• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

NXG - HZD 60fps Patch

In fact Gamer Nexus made a video on how a GTX 1060 can crush a PS5 (which is just ridiculous, which I know too well since I have both a 1060 and a PS5) and people didn't lose their minds like they are doing know over this new Horizon analysis. 🤷‍♂️

Not joking. A fucking 1060! 😂 Read the comments and you'll see people joking about PS5 Pro targeting a GTX 1070.


Oh I can assure you, people here lost their minds over that video.
 

GymWolf

Gold Member
I found it hard to find a good 24" 1440p and ended with a gigabyte m27q which I'm really enjoying. It's quite good for the price as well. I think AOC makes some very good budget monitors with freesync. Harder to find a budget one with gsync
My current monitor is aoc or dell, don't remember.

I hate playing on a monitor because they are small and colors\blacks are nowhere near a good oled tv, for me it's barely better than playing on mobile with a switch (i know, harsh comparison).
 

MrFunSocks

Banned
Edit: it seems like gsync monitor are more mainstream than what i thought, not sure about vrr oled tho.


But listen, we are discussing about nothing really here, i do think that serious pc gamers are more disturbed by uneven framerate compared to console players, do you think that i'm wrong about this simple concept?!
Yes, I do, because again - an uneven framerate is irrelevant with VRR. Do you even understand what freesync/gsync is?
 

NXGamer

Member
I think most people are being very civil. If this was a DF video where PS5 got even a little flack or a wild conclusion was jumped to it would be 200 pages now with a million personal insults towards Alex or Tom.

It's clear it's his CPU with single core performance of maybe a 2013 to 2015 Intel CPU that is causing drops not his 2070.

I have to go to 4k native before I can get the bridge section to drop to the same level as in his video and even then its only 0.1% lows and it still doesn't drop as low as 46 FPS as in his video - I must have a very strong 2070S:

4lcZD7K.jpg


And here is 1577p which is still more pixels than the PS5:

mg3t6os.jpg


Should my conclusion be that it's the unified architecture of the PS5 that is holding it back? No, this game is a poor port on PC and a BC game on PS5 so any conclusions you jump to are immediately stupid.

The only conclusion I can come to is that I have a similar GPU to him but a much better CPU and my performance is miles better so clearly this game is CPU bound at higher framerates.
It's not my SC CPU, it is the RAM/Bandwidth most likely causing stalls, if I lower resolution (as I say in the video) it improves. You increase to 4K is just putting more load on GPU and bandwidth not affecting CPU that much.
 

martino

Member
In fact Gamer Nexus made a video on how a GTX 1060 can crush a PS5 (which is just ridiculous, which I know too well since I have both a 1060 and a PS5) and people didn't lose their minds like they are doing know over this new Horizon analysis. 🤷‍♂️

Not joking. A fucking 1060! 😂 Read the comments and you'll see people joking about PS5 Pro targeting a GTX 1070.


this shit again. classic taking out of context the subject, is even reading the title difficult nowadays ?
it was a precise test about targeting 120fps and it shows more cpu bootleneck( because of high framerate than gpu ones)
No, it is the best PC I have access to (I am independent and do not get free hardware) but this does not negate the test OR the results.
Covering more than 85% of the PC space is a better example of the Console versus PC conversation though. I would love, and have been trying, to get higher end hardware to get an even better coverage than I have, it continues.

However, the same argument you are making can be levied at the 1% that have a 3090 - I9/Threadripper PC and then say PC beats console, I believe this test is more reflective of the reality.
At least, i'm glad where have now a more honest reason on your part that rely on belief.
 

GHG

Gold Member
Yes, I do, because again - an uneven framerate is irrelevant with VRR. Do you even understand what freesync/gsync is?

No its fucking not. Framerate drops are still framerate drops but the impact of those are reduced due to the fact that you won't get any screen tearing or sharp drops/stutters due to the way that vsync works when enabled (you disable it when using a fresync/gsync display).

