The Last of Us Part I Benchmark Test & Performance Analysis

[ThrashZone]

Sorry, This pile of bugs deserves none of the praise mentioned in the article.

Hi,
Yeah it's missing the editors choice award is this oversight or are they making a new one just for this crappy release and pooch is constipated atm

 

[CyberCT]

Okay yeah, everyone has their own preferences.

I've noticed on 4090 DLSS 3.0 looks better then turning on 2.0, maybe it's my eyes or something lol. Of course, you can run both together too if you want. DLAA is a bit tough to run but that can yield some amazing AA.

Luckily it seems like we have slightly better choices this gen than last. Next gen will be like "HAS TECHNOLOGY GONE TOO FAR?" Going to be FULL path tracing

I tried DLSS 3 for the first time on my 4090 with Cyberpunk. I noticed no graphics abnormalities and, while there was a slight input lag addition using the controller, it wasn't much and I really love the high frame rate on the projector, where there is no VRR. I'm very impressed with DLSS 3 so far.

 

[GamerNerves]

Why would I want to use upscaling at 1440p with an $800 GPU already?

DLSS at lower resolutions is not as good as at 4K and frame gen is worse at low frame rates because the difference between each frame is larger so more room for artifacts. Let alone the fact that the main advantage of high FPS is reducing input latency.

Off-topic: Fanboys raving like crazy now about DLSS 3 and whatnot, even if it doesn't look great.
I'd like to add to your comment that more FPS is always more fluidity, but I haven't tested where the limit is when I personally cannot tell the difference anymore. It seems above my 165 Hz screen there is more fluidity to be had, to me at least.

On-topic: I yet need to see if this game looks as good as some say, since Youtube material doesn't cut it, but alas, it is a buggy awkward port, though might not be that badly optimized if medium quality settings look nice.

 

[LupintheIII]

It's "acceptable" because an endless stream of moronic "Real Gamers (tm)" constantly lower their expectations into the gutter whilst simultaneously have all the "don't pre-order self-control" of Steven Seagal at a burger stand... Until PC gamers regrow their spines and start calling out sh*tty console ports for what they are again, PC gamers will get to continue to 'enjoy' a new wave of bad console ports whilst the PC tech sector quietly cheers on 12GB VRAM usage for 900p gaming as a reason for 'needing' a $1200 GPU as the new 'budget gamer' option...

You can get a 6750XT with 12GB of Vram for $370.
The real morons are the one who got a 3070Ti for $900 when 8GB was already mid/low-end in 2016.

 

[Winssy]

Thanks to DLSS, even with 8GB of memory the 3070ti is significantly better than the 6750xt, because FSR looks really bad in most games.

 

[watzupken]

To me, this is simply a bad port. Visually it is not a big uplift from the console version. But the high VRAM usage, slow shader compiling, and crashing are show stoppers. I guess one is better off just buying a PS5 to play this game.

 

[luciferxy]

kinda ironic isn't it ? TLoU was one of the most optimized game for such an underwhelming hardware that ps3 has.
And now over a decade later, it is one of the worst UNoptimized game, even on most current gpus.

 

[masterdeejay]

Where is Intel Arc A770 16Gb?

 

[SyCoREAPER]

The game looks good, there's no denying that but the fact that a 4090 (which I have) is needed for 4K 60+fps is ridiculous and shows a lack of optimization. The least they could have done is implement FSR of DLSS despite those being crutches.

 

[Chrispy_]

Or you can use 10-30 minute of your machine's time to compile that

10-30 minutes? Some people are saying 1-2 hours, including multiple crashes and having to restart the game to continue building the shaders. I'm inclined to believe that because streamers broadcast this happening as they tried to play and that's why this has blown up in the both the independent and social media.

 

[JustBenching]

According to HWunboxeds numbers, a 3080 @ 1440p high basically ties a 6800xt @ 1440p MEDIUM. Also, a 3080 @ 4k high ties the 6800xt @ 4k medium.

