I believe the ML community will strongly disagree. CUDA is everything
The most surprising aspect of this is that it's a solo paper accepted for a major conference!
There's another case of reverting the codebase to an earlier version just to run it
My guess that it's happening mostly due cache conflicts. With 1024 for a simplified L1 with 32kb you can fit exactly 8 lines of the inner dimmension in the cache, which means that (0,8,0) would have the same cache…
As a side note - loop unrolling has almost no effect, it just allows scalar replacements for latency bound instructions and some other stuff
But it can be 20x
So no custom CUDA kernels, just CuPy? Isn't that a performance issue? (Based on installation notes)
People would probably find out what hardware they use for benchmarks and optimize for that, leading to performance decrease for many othes
One could argue if those can be called "technically knowledgeable"
Was it a university like ETH Zürich, where anyone can get in but there is a limited number of spots from the second year onwards? Then it really made sense to suggest that
Same here. I have taken many Calculus and Algebra classes, SAT Math level 2 and have no clue why these graphing calculators are necessarily...
Are you a software developer? No judging, just curious
And your nickname says everything
Interesting is very subjective term. E.g. some find theory much more interesting than games
2$ for a bag is expensive (in Zurich)?
The name of the language and this title sound so wrong... Never ever name something D Many years ago while being too naïve I created a website d-something and the name was big problem, now I see why
2FA authorization needs JS if you want to use it conveniently, otherwise you would have to refresh all the time
And then classic React enters the building
Microsoft once already failed to promote a mobile OS which was ten times better than Android, why would they be successful with shitty Android fork...
That's why Proof-of-stake is the way to go, you will never see a major coinholder undermining the coin
And what if it is not your fault?
As a tourist I am pro-change, because it can be nightmare visiting Cyrillic countries - every street sign, label, bus stop name etc. just doesn't make any sense and hard to remember if you don't know Cyrillic
Not that ideal when you want a diversity and one of your devices is not from the same company
As men courses, especially in SL and GS, often tend to be more complicated, I don't see any value from these analysis. Just my 2 cents, but a better way to prove a point would be to compare results in lower (some…
Every random() function is kind of deterministic, but still very interesting discovery!
I believe the ML community will strongly disagree. CUDA is everything
The most surprising aspect of this is that it's a solo paper accepted for a major conference!
There's another case of reverting the codebase to an earlier version just to run it
My guess that it's happening mostly due cache conflicts. With 1024 for a simplified L1 with 32kb you can fit exactly 8 lines of the inner dimmension in the cache, which means that (0,8,0) would have the same cache…
As a side note - loop unrolling has almost no effect, it just allows scalar replacements for latency bound instructions and some other stuff
But it can be 20x
So no custom CUDA kernels, just CuPy? Isn't that a performance issue? (Based on installation notes)
People would probably find out what hardware they use for benchmarks and optimize for that, leading to performance decrease for many othes
One could argue if those can be called "technically knowledgeable"
Was it a university like ETH Zürich, where anyone can get in but there is a limited number of spots from the second year onwards? Then it really made sense to suggest that
Same here. I have taken many Calculus and Algebra classes, SAT Math level 2 and have no clue why these graphing calculators are necessarily...
Are you a software developer? No judging, just curious
And your nickname says everything
Interesting is very subjective term. E.g. some find theory much more interesting than games
2$ for a bag is expensive (in Zurich)?
The name of the language and this title sound so wrong... Never ever name something D Many years ago while being too naïve I created a website d-something and the name was big problem, now I see why
2FA authorization needs JS if you want to use it conveniently, otherwise you would have to refresh all the time
And then classic React enters the building
Microsoft once already failed to promote a mobile OS which was ten times better than Android, why would they be successful with shitty Android fork...
That's why Proof-of-stake is the way to go, you will never see a major coinholder undermining the coin
And what if it is not your fault?
As a tourist I am pro-change, because it can be nightmare visiting Cyrillic countries - every street sign, label, bus stop name etc. just doesn't make any sense and hard to remember if you don't know Cyrillic
Not that ideal when you want a diversity and one of your devices is not from the same company
As men courses, especially in SL and GS, often tend to be more complicated, I don't see any value from these analysis. Just my 2 cents, but a better way to prove a point would be to compare results in lower (some…
Every random() function is kind of deterministic, but still very interesting discovery!