If your framerate is bouncing around between 20fps and 80fps you will still see it. It just won't be as bad as if that is happening on a conventional monitor. It's not a silver bullet.
 

NXGamer

Member
The fuck am I reading?
Is this one of those "youre an IT guy you know everything about computers"

John also has a tech degree and was a software engineer and has been working in IT pretty much as long as NXG second they have deep dive talks with actual game and engine developers....the chances of someone guessing better than people who got knowledge from the horses mouth is very slim.


Im not dissing NXG im sure he knows quite alot about game development and has had a few chats with people.
But the DF guys are basically industry standard now and are constantly improving their knowledge with direct interviews and interactions with actual game developers and hardware developers.
On that, lets be clear I have no view on John's history but he has not been working as long as me as he is quite a bit younger than me and has been a journalist for the past 10 years or so, so much, much less.

This is not a dick measuring competition, being an IT support guy is not the same a Backend or Front End dev. Being 1st/2nd or 3rd line support is different again. The fact is some in here attack a view they do not like, stick to facts and not attacks, I hate and despise with a passion any appeal to authority at any time. I do not, and no-one should, believe in experts, they do not exist, specialist do and at some point in anything including Technology you have to specialise in something as everything gets pointy. You can have jack of all trades but by nature they are masters of none, this is why you now have AI, Game, Animation, Graphics, Sound, Data etc etc specialist and not just like the old days when I was growing up and making my own games on 8/16-bit computers. 1-5 man teams making entire games. The basics are the same, the complexity has spiralled.

Let's discuss the info not the distractions as is always the case and here.
 

assurdum

Banned
And this DF weekly can illustrate it on the current subject :

The level of idolatry some of you have for DF it's hilarious. They are nothing more than enthusiast with some tech knowledge. Doubt they can ever touched a code for gaming. A stupid example was their surprise when they did their first face off of DMC on next generation . I'm not sure what exactly was surprising about it.
 
Last edited:

GymWolf

Gold Member
Yes, I do, because again - an uneven framerate is irrelevant with VRR. Do you even understand what freesync/gsync is?
Never tried myself but i know what it does.

My point was different tho, let me put it this way, if you are sensible to uneven framerates like me, you choose pc as your main platform, because even without a gsync monitor i can achieve a perfect framerate most of the times with the game settings.

People who don't care about these things can play on console and being happy, so for me it's natural to say that people on pc is usually more disturbed by uneven framerate because they chose a platform where they can do something to eliminate the problem (if they don't have a gsync monitor like me), on console they can't do nothing except having a still non-common vrr tv, unless we wanna act like every console gamer has a vrr oled in their home or every console game use vrr, it is a new thing and i'm talking about the past years\present, not the future.

So for now the only way to achieve perfect framerate in almost every game is to play on pc where even the lack of a special monitor can be solved by playing with settings.


I don't believe for a second that a guy very disturbed by uneven framerate can chose a console as a main platform where to play the majority of games.
 
Last edited:

Lysandros

Member
this shit again. classic taking out of context the subject, is even reading the title difficult nowadays ?
it was a precise test about targeting 120fps and it shows more cpu bootleneck( because of high framerate than gpu ones)

At least, i'm glad where have now a more honest reason on your part that rely on belief.
What do you think about his statement of "It's equivalent in performance to a mid-high end gaming PC about five years ago, that's where this thing is." at the end of the video (32:50) do you agree with it?
 

assurdum

Banned
It seems to me i bring info on the current subject....
why are you upset and relying on ad hominem ? why de you feel attacked here ?
In what way I feel attacked lol the hell are you trying to argue. I just said the level of credit some of you give to them it's surprising. Meanwhile give the fanboy to NXG. I don't see great difference in what NXG does compared to them.
 
