Bridgman Is No Longer "The AMD Open-Source Guy"

[phoronix]

Phoronix: Bridgman Is No Longer "The AMD Open-Source Guy"

After five years of the open-source AMD strategy, John Bridgman is no longer managing these efforts...

http://www.phoronix.com/vr.php?view=MTE4ODU

[Drago]

Since AMD HSA for Linux will be open-source, Bridgman is still in this realm.

[89c51]

Good luck on the new job bridgman.




Tell the new guy to join the forums so we can holler at him about PM and stuff (j/k)

[Nille_kungen]

First i was a bit worried about AMD cutting open source team after only reading the title, but after reading the article it seams like something great, so best of luck Bridgman.

[bug77]

Wait, AMD still has a heterogenous portfolio?

[blackiwid]

it can be good or bad, I thank bridgman and all that helped him with the driver efforts for the work, I think that in the beginningg there were big advancements, but in the last year nearly nothing happend on the opensource front. but maybe I am blind.

So hopefully it gets new drive with this change. What I would like is a plan what is worked on.

And there should not be the only point "bringing support for newer chips" then I could even understand when it would take longer than prognosed.

But a todo list where stuff stands on todo status for several years is not good.


But again that was shurly not (only) Bridgmans fault, so go on thanks for all positive results you helped with and your communication here.

[bridgman]

but in the last year nearly nothing happend on the opensource front. but maybe I am blind.

Not blind, but a lot of the work was less visible than usual. There were three big contributors to that :

1. Moving to a new shader architecture (GCN), new memory management (GPUVM) and new shader compiler (llvm) at the same time. This was kind-of necessary but it meant that we had far more work in process where you couldn't see an obvious benefit. Using llvm was partly to build a good foundation for an open source OpenCL stack, and partly to get a more capable shader compiler into the graphics stack.

2. Doing speculative development on the two remaining weak spots of the current driver stack -- power management and accelerated video decode. Historically there was so much catch-up work to be done that we focused development efforts in areas where we were pretty sure we would be able to release the resulting code, so there was little risk of wasted effort and (relatively) short delays between writing code and having the results publicly visible.

In these last two areas we knew that the review/revise/repeat process would be a lot longer, but we reached the point where it made sense to dive in anyways. We had to write/debug a lot of code just to get the process going, so there was a bunch of work done there which hasn't shown you any real benefit yet.

3. Making a big push to have launch-time open source support. This is the ultimate "developers do a lot of work but you don't see anything" investment because you don't see the results until closer to hardware launch date, but it is an important step.

It's probably fair to say that more work has been done in the last year than any previous year, but relatively less of it has been immediately visible.

[disi]

Good luck in the new job, hope it had a nice raise in payments

[bridgman]

Thanks. Pay and title don't change, but the big deal for me was that we hired some good new people to take over other parts of my job unrelated to open source, so now I get a lot closer to being full-time on Linux development work.

[LinuxID10T]

Just wanted to let you know Bridgman, thanks for all the work. I know you get a lot of flak around here, but I'm glad you guys at AMD work so hard for the cause of open source.