The Last of Us Part 1 for PC arrived towards the end of March in a rather desperate state – a good experience only available to those with more powerful hardware, with deep problems facing gamers on more mainstream components. However, progress has been made on a series of patches, and we thought it might be a good time to check out the port to provide an update on the game’s current level of quality. There’s a lot of good news to share here, but also some lingering issues — and we’re concerned they may never be addressed.
But let’s get to the good news first. At launch there were serious texture quality issues on 8GB graphics cards – which is very problematic given how many of them are on the market. Setting high texture quality through the GPU’s VRAM limit leading to massive stuttering issues. The alternative was to fall back to medium or even low quality textures, depending on your resolution – the problem was that the quality level looked more like “ultra low”. In fact, the textures from the PS3 original were more detailed.
I would venture to assume that this issue has already been resolved. Medium textures now look absolutely fine, with only a minimal hit to quality compared to the high preset. Memory management optimizations now mean that 8GB GPU owners can actually use the high texture preset instead if they wish, giving a look very close indeed to the PlayStation 5 version. Naughty Dog deserves credit for this. It demonstrates how compute scaling should work and why we shouldn’t blindly claim that 8GB GPUs are already obsolete.
So how was this achieved? Based on the difference in the average textures, it looks like there are completely different mipmaps showing up when the game is set to medium. It’s like the game had an art pass done to the medium textures to keep the resolution higher and retain more detail – so textures that may not even have existed before. On top of that, there’s also a change to the way texture streaming works to accommodate GPUs with different memory sizes. This is controlled by a new option called “texture flow rate”, which is a brilliant addition.
This option defaults to “normal” on medium and high presets. When set to normal, the texture cache size is smaller, thus freeing up more VRAM, but keeping the texture quality the same in cropped view. However, if the camera is moving quickly, some environmental textures may appear later. At fast or fastest setting, the texture pop is practically gone, but this comes at the expense of higher VRAM usage. If you’re playing the game on an 8GB GPU, I recommend high textures on normal or fast settings, which looks absolutely fine.
In the video embedded on this page, you’ll see that there are repeated improvements in GPU and CPU usage, but make no mistake – The Last of Us Part 1 is still very heavy on PC users, and looking at unlocked performance based on In the PS5’s VRR mode the game’s performance, the console version, is 55 percent faster in a GPU-limited scenario compared to the RTX 2070 Super, which generally matches the PS5 in most other rasterized games.
On the CPU side, Naughty Dog still uses the CPU heavily to do background streaming, which makes life very difficult for users of more modest mainstream CPUs, such as the Ryzen 5 3600, for example. I’d like to see more work done here – like the implementation of DirectStorage – but it’s clear that, at least for now, Naughty Dog’s priorities lie elsewhere
An area where there is good improvement – but at a cost – comes from the game’s shader compilation strategy. In short, to avoid stuttering, The Last of Us Part 1 has a long compilation step at the beginning of the game. At launch it was around 41 minutes on the Ryzen 5 3600. Now it’s 25 minutes for the base game, with an additional four minutes for the Left Behind DLC.
This is good news, except that my game is suggesting that the build step is incomplete – I’m witnessing an obvious shader build stutter that wasn’t there before. We know this is the problem as it goes away on a second playthrough, then comes back again when I enable GPU drivers and shader caches. Hopefully this can be resolved once and for all, as I simply shouldn’t expect to see stuttering issues like this in a major triple-A release from a studio of Naughty Dog’s pedigree.
Other good news? Apparent game load time issues have been improved. At startup, just launching the game from the main menu resulted in almost a minute load on the Ryzen 5 3600 with a 3.5GB/s NVME drive. This load time was excessive in an era where games like Marvel’s Spider-Man on PC moved from menu to game in just six seconds. With the latest patch 1.05 the initial load is now 30 seconds. This also applies to in-game loads – for example, skipping a cutscene and loading the next chapter took almost 27 seconds when shown in our original coverage of the game. The same load is now achieved in just under 12 seconds on the same hardware at the same settings.
The Last of Us Part 1 is in better shape then, but it’s still far from the optimal experience I was hoping for. And other issues still need to be resolved. DLSS and FSR2 manage to affect the quality of the shadow map, which really shouldn’t be happening. And of course, there are still bugs to contend with: in my game, an enemy died and the key needed to progress appeared under his body, meaning I couldn’t pick it up. It’s left feeling like PC users remain second-class citizens, putting up with problems that Naughty Dog would never inflict on its console audience.
Ultimately, though, The Last of Us Part 1 for PC is in better shape than it was at launch. The texture quality is much improved, a good experience is possible on the 8GB GPU and the game runs a bit better overall. However, many problems remain and new ones have been added. If Naughty Dog and Iron Galaxy can further relax CPU limitations for background streaming, further increase GPU performance, and completely remove stuttering when compiling shaders, the port will be in a good position. But at least right now, I still think it needs a lot more work.
Leave a Reply