This is nvidia fine wine, you can use considerably higher settings on the 3080 for pretty much same performance as the amd alternative. Ouch

 

[Chrispy_]

According to HWunboxeds numbers, a 3080 @ 1440p high basically ties a 6800xt @ 1440p MEDIUM

Why would you try and claim that when it's so easily disprovable?

1. HUB's numbers show the 6800XT's minimums are equal to the 3080's averages at the settings you just stated. Given the stuttering of the 3080, that's doubly important.
2. The 3080Ti is actually closer to the result you claimed, but neither it nor the 3080 10GB results fit your claims very well in the first place.

This is nvidia fine wine, you can use considerably higher settings on the 3080 for pretty much same performance as the amd alternative. Ouch

The AMD alternative is a 6950XT by street price, and the much cheaper 6800XT outperforms the 3080 at higher settings due to the lack of VRAM on the 3080 impacting miminum framerates and causing a lot of unpleasant stutter that is commented on heavily throughout the video. Realistically both cards are EOL now so used prices might be more relevant, and the 6900XT is around the same price on ebay as a 3080 10GB.

Trying to wave a green or red banner around is pointless anyway when the results of different tiers of GPU are so compressed in this game that quite clearly has optimisation issues and performs poorly compared to many other equally or more graphically-impressive titles.

The take of all the various reviews and channels covering this game is that 8GB really isn't enough any more. The fact that a 3080 has 10GB buys it a tiny bit of wiggle room but the writing is already on the wall, as it has been for my 3070 for some time now. I've listed it on ebay, the 6800XT is already in my system and I'm no longer in the market for any GPU with only 8GB, not even for 1080p given how things are going with new releases. At this rate, the only green-team GPU I'll have left will be the Quadro RTX 6000 (downclocked 3090)

 

[JustBenching]

Why would you try and claim that when it's so easily disprovable?

1. HUB's numbers show the 6800XT's minimums are equal to the 3080's averages at the settings you just stated.
2. The 3080Ti is much closer to the result you claimed, unsurprising for a $1199 vs $649 matchup (MSRP's are such a joke now).

View attachment 290013 View attachment 290014

Wait what? At the settings I stated, the 3080 averages 83 fps @ 1440p high and 47 fps @ 4k high. The 6800xt averages 88 fps @ 1440p medium and 50 fps @ 4k medium. The difference in both resolutions is exactly 6%. So the 3080 plays at considerably higher settings with just 6% fps difference. That's nvidia fine wine

Posting a bunch of pictures and a wall of text will not change the numbers man

 

[Chrispy_]

Wait what? At the settings I stated, the 3080 averages 83 fps @ 1440p high and 47 fps @ 4k high. The 6800xt averages 88 fps @ 1440p medium and 50 fps @ 4k medium. The difference in both resolutions is exactly 6%. So the 3080 plays at considerably higher settings with just 6% fps difference. That's nvidia fine wine

Posting a bunch of pictures and a wall of text will not change the numbers man

At 1440p (the only resolution that provides playable framerates here) the 3080Ti is closer at just 5% fps difference.

Why would you play at medium settings on a GPU this expensive anyway? Just turn down the resolution, because if you want dips to 30fps at lower image quality you can just throw the disc back into your PS3, lol.

 

[JustBenching]

At 1440p (the only resolution that provides playable framerates here) the 3080Ti is closer at just 5% fps difference.

Why would you play at medium settings on a GPU this expensive anyway? Just turn down the resolution, because if you want dips to 30fps at lower image quality you can just throw the disc back into your PS3, lol.

You don't play at 30 fps, I agree. What im saying is, a 3080 @ 1440p HIGH is just 4% behind a 6800xt @ 1440p Medium.

This from computerbased at 4k Ultra + DLSS Q, ,which is basically the settings youd play on a 3080

 

[Chrispy_]

You don't play at 30 fps, I agree. What im saying is, a 3080 @ 1440p HIGH is just 4% behind a 6800xt @ 1440p Medium.

