easy to write down.
That’s why computer only gives one frame every ten seconds. Because it has to count down from ten before rendering the frame
взрослые дядьки а занимаются херней
That was a BLAST
i can do that without a computer ffs
The way it looks to me as that each processor type is gradually evolving, while adding features of other types. Some companies have built experimental neural processors. Those will probably take over most work that quantum computers are not suitable for. GPUs will probably go back to being hard wired specifically for fixed purpose pixel programming or each core having only all functions needed for graphics rendering.
Such a suckass letdown!
1:16
Good job, that is what parallel processing is, doing thousands of actions all at once. But what people often forget is the extensive time and cost of the overhead to prepare for that one fascinating shot! Nevertheless parallel processing is still a powerful concept but the software community still have miles of development to make a meaningful use of true parallelism.
This implies that the GPU is superior to the CPU, but in reality, the CPU would be handling much larger tasks, one at a time. They both have their uses. I wanna see the AMD CPU/GPU hybrid.
At 1:22, one overclocked GPU core broke a tube 😛
how long does it take to reload
This idea blew my mind.
Is GPU used for gaming?
My CPU’s temperatur stands 73°C on the app in my laptop but my GPU ‘s temperatur was UMA. What does UMA stand for?
Was my GPU not turn on?
This actually is a an interesting way to explain how a GPU works as opposed to a CPU
This is only for graphics processing?
Why was this only just recommended to me?
A CPU connected to the 2.0 would do just as fast. It is true that GPUs can simultaneously compute multiple functions, but the speed of an individual computation is so fast that nobody would notice it, even on a scale of 1000×1000.
Really HAHAHAHAHAH
Linus torvalds: “Nvidia, Love you!”
While I agree that GPU computing has massive potential, this video at best is a stunt for high-schoolers to all scream WOW while the camera captures their reaction.
Even commodity CPUs now have 4 hyper-threaded cores, the video emulates an 8-bit single core CPU with a clock rate of 1 Hz at best.
It would have been great if Mr 2.0 was able to pick up the right ink colors out of a pool load them into the parallel pipes (emulating CUDA cores) in order to match a target pattern and fire them on time.
америкосы хуесосы
It’s… just a bunch of tubes pre-loaded with a specific colored paintball and fired at the same time. I don’t get how people find this amazing.
K
Very intuitive demo
like novideo cards didnt explode enough. they created a nuke?
It’s like Homer Simpsons makeup shotgun 😀
this is absolutely not a good way of portraying the comparison. Why don’t they show where a CPU shines? my 1080 ti is supposedly super fast, but it still lags every few seconds. i7-7700k setup
I miss mythbusters.. i love their episodes..
wheres the science ?
this was just engineering..
american cocksuckers
Thank you for sharing in https://www.facebook.com/groups/CGShareNet/
Visit http://www.cgsharenet.com for digital art
Se habla de CPU, y GPU, si pero me trae a la mente el símil de SQL, y BIGDATA, parecerá loca mi comparación, pero es así, en SQL, se va armando el conjunto de datos, usando join, unio, y demás , en BIGDATA, es simplemente muéstrame el todo.
Loco pero NO..
I’m confused??
Anyone got video of the full thing?
yah…that was useless and not at ALL what happens. stupid
2009?!!! Wow I was really living in a cave back then.
1:16 Creampied.
wait so this seems backwards the first one was way for impressive the second one was just gas going through tubes or i just don’t understand
Not a fair comparison. The GPU machine is fast, but it is also massive in size. The puny CPU machine is only about 1/20 the size. A true CPU machine would also be able to switch colors with its one nozzle and paint more than just blue.
will it blend?
1. 지금. 다시보니. 물체가 맞고. 파편’이 튄다… 그것도 보인다는 말이 된다.2. CPU나 GPU 사이에 가속 판넬을 연결할 수 있다.