When using Metal for compute stuff am I going to be limited by the replaceRegion/getBytes calls even on a device with unified mem?
-
-
-
Replying to @warrenm
the two memory copy operations are much slower than my kernel.pic.twitter.com/gU5wMCr932
1 reply 0 retweets 0 likes -
Replying to @tapbot_paul
Yeah, it’s common for I/O to be the bottleneck when streaming resources, especially when the compute workload isn’t massive.
2 replies 0 retweets 0 likes
Replying to @warrenm @tapbot_paul
I would expect shared resource copies to be proportionately faster on integrated GPUs, though.
9:27 AM - 6 Jun 2016
0 replies
0 retweets
0 likes
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.