Last edited:

Arioco

Member
this shit again. classic taking out of context the subject, is even reading the title difficult nowadays ?
it was a precise test about targeting 120fps and it shows more cpu bootleneck( because of high framerate than gpu ones)


How can be a test where you run one version at 4K and the other al 1080p be "precise" or show CPU bottleneck when on PS5 the bottleneck is clearly the GPU? Unless you're saying PS5 CPU is causing the game to dip to the 30s, which is nuts.
 

Arioco

Member
What do you think about his statement of "It's equivalent in performance to a mid-high end gaming PC about five years ago, that's where this thing is." at the end of the video (32:50) do you agree with it?


What does he think? Probably that you're taking the statement out of context too. 😂

Those guys are so funny.
 

martino

Member
What do you think about his statement of "It's equivalent in performance to a mid-high end gaming PC about five years ago, that's where this thing is." at the end of the video (32:50) do you agree with it?
out of 120fps mode, of course not.
 

Arioco

Member
out of 120fps mode, of course not.


And if you agree in 120 fps mode you're just wrong. Watch DF analysis (or any other analysis, It doesn't matter at all which one you choose) and you'll see in 120 mode PS5 is running DMCV at "reconstructed 4K" and between "110 to 120 fps" (I'm quoting Jonh and Tom now).




Any idea on how Gamer Nexus got PS5 to dip into the 30s? 🙄 Since I don't even know how to read the title I'm very interested in your explanation. 🙂
 

martino

Member
And if you agree in 120 fps mode you're just wrong. Watch DF analysis (or any other analysis, It doesn't matter at all which one you choose) and you'll see in 120 mode PS5 is running DMCV at "reconstructed 4K" and between "110 to 120 fps" (I'm quoting Jonh and Tom now).




Any idea on how Gamer Nexus got PS5 to dip into the 30s? 🙄 Since I don't even know how to read the title I'm very interested in your explanation. 🙂

This post gave you the wrong idea if it give you the impression i think the tests done there are good.
My point was this video has a context "120 mode".I think we can agree the way games are tested show same kind of problems than here.
And thank to illustrate the out of context thing again....what was said before your quote ? not something like "the first level(s)" ? what does the video show later ? dips in the 60's ?
So can we imagine gamer nexus found and benched a part of the game where load and transition are harder than the one df found ? (without agenda and brand bias you can't be certain it's impossible but i will join you here and i also believe his tools are worst)
You should at least entertain more uncertainty on the finding whatever the source....this exercise cannot provide absolute results.
And the more they are rushed the less representative of possible limitations the findings can be..(youtube practice doesn't help us here)
 
Last edited:

N1tr0sOx1d3

Given another chance
Whelp, now its confirmed, ps5 is better than PC. Hopefully we will see a pc2 soon that can match the hardware.

Who here is saying this because it certainly isn’t me?
I’m pleased that there is a significant performance upgrade over PS4 pro that’s all…
 

Shmunter

Member
Looks like DF members are actually starting to get wind of this

Lol, this guy is something else. Cannot bench console against pc, because no checkerboarding? So when you bench dlss vs native, or ultra vs low settings, or df’s own console vs pc - it was always for nothing, none of it ever mattered.

By this sweaty post alone, it would seem PS5 is indeed punching way above his preferred platform. No need to look any further for evidence.
 
Last edited:

Arioco

Member
Lol, this guy is something else. Cannot bench console against pc, because no checkerboarding? So when you bench dlss vs native, or ultra vs low settings, or df’s own console vs pc - it was always for nothing, none of it ever mattered.

By this sweaty post alone, it would seem PS5 is indeed punching way above his preferred platform. No need to look any further for evidence.


Here you can see Alex (not DF, Alex himself) comparing Days Gone on PS5 (which runs at 4k CB) against PC. Still performance was of course measured on several GPUs. 🤷‍♂️

 
Last edited:

martino

Member
Lol, this guy is something else. Cannot bench console against pc, because no checkerboarding? So when you bench dlss vs native, or ultra vs low settings - it was always for nothing, none of it ever mattered.

