Linux Graphics, RX 480 & Other July Excitement; Did You Notice Something Different?

[bridgman]

Why shall I support you with "not using AdBlocker on your page" or donation? - I mean you don't even use free Javascript!

Do you read the articles or use the forums ?

[nf24]

Why shall I support you with "not using AdBlocker on your page" or donation? - I mean you don't even use free Javascript!

[microcode]

For what it's worth, I think that the volume of articles in July has been just fine. If this is a more sustainable pace for you, I'm in favour of maintaining it (and I'm sure Fatima is as well). That said, I'm sure you need the money anyhow, I might make an additional donation on top of the existing premium subscription later today.

[michael-vb]

Mentioned this before, but a reminder just in case. I am deliberately using PrivacyBadger instead of an ad-blocker, but still get a message on Phoronix asking me to turn off my ad-blocker. Perhaps you should look into this, and see if you really are loosing money on visitors using PrivacyBadger who don't actually want that. If there are more than a few then fixing that might be a low-hanging fruit for a bit of extra revenue.

[sdack]

I've been working on such tests and will have them out in a day or two.

From what I've gathered has AMD been using these "dumb" buffers for a while in their driver. I am curious what the deal with these is (possibly less stress on CPU caches because of the combined, write-through behaviour) and why Nvidia waited this long to use them.

[Michael]

I haven't seen anything insightful about the Nvidia driver 367.35 performance anywhere so far. Everyone has just been parroting the release info by Nvidia without actually looking into it, including Phoronix. The change to buffer writes within the driver, which should have been a "low hanging fruit" for every blogger, seems to have fallen right through the centre of this summer hole.

I've been working on such tests and will have them out in a day or two.

[cj.wijtmans]

At least for me the lastest drivers are buggy. Sometimes it stops performing well and I get just 10-25 fps maximum at CS:GO where it normally is 90-120. The same happens with Kicad 3D view so it's not game specific. Closing the 3D view window and reopening it fixes it sometimes (may need multiple tries).

It's too bad that RPMFusion doesn't provide a slightly older driver.

I noticed this as well, i thought it was the game(ns2) maybe not.

[Tomin]

I haven't seen anything insightful about the Nvidia driver 367.35 performance anywhere so far. Everyone has just been parroting the release info by Nvidia without actually looking into it, including Phoronix. The change to buffer writes within the driver, which should have been a "low hanging fruit" for every blogger, seems to have fallen right through the centre of this summer hole.

At least for me the lastest drivers are buggy. Sometimes it stops performing well and I get just 10-25 fps maximum at CS:GO where it normally is 90-120. The same happens with Kicad 3D view so it's not game specific. Closing the 3D view window and reopening it fixes it sometimes (may need multiple tries).

It's too bad that RPMFusion doesn't provide a slightly older driver.

[Mike Frett]

Don't worry about it, you are doing a good job. You should eventually set some time aside for a vacation so you don't get burnt out.

[sdack]

I haven't seen anything insightful about the Nvidia driver 367.35 performance anywhere so far. Everyone has just been parroting the release info by Nvidia without actually looking into it, including Phoronix. The change to buffer writes within the driver, which should have been a "low hanging fruit" for every blogger, seems to have fallen right through the centre of this summer hole.