I've been playing Zero-k for a little while now, and I've noticed that when I launch on Windows I get solid performance on High graphics. I have no trouble playing in cluster matches (except for times when it gets real ridiculous >10v10 with lots of tremors etc). Load time is somewhere around 10 seconds for a match.
Yet linux, which is my preferred operating system, is an entirely different story. Load times are often up to and over a minute. People regularly force start a game before my client loads, which ruins a game experience when you can't choose where your commander is. Yet my linux nvme SSD is much faster than my windows SSD and that shows in other write intensive aspects.
Furthermore, the SIM engine seems to be significantly slower on linux. Large cluster matches are unplayable. Any game with 10 or more players (>5v5) is not worth joining because I will lag out and be unable to play. Even with lowest graphics. WITH LOWEST GRAPHICS. This is a fairly modern mid-high end system.
I do not think this discrepancy is due to drivers. I run other games that load as fast in linux (Company of Heroes (1+2), Xonotic, Doom 2016) as in Windows. I get around 10% less FPS in linux than windows for the listed games, but not nearly the order of 1/2 to 1/3 that I see in Zero-K.
So my question is to the devs, where do I start troubleshooting this? The first step seems to me to grab benchmarks that show SIM/DRAW fps in both linux and windows in similar matches.
Systems Specs for the curious:
quote:
System: Host: azimuth Kernel: 4.18.9-zen1-1-zen x86_64 bits: 64 compiler: gcc v: 8.2.1
Desktop: KDE Plasma 5.13.5 Distro: Arch Linux
Machine: Type: Desktop Mobo: ASRock model: X370 Gaming-ITX/ac
UEFI: American Megatrends v: P4.50 date: 02/12/2018
CPU: Topology: 8-Core model: AMD Ryzen 7 1700X bits: 64 type: MT MCP arch: Zen rev: 1 L2 cache: 4096 KiB
flags: lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm bogomips: 108642
Speed: 2706 MHz min/max: N/A Core speeds (MHz): 1: 2706 2: 2718 3: 2719 4: 2721 5: 2721 6: 2725
7: 2722 8: 2723 9: 2721 10: 2723 11: 2720 12: 2960 13: 2718 14: 2722 15: 2716 16: 2722
Graphics: Card-1: Advanced Micro Devices [AMD/ATI] Vega 10 XT [Radeon RX Vega 64] driver: amdgpu v: kernel
bus ID: 28:00.0
Display: x11 server: X.Org 1.20.1 driver: amdgpu unloaded: modesetting resolution: 2560x1440~60Hz
OpenGL: renderer: Radeon RX Vega (VEGA10 DRM 3.26.0 4.18.9-zen1-1-zen LLVM 6.0.1) v: 4.5 Mesa 18.2.0
direct render: Yes
Drives: Local Storage: total: 6.37 TiB used: 1.89 TiB (29.6%)
ID-1: /dev/nvme0n1 vendor: Samsung model: SSD 960 EVO 500GB size: 465.76 GiB
ID-2: /dev/sda vendor: Hitachi model: HUS724030ALE641 size: 2.73 TiB
ID-3: /dev/sdb vendor: Crucial model: CT500MX200SSD1 size: 465.76 GiB
ID-4: /dev/sdc vendor: Hitachi model: HUS724030ALE641 size: 2.73 TiB
Partition: ID-1: / size: 425.46 GiB used: 216.07 GiB (50.8%) fs: ext4 dev: /dev/nvme0n1p2
ID-2: /boot size: 511.0 MiB used: 125.8 MiB (24.6%) fs: vfat dev: /dev/nvme0n1p1
ID-3: swap-1 size: 32.00 GiB used: 256 KiB (0.0%) fs: swap dev: /dev/nvme0n1p3
Info: Processes: 366 Uptime: 5h 41m Memory: 15.60 GiB used: 4.46 GiB (28.6%) Init: systemd Compilers:
gcc: 8.2.1 clang: 6.0.1 Shell: zsh v: 5.6.2 inxi: 3.0.20
|
This is a major block for me and makes zero-k unplayable. 90% of games seem to always just be a large cluster match. I cannot join them because of my system. I know it is a meme that Zero-K is not meant to be played, but unless we can see decent performance fixes, it might really not be for me.
