elle —
Right now in the doNextWorkQueueEntry function, we first check if nextEntry < entryCount and i…
You're right of course. If the hardware resource doing SIMD operation X is busy, the other hypert…
I started following HMH back in December and my coding style has been rapidly mutating towards Ca…
This is very cool. I wish Sony would open Native SDK for Vita for everybody, not just the managed…
You can look at this a lot like the port pressure stuff we were doing with IACA. With HyperThrea…
The point here is that the only possible way to actually have what you're talking about work is _…
In theory, as long as the SIMD unit executes only one hyperthread's stuff at a time it'll be as f…
I was wondering if it is possible to combine HT with SIMD.
They are different types of parallel…
Hey guys.
I've got a bug on the code, after to watch the video "Day 012 - Platform-independent So…
Aidan —
So I recently got access to a PS Vita Dev Kit through the PlayStation First program where they gi…
Makes sense. I guess I'm more curious about there being some OS magic in between releasing semaph…
If there's a performance gain to be had, it's more likely to be in having the job queue take n jo…
Good video.
I liked the "juggling razorblades" analogy.
Yep, it makes sense and I agree it's the better option. I was initially thinking the implementati…
To clarify on the anti-aliasing:
Since we are always drawing from bitmaps which have alpha-blend…
There's really no such thing as non-32-bit-aligned 32-bit value on the stack unless you really go…
What other platforms are you talking about? You have working TLS also on Linux, OSX, Android and …
The point of the queue here is not to maximize the amount of sleeping, but rather minimize the am…
We haven't really gotten to the point where we need to start caring about thread_context. It's r…
I was wondering what the purpose of thread_context is now that multi-threading has added to HH. D…
Such approach is fine. It will work. Just don't forget to increase max count. Semaphore value has…
If I was to implement something like this I'd probably go the route of the WaitForSingleObjectEx …
Oh, you're right. Value will be 0 if no other jobs are being added. I don't know what I was think…
Ok, well I grabbed the source code and looking through it properly I can see there's no bug. I do…
Why do you think semaphore is at 0 at 4th step? Semaphore decreases only when WaitForSingleObject…
Apologies if I'm mistaken, but I just watched episode 126 and spotted what might be a possible ra…
then there is _aligned_malloc which will take care of alignment to arbitrary boundaries for you:
…
_mm_mul_epu32 only multiplies 2 32-bit integers and stores result as 2 64-bit integers. So you wo…
elle —
After debugging my vectorized version for a long time, I thought of a few minor things that might…