Hitman 3 PC Analysis : PS5/Series X Comparisons

7SMkbeu.jpg
 
So the 83.33% is bullshit?
It's not. Supersampling option in game usually allows increasing rendering resolution at fixed steps. At 1080p, the in-game supersampling setting at 2.00 renders the game at 4k. The 2.00 donates the multiplier applied at both horizontal and vertical axis (1920x2=3840). Obviously, the in game option does not allow you to decrease the rendering resolution, so he selected game resolution of 3840x2160 and applied a custom supersampling setting of 83.33% which yields...1800p.
 
Last edited:
It's not. Supersampling option in game usually allows increasing rendering resolution at fixed steps. At 1080p, the in-game supersampling setting at 2.00 renders the game at 4k. The 2.00 donates the multiplier applied at both horizontal and vertical axis (1920x2=3840). Obviously, the in game option does not allow you to decrease the rendering resolution, so he selected game resolution of 3840x2160 and applied a custom supersampling setting of 83.33% which yields...1800p.

Ok I see. The 83.33% is just specific to this games multiplier and not like other engines internal resolution scalers that use incremental percentages of the the games set resolution.
 
Last edited:
Interesting...

Back in the summer, after the spec reveals, a lot of people compared these consoles' GPU to a 5700 XT. I guess it's basically a 5700 XT+, with the + being higher clock and extra RDNA 2 features? I would have liked to see both consoles at least be on par with a 2070 Super, but I guess that could vary game to game as well (didn't PS5 beat a 2080 in Valhalla?).
If I remember it was Digital Foundry and also VFXVeteran VFXVeteran which...wasn't really pretty.

Also Valhalla prefers AMD card heavily.
 
Ok I see. The 83.33% is just specific to this games multiplier and not like other engines internal resolution scalers that use incremental percentages of the the games set resolution.
Yes, exactly. The way the game denotes the supersampling level is different to some other games. But in any case Alex's methodology is correct.
 
AC you should point that you are talking about the minimums on dynamic resolution the game is not 1440p on PS5 nor 1080p on XSX. The other user asked for constant which are not the case in your example.
Completely wrong, as that's the lower bounds for like seconds, 95% of the time the games you mentioned are running at higher resolutions. Also Dirt 5 now has all the settings restored on 120hz and a higher framerate by up to 30fps on Series X
Valhalla drops down to low fifties with tearing on PS5, doesn't happen on Series X
Hitman 3 does drop frames on PS5 on the Miami level, they just showed you that on this very video🤣 even at 1800p.

So again when you have a PS5 game with a CONSTANT 44% resolution advantage and higher settings than Series X get back to me, I'll wait.

Lower bound / constant: That's because AC used a dynamic resolution scaler which Hitman devs do not have. It's quite logical and something that shouldn't even have to be explained.

Considering the high-frame rates on PS5 even in stressed areas, it is easy to see that a dynamic res scaler would make the end-result the same as Xbox Series X (which can also benefit from DRS in scenarios where it drops frames)

Riky Riky
Dirt 5: False. Dirt 5 still uses lower settings on XSX. And none of the problems were fixed in the 60 FPS mode. You can check NXGamer's video for more details.
Valhalla: Series X doesn't drop frames because it lowers to the resolution to 1080p, while PS5 maintains 1440p. How is it ANY different than Hitman 3? Only the roles are reversed here for consoles.
 
Lower bound / constant: That's because AC used a dynamic resolution scaler which Hitman devs do not have. It's quite logical and something that shouldn't even have to be explained.

Considering the high-frame rates on PS5 even in stressed areas, it is easy to see that a dynamic res scaler would make the end-result the same as Xbox Series X (which can also benefit from DRS in scenarios where it drops frames)

Riky Riky
Dirt 5: False. Dirt 5 still uses lower settings on XSX. And none of the problems were fixed in the 60 FPS mode. You can check NXGamer's video for more details.
Valhalla: Series X doesn't drop frames because it lowers to the resolution to 1080p, while PS5 maintains 1440p. How is it ANY different than Hitman 3? Only the roles are reversed here for consoles.