This from computerbased at 4k Ultra + DLSS Q, ,which is basically the settings youd play on a 3080

View attachment 290018

and what I'm saying, as a counterpoint to this dumb apples-to-oranges comparison is that a 6800XT @ 1440p medium is just 5% behind a ridiculously powerful and expensive 3080Ti at 1440p high.

Not only are you not comparing like-for-like, the 6800XT straddles two Nvidia cards equally in either direction. You picked the 3080, which is ~20% more expensive than a 6800XT, I picked a 3080Ti which is 50% more expensive than a 6800XT. Either way, the performance numbers are ~5% different which is laughable given that they're not priced comparably, the settings comparison is particularly valid or accurate in this buggy-mess of a game, and all the results of all the GPUs are so compressed that the 6% difference you initially claimed is significant enough to be true for entire tiers of GPU within the same AMD-vs-AMD or Nvidia-vs-Nvidia comparinson anyway, rendering your red-vs-green point even less relevant.

I'll quote myself from earlier:

Trying to wave a green or red banner around is pointless anyway when the results of different tiers of GPU are so compressed in this game that quite clearly has optimisation issues

I don't really like siding with one side or the other, but as 2023 progresses it's getting harder and harder to justify Nvidia's small VRAM decisions of the past, present, and future. I'm replacing green with red for the same reason I upgraded my 2060FE back in the day - it ran out of VRAM in an unacceptable number of games I was playing at the mainstream resolutions and settings I was playing at.

I agree with you that the 3080 is just about surviving this particular title with its dignity intact - something I'd bloody well hope so for such an expensive GPU; It was always a slightly-faster, and much more expensive card than the 6800XT. How much longer it can retain its dignity with 10GB remains to be seen....

 

[JustBenching]

and what I'm saying, as a counterpoint to this dumb apples-to-oranges comparison is that a 6800XT @ 1440p medium is just 5% behind a ridiculously powerful and expensive 3080Ti at 1440p high.

Not only are you not comparing like-for-like, the 6800XT straddles two Nvidia cards equally in either direction. You picked the 3080, which is ~20% more expensive than a 6800XT, I picked a 3080Ti which is 50% more expensive than a 6800XT. Either way, the performance numbers are ~5% different which is laughable given that they're not priced comparably, the settings comparison is particularly valid or accurate in this buggy-mess of a game, and all the results of all the GPUs are so compressed that the 6% difference you initially claimed is significant enough to be true for entire tiers of GPU within the same AMD-vs-AMD or Nvidia-vs-Nvidia comparinson anyway, rendering your red-vs-green point even less relevant.

I'll quote myself from earlier:

I don't really like siding with one side or the other, but as 2023 progresses it's getting harder and harder to justify Nvidia's small VRAM decisions of the past, present, and future. I'm replacing green with red for the same reason I upgraded my 2060FE back in the day - it ran out of VRAM in an unacceptable number of games I was playing at the mainstream resolutions and settings I was playing at.

I agree with you that the 3080 is just about surviving this particular title with its dignity intact - something I'd bloody well hope so for such an expensive GPU; It was always a slightly-faster, and much more expensive card than the 6800XT. How much longer it can retain its dignity with 10GB remains to be seen....

If you wanna criticize either company, you compare prices at MSRP. Period. Market prices are dictated by demand. The 6800xt came out to compete against the 3080, and it had similar MSRP. That's a fact, period. Now whether or not it's much cheaper now - cause it wasn't selling when it was priced against the 3080 is irrelevant. AMD wanted to sell the 6800xt for 3080 money, nobody was buying, they dropped the prices.

So, a 3080 at 1440p high gets very similar performance to a 6800xt at 1440p medium. That's just...a fact?

 

[btk2k2]

If you wanna criticize either company, you compare prices at MSRP. Period. Market prices are dictated by demand. The 6800xt came out to compete against the 3080, and it had similar MSRP. That's a fact, period. Now whether or not it's much cheaper now - cause it wasn't selling when it was priced against the 3080 is irrelevant. AMD wanted to sell the 6800xt for 3080 money, nobody was buying, they dropped the prices.

