Just finished and benched my xp retro build. I want it silent and efficient so instead of an early and hungry high-end card I decide to take a newer low-to-mid-end card, and I'm happy with the result.

Sys-specs:

Windows XP SP3
Asrock 1333viiiv LGA775 platform
Core2Duo E8400 at 2.99 Ghz; FSB 333
4GB Dual-Channel 4-4-4-12@333Mhz

HP OEM Nvidia 530 GT ; 2GB DDR3 128bit
GPU-Clock 699Mhz, Memory-" 792 Mhz, Shader-" 1399Mhz ; GPU-Voltage 0.99V.
(Card does not exceed 72 C/ 161 F on Furmark while at just 19% fan-speed)

Well, we can see what happens with 2018 card and Ryzen 5 5600x @ 32 GB DDR4 at 4000Mhz
OS Win10 Pro 64bits 22H2

Ryzen 5500 w/32Gb DDR4 3200, Intel ARC A770 16Gb

DXVK will greatly improve FPS but can`t complete all 3dmark2003 Tests.
(DirectX to Vulkan Wrapper)

matze79 wrote on 2024-06-11, 19:50:

Ryzen 5500 w/32Gb DDR4 3200, Intel ARC A770 16Gb […]
Show full quote

Ryzen 5500 w/32Gb DDR4 3200, Intel ARC A770 16Gb

The attachment arc700.jpg is no longer available

DXVK will greatly improve FPS but can`t complete all 3dmark2003 Tests.
(DirectX to Vulkan Wrapper)

The attachment arc770.jpg is no longer available

nice bench matze.
also I was just thinking about doing the exact same thing a few days ago with Vulkan Wrapper, haha!
Ill try.

also i've upgraded my rig to the most can do am4, look:
in general, it gain 5000 points for changing my video card and cpu, interesting but i think is little poor improvement (comparing 2070 vs 4070 ti sup @ r5 5600x vs r7 5800x3d).

in real games the story is another world:
I've gained around 50 to 70 FPS in Metro: Last Light, 20 to 30 FPS in less optimized games (Metro 2033, not the Redux version), and an incredible 100 to 150 FPS more in well-optimized games (Doom 2016).