You don't understand constant? The Series X version of Hitman 3 is ALWAYS 44% higher resolution with higher shadows than the PS5 version. Valhalla is actually lowest 1188p but even those breakdowns show that in certain cases it can run higher on Series X than PS5, it's completely variable. Most of the time they would be practically the same.
 
You don't understand constant? The Series X version of Hitman 3 is ALWAYS 44% higher resolution with higher shadows than the PS5 version. Valhalla is actually lowest 1188p but even those breakdowns show that in certain cases it can run higher on Series X than PS5, it's completely variable. Most of the time they would be practically the same.
Yes 44% higher res, atleast I am not is denying that. The shadows quality is also better on XsX.

This game looks like it is rasterization heavy, this is where we have seen and can expect PS5 beat XsX everytime.

I think what Heisenberg007 Heisenberg007 is saying, is that with dynamic res we would know, what PS5 would be able to do on this engine.

But for me that is pure speculation because the Devs decided for the locked lower res on PS5.

Clear win for XsX as it is, if that is important.
 
Yes 44% higher res, atleast I am not is denying that. The shadows quality is also better on XsX.

This game looks like it is rasterization heavy, this is where we have seen and can expect PS5 beat XsX everytime.

I think what Heisenberg007 Heisenberg007 is saying, is that with dynamic res we would know, what PS5 would be able to do on this engine.

But for me that is pure speculation because the Devs decided for the locked lower res on PS5.

Clear win for XsX as it is, if that is important.
ONLY 1 FPS
nw8xA3O.jpg


 
I don't believe this game is well-optimized for any console. If we believe it is, why does PS4 Pro outperform Xbox Series S? It doesn't make any sense.

PS5 should also be performing much better than it did in this game. But it can't even reach 2060 Super? That doesn't sound right either.
But you thought that all of the other games that favoured the ps5 were though right?!

Seeing your input in those face off threads makes me think so.
 
Uh, no. 3200x1800 is 69.4% of 3840x2160. Do the math.
You clearly don't seem to understand how it works.

I dont understand how he got 83.33%, 1% of the pixels of 3840 x 2160 is 82,944. (3200 x 1800)/82944 = 69.444% What am I doing wrong? How do you end up with 83.33%?
You can see his response above.

The only way to get 83.33% is to add the horizontal and vertical resolutions then calculate the difference. Basically butchering maths to try and spin a narrative.
Unless you want to believe pixel counts are just horizontal+vertical resolution, yes.
What is this BS? See his tweet above.
 
Last edited:
It's funny they use PS5 content... I through XSX is the best version with "glorious" 4k? But what about XSX for the framerate? Hurting question?
 
It's a major win for XBOX and IMO first REAL difference that makes one version significantly better. The reasons might be multiple but don't forget that this is the difference that most expected by looking at consoles specs. Still it is mostly last-gen game so not much of a deal. But don't forget the PS3 situations - when tools matured and devs got some experiance the games started to look amazing.

I still think PS5 and XBOX have a lot of power in them that has not been yet unleashed.

Now it's getting interesting - I am curious in performance in Resident Evil 8 on XBOX and PS5. That's gonna be next big title that will push the consoles... The question is how RE:Engine is capable of use more unusual PS5 specs - I think Capcom and Sony are working close together (Maiden demo) and hope for some optimization.. If that's not the case I think we will se better results on XBOX.
 
So ps5 is a little better than 5700, with raytracing capabilities? I'd like to see comparisons with the series X GDK, as that was the better console version. I don't see why some people don't like DF? Series X is starting to come out the womb like I claimed a while back.
The 5700 XT, not 5700, is a better comparison point, since Alex was saying that in actual gameplay, the PS5 was doing better than a 2060 Super (60 FPS vs 55 FPS).
 
I thought it was a great video that made some really specific points.

He specifically started he got the exact PC equivalent console settings from the developer themselves instead of having to guess them by eye, so where he saw unexplained differences he speculated these would be, in the instance he highlighted, memory bandwidth related. I think this was his point in relation to XSX performing relatively better to the PC equivalent than PS5... Because Xbox has better memory bandwidth possibly?