So, a 3080 at 1440p high gets very similar performance to a 6800xt at 1440p medium. That's just...a fact?

Why would you play at high or medium when the 6800XT can handle Ultra 1440p just fine at 60 FPS?

Also depending on scene the 3080 at 1440p can drop a lot more.

And just look at the 1% lows.

The 6700XT gives a smoother experience than the 3080 at 1440p Ultra.

 

[Imouto]

@fevgatos my dude. You are a fucking cancer for this community. The reason why you aren't permabanned is beyond me.

Unless you are doing some social experiment to see how far being a outrageously lying fanboy can get you. Then hurray for science because you are doing an impressive job.

 

[JustBenching]

Why would you play at high or medium when the 6800XT can handle Ultra 1440p just fine at 60 FPS?

Also depending on scene the 3080 at 1440p can drop a lot more.

And just look at the 1% lows.

The 6700XT gives a smoother experience than the 3080 at 1440p Ultra.

You cant get 60 without drops on either card at ultra. TPU is probably testing a lighter scene,, hwunboxed is testing a heavier scene and both drop drop with averages being just above. I want to play at around 80-90 fps,, what are my options? A 3080 @ 1440p high or a 6800xt @ 1440p medium. That's nvidia fine wine

@fevgatos my dude. You are a fucking cancer for this community. The reason why you aren't permabanned is beyond me.

Unless you are doing some social experiment to see how far being a outrageously lying fanboy can get you. Then hurray for science because you are doing an impressive job.

What did I actually lie about? The numbers im quoting are directly taken from hwunboxeds video. You not liking the results and the nvidia finewine doesn't mean im lying.

 

[Cryio]

5700 XT: 1080p, FSR2 Quality, Ultra, Textures on High (Effect Textures on Medium): pretty much 60 fps locked.

Game is too ridiculously demanding on the CPU part for no reason.

 

[csendesmark]

I received the game for my 7900XT
The shader building took me 107 minutes, then I played the game in two sittin around 10~11 hrs
got only one crash at the last ingame frame before the naughty dog logo and the credits came
I suspect the negative reviews are mostly coming because old drivers and crappy machines.
Also, great game!
Did not see any of this last of us before and it's awesome!

 

[JustBenching]

I received the game for my 7900XT
The shader building took me 107 minutes, then I played the game in two sittin around 10~11 hrs
got only one crash at the last ingame frame before the naughty dog logo and the credits came
I suspect the negative reviews are mostly coming because old drivers and crappy machines.
Also, great game!
Did not see any of this last of us before and it's awesome!

Is it a mistype? You mean 17 minutes? Or do you actually mean 107? LOL that's a lot

 

[csendesmark]

You cant get 60 without drops on either card at ultra. TPU is probably testing a lighter scene,, hwunboxed is testing a heavier scene and both drop drop with averages being just above. I want to play at around 80-90 fps,, what are my options? A 3080 @ 1440p high or a 6800xt @ 1440p medium. That's nvidia fine wine


What did I actually lie about? The numbers im quoting are directly taken from hwunboxeds video. You not liking the results and the nvidia finewine doesn't mean im lying.

The lowest FPS count I've got was when you arrive to Salt Lake City when you arrive and see the long highway with the cars plus the city vista at 54FPS
the rest pretty much had 70-+5, interior places often got 90+
I played 2560×1440 with everything ultra, 3950X with 128GB RAM and a 7900XT, and the game on a 2TB Samsung 980PRO

Is it a mistype? You mean 17 minutes? Or do you actually mean 107? LOL that's a lot

I did write 107 and you can see there are no edits on the post,
Wish it took only 17 minutes

 

[LifeOnMars]

PS5

5700 XT: 1080p, FSR2 Quality, Ultra, Textures on High (Effect Textures on Medium): pretty much 60 fps locked.

Game is too ridiculously demanding on the CPU part for no reason.

Exactly, same on the 5700 XT I had and it run smoothly. 5700x CPU coped with it easily, what CPU do you have?