By this sweaty post alone, it would seem PS5 is indeed punching way above his preferred platform. No need to look any further for evidence.
it depends on the objective of the test(crazy !)... evalutate tech , show differences, compare hardware performance
Here the objective is supposedly to compare two hardware performance when the same conditions cannot be replicated...is that difficult to see the problem ?
How course DF did it too, but is that a valid excuse ? Recently they said they will not do this anymore if the exact console settings cannot be replicated. They are probably at a point of this kind of bad move cause more problems than the clicks bring money.
Ho, i'm badmouthing DF can i get some PS cultist like now ?
 
Last edited:
It's not my SC CPU, it is the RAM/Bandwidth most likely causing stalls, if I lower resolution (as I say in the video) it improves. You increase to 4K is just putting more load on GPU and bandwidth not affecting CPU that much.
Not trying to be awkward but I'm not sure I understand your point.

Gamer's Nexus have tons of benchmarks using the same motherboard and RAM where even something like a 6 core 3600 is well ahead of a 2700 at 1% and 0.1% lows no matter the resolution.

I'm on PCIE 3.0 too using a card that on paper should only be 10% ahead of yours yet my minimum frame rates are near double yours at 1440p (at the bits I could compare to your video.)

I just played through the proving grounds up to the spider boss thing and my 0.1% low was 74fps. There may be a bandwidth bottleneck at 4k but it looks like you are hitting CPU limitations long before that becomes a problem.

I guess I just don't understand the point you are trying to make about bandwidth as you would be similarly CPU limited in loads of multiplatform games at least assuming Tech Jesus knows how to benchmark.

I think people are underestimating how huge the leap from 2700 to the Ryzen 3000s was even though they were released so close together. It was like the SC jump between an 2013 and 2018 Intel which is huge if your game is CPU limited at all.
 
It's not my SC CPU, it is the RAM/Bandwidth most likely causing stalls, if I lower resolution (as I say in the video) it improves. You increase to 4K is just putting more load on GPU and bandwidth not affecting CPU that much.
MD ray said he believed the stutters lessened as you traverse the area a bit more. Did you run back thru the same area a few times or just the first. I know fortnite stutters like crazy the first time you play the map and is better the next round.
 

Shmunter

Member
MD ray said he believed the stutters lessened as you traverse the area a bit more. Did you run back thru the same area a few times or just the first. I know fortnite stutters like crazy the first time you play the map and is better the next round.
Apart from gameplay, Fortnite pc always stutters like crazy every time in the pre-game lobby as new players join..…suspect it’s loading in the character skins and dances etc. This does not happen on console - the way the i/o plays out between the architectures is significant here. Not a stretch for these things to manifest in other titles.
 
Last edited:

Shmunter

Member
it depends on the objective of the test(crazy !)... evalutate tech , show differences, compare hardware performance
Here the objective is supposedly to compare two hardware performance when the same conditions cannot be replicated...is that difficult to see the problem ?
How course DF did it too, but is that a valid excuse ? Recently they said they will not do this anymore if the exact console settings cannot be replicated. They are probably at a point of this kind of bad move cause more problems than the clicks bring money.
Ho, i'm badmouthing DF can i get some PS cultist like now ?
When you compare the same game across platforms of various capability the comparison is already legitimate. No one cares to know how they compare? - no such thing.

Over and above the basic game comparison, NX went beyond under best endeavours by matching the compute power between the console and pc, the settings in the game, and render load by compensating pc rez for the cb4k on ps. Is it perfect? It can never be, but it’s as close as one could expect. The effort should be praised, not many would bother going to such lengths.
 
Last edited:
When you compare the same game across platforms of various capability the comparison is already legitimate. No one cares to know how they compare? - no such thing.