Memory bandwidth seems to matter more the closer to 4k you try to render at. I don't think you will see advantages at all between PS5 and XSX at 1440p because memory bandwidth does not matter at that resolution and this maybe is why PS5 is seeming to be winning, although close, in these early games as its faster clocks may make a difference.

If devs start putting out 4k or near 4k native games regularly then XSX will pull well ahead due to power and bandwidth advantages.

Worth noting also, he shows that neither PS5 or XSX use VRS in this game. PC does however, using Tier 1, and in this game VRS gives 3% frames advantage on Quality mode. He doesn't say what the increase is in Performance mode as to him it degrades the image too much.

VRS tier 2 I am assuming, because it is hardware based, will offer even better performance increase than 3%. Certainly looks like it in Gears 5 Hivebusters.

Important to note that VRS does increase performance and it will be a miss if PS5 does not have it.
 
Worth noting also, he shows that neither PS5 or XSX use VRS in this game. PC does however, using Tier 1, and in this game VRS gives 3% frames advantage on Quality mode. He doesn't say what the increase is in Performance mode as to him it degrades the image too much.

VRS tier 2 I am assuming, because it is hardware based, will offer even better performance increase than 3%. Certainly looks like it in Gears 5 Hivebusters.

Important to note that VRS does increase performance and it will be a miss if PS5 does not have it.

Full breakdown here:
The team saw similarly large perf gains from VRS Tier 2 – up to 14%! – this time with no noticeable visual impact.
 
As Alex points out, this game and Watch Dogs are the only instances where we have precise settings for consoles directly from the developer. Theres no guess work involved nor subjective involvement for one vendor or another. These 2 games paint a vanilla 2070 ballpark performance for the ps5. And the settings used are low, medium and turned off other than textures. It seems the need for higher resolutions is strangling the performance of these machines right from the get go.
 
Hitman 3 is my most anticipated game of the year. I went with the PS5 version despite having a capable gaming rig (2080). I'm starting to wish I got the PC version, but the lack of a steam release and the inclusion of VR on console was just enough to swing in favour of the PS5 version for now. I'll buy again when it releases on Steam.

The game has delivered in content, quality of the port and definitely when it comes to these performance analysis because the reactions have been priceless.

All versions are great, but the pecking order is clearly established.
 
Waitin on them Nintendo Switch Benchmarks.
*and it's the cloud version. Damn. Would have been quite the contrast to see a native Switch version of this.
 
Last edited:
Why are so many in denial about next gen consoles and their PC equivalent gpu's, more specifically, the Sony fanboys? I don't even think Cerny mentioned a comparable gpu, because of this exact reason.
 
In denial about


That is a game that performs poorly on nvidia compared to AMD and which settings on PC were matched by using guesswork with the naked eye. Thats not comparable to the situation at hand, where the game has the exact console settings provided by the developer. Same with Watch Dogs Legion. Those 2 dont paing a picture where the PS5 is at 2080 levels
 
One game is better optimized for one engine then the other. We will constantly see differences between these consoles, and no xbox fanboys, the tables are not turned based on 1 game. Otherwise the tables were already turn multiple times in favor off the PS5.

The PS5 is still overall the better performing console in most games right now (even this game if that res different doesn't do anything at the end for most people).
 
Last edited:
Strange results from PS5 l don't get it another DF video (Assassins Creed) had the PS5 delivering near 2080 performance.
 
Last edited:
In denial about
Yeah, in an AMD sponsored title. What GPU would you compare it to on AMD side of things? And comparing that gpu to Nvidia's gpu's, in a plethora of benchmarks, you'll see why Hitman comparison gives more proof to what he been said all along. Ps5 isn't some mystical gpu that can surpass it's physical core count, clockspeed, vram amount. We gotta go with reality instead of wishing upon a star.
 
