Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Hasturtium

Pages: 1 [2] 3 4 ... 9
16
Talos II / Re: Talos II reboots itself
« on: December 02, 2023, 06:09:33 pm »
Looks like it might be tied to this, though I couldn’t say what’s triggering it… In any case it looks like updating the firmware to 2.18+ could mitigate the issue. I'm running a Blackbird, so I can't really provide further insight on this one. But I hope it helps.

17
Operating Systems and Porting / Re: [NEWS] Fedora 39 is here!
« on: November 10, 2023, 04:50:07 pm »
I might be a little delayed this time around since I'm working about 400 miles from home at the moment. Strange times, new day job, but gotta pay the bills.

Godspeed - I've been through a number of job hops in the past decade. Here's hoping things settle into a stable rhythm for you soon.

18
Operating Systems and Porting / Re: [NEWS] Fedora 39 is here!
« on: November 10, 2023, 08:36:48 am »
I’d wait for the honorable ClassicHasClass to render his verdict over on Talospace before making the jump, as I did with 38.

19
Applications and Porting / Re: S.T.A.L.K.E.R. / openXray
« on: October 28, 2023, 09:08:16 am »
Sorry I didn't get back to you on this, I realized later than I should have that openXray apparently doesn't support the original STALKER: Shadow of Chernobyl yet... so even if it compiled successfully it probably wouldn't run for me at this point. What a weird release.

20
General OpenPOWER Discussion / Re: POWER11 on the horizon?
« on: October 24, 2023, 03:31:18 pm »
If anything I'm most interested in is IBM returning to its core open source design, so that it has a Power 11 processor that's exactly as open as a Power 8 or Power 9, I'm not going to buy Power CPUs. that they are not IBM because their technology is the best ever for their processors and I don't think that any other company, no matter how good, can reach their level and therefore I don't intend to change companies for these microprocessors. IBM creates it and IBM must produce it or have it produced by one of its direct partners to be the best, this at least is my humble opinion and then obviously it is debatable by anyone.

I mean, if it makes you feel better it's all but certain that the Solid Silicon S1 is a Power10 with open components replacing the previous closed-source bits. There's not another explanation that makes sense within the timeline of their operation.

21
General OpenPOWER Discussion / Re: POWER11 on the horizon?
« on: October 24, 2023, 08:58:52 am »
I'm fine with my eight core Power9. The only other desktop I'm plotting is a baseline Mac Studio for video editing and multimedia use.

22
It looks like the truth is considerably more interesting - per our own ClassicHasClass, this won't be a Power10 CPU, but it will be SMT-4 capable and offer competitive features.

23
So Solid Silicon + Raptor have created a new BMC, and Solid Silicon + Lattice have cooperated to create a Power10 that isn't encumbered by source availability issues? Is that a possibility? If they're changing the CPU socket that means it's definitely not reheated Sforza + Blackbird/Talos II... but I guess time will tell.

24
Applications and Porting / Re: Building Jellyfin for ppc64le?
« on: October 19, 2023, 09:53:10 am »
I seem to recall IBM hosting a bounty on Bountysource for Jellyfin years ago because it was falsely failing some kind of test during compilation. My memory of it is hazy so I've probably gotten the crux of the issue wrong, but if you can manage to dig up that bounty's page, it won't hurt your efforts.

25
GPU Compute / Accelerators / Re: AMD Pro W7xxx series is out folks!
« on: October 09, 2023, 09:29:16 pm »
This?
https://gitlab.freedesktop.org/drm/amd/-/issues/2288

Hey, you found it. Appears it's a non-issue now, which is great news.

26
GPU Compute / Accelerators / Re: AMD Pro W7xxx series is out folks!
« on: October 09, 2023, 09:08:19 pm »
What kind of regressions?
I do have regressions. But these do not interfere further.

He mentioned it over on the dead bird site, some kind of kernel data structure corruption. I don't have the Radeon Pro W5500 I originally used in this machine or I'd toss it in and give it a try.

27
GPU Compute / Accelerators / Re: AMD Pro W7xxx series is out folks!
« on: October 09, 2023, 06:14:13 am »
The XFX Speedster SWFT 210 RX 6600. It’s been perfect since kernel 6.1; I haven’t run into any problems at all.

Tangentially, I wonder if Q66 or anyone else ever dealt with regressions noted in RDNA1/RX 5x00 series hardware as of kernel 5.17...

28
GPU Compute / Accelerators / Re: AMD Pro W7xxx series is out folks!
« on: October 06, 2023, 02:50:30 pm »
Why a W7600 and not a RX7600XT?

I’d guess lower power requirements, better binned silicon, and less likelihood of firmware gremlins. That’s been my experience - it can be a roll of the dice with some consumer cards. I had a Sapphire RX 6600 that flat out wouldn’t run stably in ppc64le and had visual artifacts, but an XFX card’s been running without issue or incident for me since kernel 6.1.

29
GPU Compute / Accelerators / Re: AMD Pro W7xxx series is out folks!
« on: October 05, 2023, 05:49:20 pm »
It’s a step up from the RX 6600, which has been a rock star in ppc64le since kernel 6.1. Until somebody takes a gamble on RNDA3 on the platform I guess this will remain a mystery.

Elsewhere, I wonder when Intel’s xe driver will be mainlined…

30
Applications and Porting / Re: S.T.A.L.K.E.R. / openXray
« on: October 04, 2023, 10:59:33 am »
I’ve had no luck, but the project notes indicate a ceiling of GCC 10 is supported and I’m running Fedora 38, so my woes are probably tied to that. Need to give it another try this weekend.

Pages: 1 [2] 3 4 ... 9