Over and above the basic game comparison, NX went beyond under best endeavours by matching the compute power between the console and pc, the settings in the game, and render load by compensating pc rez for the cb4k on ps. Is it perfect? It can never be, but it’s as close as one could expect. The effort should be praised, not many would bother going to such lengths.
I agree, I'm just not sure the reasoning is perfectly sound. Game released in a very poor state to begin with and has gotten better over time, but that doesn't mean it's currently a great port either.
 

hlm666

Member
It's not my SC CPU, it is the RAM/Bandwidth most likely causing stalls, if I lower resolution (as I say in the video) it improves. You increase to 4K is just putting more load on GPU and bandwidth not affecting CPU that much.
Why does the 5700xt with the same memory bandwidth not have the issue and performs better?
 
No, it is the best PC I have access to (I am independent and do not get free hardware) but this does not negate the test OR the results.
Covering more than 85% of the PC space is a better example of the Console versus PC conversation though. I would love, and have been trying, to get higher end hardware to get an even better coverage than I have, it continues.

However, the same argument you are making can be levied at the 1% that have a 3090 - I9/Threadripper PC and then say PC beats console, I believe this test is more reflective of the reality.
I didn't see this before my other post and I don't disagree and hopefully someone can help you out with a bit of a selection of gear in the future.

I don't think the PC port of Horizon should be a hill anyone dies on as it started life as a complete mess and is still is to some degree.
 

sendit

Member
Being a software dev on the commercial side is irrelevant. I myself have almost 20 years experience as a software dev in multinational billion dollar companies. Currently employed at one. Doesn’t mean I pretend to know about the Xbox series Xs ray tracing APIs being unfinished like he did.

oh you are one of the people that can’t understand that hotter exhaust does not equal better cooling lol. Enough said. To the ignore list you go. Have fun.
breakdown GIF
 
Yeah i know, some friends in my ex italian forum were getting in a race for who had the most pricey monitor, i'm never gonna spend that amount of money on a fucking monitor :lollipop_grinning_sweat:
I paid 400$ cdn for mine at the beginning of the year and I'm sure they are cheaper now. 170hz 1440p freesync premium IPS.
Why does the 5700xt with the same memory bandwidth not have the issue and performs better?
There is someone with a 2070s and 3600 claiming to not have the issues either.
 
The frame-rate dips he saw were very much likely due to shader caching, IMO. Did a fresh install on my 3070/3700X rig after watching NXG's vid, waited for it to complete optimizing shaders, and then jumped into the gameplay at a very conservative 1440p, DF optimized (med-high) settings with a 60fps cap. The result?

Well, see it for yourself:

Oh, before that, here's a quick screenshot of the HW resource monitor to show idle CPU/RAM/GPU util. before booting up the game (0 background apps running except for OS-related tasks):
JLSQfFQ.png


Here:


SkjWarK.png



It stutters so much initially on the first run to the point it feels unplayable for a moment, but after a while, the perf gets stable when you go into the same location where it stuttered heavily before, there are still one-off stutters here and there but nothing major. So this has got to be shader caching related problem and likely what we see in NX's vid as well, nothing to do with his HW of choice, IMO.

Mind you, the CPU I'm using is slightly faster than PS5's and what Michael has used in his vid and the GPU is also quite a bit more powerful + I'm putting a lot less rendering load on it, yet I see some nasty hitches, frametime spikes, something that not even PS4 suffers from. The PC version definitely has issues and isn't as stable as the console versions as he concluded at the end.

Here's HZD's CPU cores utilization if anyone's curious:
Nz68JWL.png


Some of you guys missing the point entirely. Listen to what is being said at @ 12 min mark in OP's linked vid. If the man said only a PC with my specs or below runs worse than PS5 nobody batted an eyelid including myself. Maybe just an oversight, as if.

You can absolutely have a locked 60 fps at 4K Ultra settings in horizon zero down with up to date specs. That would be double resolution and with all the improvements over ps5 version. [I listed a video in this thread that show those]

Fact is right now the best looking and best performing version of the game is on PC, period.