The Watch Dogs: Legion results were using ray tracing where AMD clearly falls behind Nvidia. And for the Hitman gameplay comparison, framerates were capped at 60 FPS. So we still don't have a rasterization comparison at confirmed identical settings with uncapped framerates.
 
Yeah, in an AMD sponsored title. What GPU would you compare it to on AMD side of things? And comparing that gpu to Nvidia's gpu's, in a plethora of benchmarks, you'll see why Hitman comparison gives more proof to what he been said all along. Ps5 isn't some mystical gpu that can surpass it's physical core count, clockspeed, vram amount. We gotta go with reality instead of wishing upon a star.

This was the denial I was speaking about.
 
This was the denial I was speaking about.
You seem to be stuck in denial. Ps5 doesn't touch the top of the line gpu's like many proclaim. It's like a glorified 5700 with limited raytracing capabilities.. Just about all comparisons have shown this, multiple times already. Or are you denying this as well?
 
As more games come out and this keeps happening it's going to be brutal if the panic over this one game is prologue - endless cries of fake news, claims of biased comparisons stolen by Microsoft and eventually, PSAnon are in our futures #PhilSpencerisalizardoverlord etc.
 
So ps5 is a little better than 5700, with raytracing capabilities? I'd like to see comparisons with the series X GDK, as that was the better console version. I don't see why some people don't like DF? Series X is starting to come out the womb like I claimed a while back.
This is a way oversimplification of what was actually shown in the video. In specifically gpu mem bandwidth the PS5 looks to be between a 5700 and 2060s however, they couldn't determine the overall grunt of the PS5 due to the framerate cap and v-sync. Also what i found interesting is that even though the XsX hold a 44% res performance gam in this game based on the benchmarks in the video and using that mem bandwidth heavy scene, the reality is the PS5 is at 109% of a 5700 and the XsX is only at 123% of a 5700. To me this proves that the mem bandwidth of the XsX is infact more performant than that of the PS5 even with its mem setup however not as large as the raw numbers would indicate.
 
Interesting...

Back in the summer, after the spec reveals, a lot of people compared these consoles' GPU to a 5700 XT. I guess it's basically a 5700 XT+, with the + being higher clock and extra RDNA 2 features? I would have liked to see both consoles at least be on par with a 2070 Super, but I guess that could vary game to game as well (didn't PS5 beat a 2080 in Valhalla?).
I think both consoles are above that of a 5700xt, this was really just a test of mem bandwidth as Alex mentioned. Due to the frame cap at 60 who knows how the compare outside of that mem bandwidth starved dip?
 
This is a way oversimplification of what was actually shown in the video. In specifically gpu mem bandwidth the PS5 looks to be between a 5700 and 2060s however, they couldn't determine the overall grunt of the PS5 due to the framerate cap and v-sync. Also what i found interesting is that even though the XsX hold a 44% res performance gam in this game based on the benchmarks in the video and using that mem bandwidth heavy scene, the reality is the PS5 is at 109% of a 5700 and the XsX is only at 123% of a 5700. To me this proves that the mem bandwidth of the XsX is infact more performant than that of the PS5 even with its mem setup however not as large as the raw numbers would indicate.
So in other words, a glorified 5700, not a 2080s, 3090, or whatever other cards many throw around for the sake of being completely wrong. They won't ever touch the abilities or those cards because of the hardware limitations. This was my point. People need to come back down to reality. Consoles are great price too performance machines, but they won't hold their weight compared to the cards mentioned above. As soon as people can grasp this reality, the sooner these "battles" can be over, as the war has already been decided.
 
So in other words, a glorified 5700, not a 2080s, 3090, or whatever other cards many throw around for the sake of being completely wrong. They won't ever touch the abilities or those cards because of the hardware limitations. This was my point. People need to come back down to reality. Consoles are great price too performance machines, but they won't hold their weight compared to the cards mentioned above. As soon as people can grasp this reality, the sooner these "battles" can be over, as the war has already been decided.
No. In mem bandwidth yes, in other areas no. Thats the thing with custom HW its not as easy as GPU A= GPU B. When you consider things like the Geometry engine and chach scrubbers who knows where it really sits. The unit has been shown to operate as high as a 2080s and as low as a 5700, there are way too many factors at play here. And in my personal opinion the phrasing you are using "glorified" implies negativity and is inciting negative responses towards you. But i think that was your intent anyway.
 
No. In mem bandwidth yes, in other areas no. Thats the thing with custom HW its not as easy as GPU A= GPU B. When you consider things like the Geometry engine and chach scrubbers who knows where it really sits. The unit has been shown to operate as high as a 2080s and as low as a 5700, there are way too many factors at play here. And in my personal opinion the phrasing you are using "glorified" implies negativity and is inciting negative responses towards you. But i think that was your intent anyway.
Glorified isn't a negative term to describe something, and it's no different in this case either. The GPU is "custom" in that sense, but not like the actual custom gpu's like in the ps2 era of time. They are simply modified versions of prexisting tech from AMD.

Just because you have an outlier, which just so happens to be an AMD sponsored title, doesn't mean it's an equivalent to a 2080s. Anyone should be able to come to that conclusion from seeing PC benchmarks between AMD and Nvidia on a plethora of games. It translates to consoles as well, which has been obvious with all the comparisons.
 
Glorified isn't a negative term to describe something, and it's no different in this case either. The GPU is "custom" in that sense, but not like the actual custom gpu's like in the ps2 era of time. They are simply modified versions of prexisting tech from AMD.

Just because you have an outlier, which just so happens to be an AMD sponsored title, doesn't mean it's an equivalent to a 2080s. Anyone should be able to come to that conclusion from seeing PC benchmarks between AMD and Nvidia on a plethora of games. It translates to consoles as well, which has been obvious with all the comparisons.
Ok then a 2080s is a glorified PS5. Nice logic
 
You seem to be stuck in denial. Ps5 doesn't touch the top of the line gpu's like many proclaim. It's like a glorified 5700 with limited raytracing capabilities.. Just about all comparisons have shown this, multiple times already. Or are you denying this as well?
5700 XT, not 5700. (And we still don't have any real rasterization benchmarks.)
 
I applied the same logic you used to call the PS5 a glorified 5700. And it is not negative as you stated so what gives? Can you elaborate?
Ps5 has very similar performance of a 5700 gpu. It's no where remotely close to a 2080s, which most people have realized by now. Calling a 2080s a glorified ps5, is a huge disservice. It has way better bandwith, raytracing abilities, cores, etc. It's not even in the same league. It's as if you'd believe a Bugatti is a glorified ford pinto, which is similar to your comparison.
 
Uh, Alex made a huge mistake by setting the PC resolution to 83.3% of 4K? That'd be 6.8m pixels, while 1800p is actually 5.7m pixels.

The correct scaling would be 69.4% of 4K. Huge blunder by Alex.

I dont understand how he got 83.33%, 1% of the pixels of 3840 x 2160 is 82,944. (3200 x 1800)/82944 = 69.444% What am I doing wrong? How do you end up with 83.33%?
Did you forget which user was logged into which screen? Lol
 
You don't understand constant? The Series X version of Hitman 3 is ALWAYS 44% higher resolution with higher shadows than the PS5 version. Valhalla is actually lowest 1188p but even those breakdowns show that in certain cases it can run higher on Series X than PS5, it's completely variable. Most of the time they would be practically the same.
1440p is 77% more pixels than 1080p.
 
Ps5 has very similar performance of a 5700 gpu. It's no where remotely close to a 2080s, which most people have realized by now. Calling a 2080s a glorified ps5, is a huge disservice. It has way better bandwith, raytracing abilities, cores, etc. It's not even in the same league. It's as if you'd believe a Bugatti is a glorified ford pinto, which is similar to your comparison.
Its in the same league as the 2080s as the 5700 is in the same league as the PS5. When the PS5 holds those same advantages over 5700 that the 2080s does over the PS5 how is it not a "disservice" to the PS5 to call it a "glorified" 5700? Watching you struggle in this logic loop you created is quite entertaining.
 
Top Bottom