But man this game didn't age well. Its like PBS is missing entirely, making it look like nearly everything is made of the very same clay, or it's an exceptionally bad implementation if there is any

edit: typo

 
Last edited:

sendit

Member
Some of you guys missing the point entirely. Listen to what is being said at @ 12 min mark in OP's linked vid. If the man said only a PC with my specs or below runs worse than PS5 nobody batted an eyelid including myself. Maybe just an oversight, as if.

You can absolutely have a locked 60 fps at 4K Ultra settings in horizon zero down with up to date specs. That would be double resolution and with all the improvements over ps5 version. [I listed a video in this thread that show those]

Fact is right now the best looking and best performing version of the game is on PC, period.

But man this game didn't age well. Its like PBS is missing entirely, making it look like nearly everything is made of the very same clay, or it's an exceptionally bad implementation if there is any

edit: typo


PC is such a broad term. Instead of PC, cpu/gpu combo should be noted.

Comparison's like this are pretty pointless IMO. A well equiped PC will always win. For example, if you want the best Xbox experience for day 1 releases. You get a PC, not a XSX.
 

NXGamer

Member
Not trying to be awkward but I'm not sure I understand your point.

Gamer's Nexus have tons of benchmarks using the same motherboard and RAM where even something like a 6 core 3600 is well ahead of a 2700 at 1% and 0.1% lows no matter the resolution.

I'm on PCIE 3.0 too using a card that on paper should only be 10% ahead of yours yet my minimum frame rates are near double yours at 1440p (at the bits I could compare to your video.)

I just played through the proving grounds up to the spider boss thing and my 0.1% low was 74fps. There may be a bandwidth bottleneck at 4k but it looks like you are hitting CPU limitations long before that becomes a problem.

I guess I just don't understand the point you are trying to make about bandwidth as you would be similarly CPU limited in loads of multiplatform games at least assuming Tech Jesus knows how to benchmark.

I think people are underestimating how huge the leap from 2700 to the Ryzen 3000s was even though they were released so close together. It was like the SC jump between an 2013 and 2018 Intel which is huge if your game is CPU limited at all.
What I am saying is the game is using the System Ram and Shader Cache often, this means it suffers GREATLY from System Ram speeds and PCIe speeds.

The 2700X has a limit on its memory speeds, I run at 3200Mhz so it suffers and many games do need the full extent, this game does and thus causes big stutters.

It is a combination of many things which can be improved if you reduce resolution but it causes GPU stalls which is what we are seeing often.

In the end a PC or a Console is more than just a GPU or CPU or even memory and when it is stressed (like this) any slower points crop up.

The entire focus on HW is the issue here, I was not and have not made this a test of PC, a better PC can run it better, as can AMD cards can but the game stutters on every spec I have played.
 
Last edited:
PC is such a broad term. Instead of PC, cpu/gpu combo should be noted.

Comparison's like this are pretty pointless IMO. A well equiped PC will always win. For example, if you want the best Xbox experience for day 1 releases. You get a PC, not a XSX.
I'm still confused what makes PS5 version the 'best' , when you can run the game at capped 60 fps at double the resolution and with higher settings on PC ?
 

Tchu-Espresso

likes mayo on everthing and can't dance
Lol, this guy is something else. Cannot bench console against pc, because no checkerboarding? So when you bench dlss vs native, or ultra vs low settings, or df’s own console vs pc - it was always for nothing, none of it ever mattered.

By this sweaty post alone, it would seem PS5 is indeed punching way above his preferred platform. No need to look any further for evidence.
I think you nailed it. He’s a special kid this Alex especially when he feels the need to inject that he doesn’t play console games into discussions as often as possible.
 

Hoddi

Member
It's not my SC CPU, it is the RAM/Bandwidth most likely causing stalls, if I lower resolution (as I say in the video) it improves. You increase to 4K is just putting more load on GPU and bandwidth not affecting CPU that much.
This thread has gotten way off topic. But in case you're interested then I ran the game through Intel's VTune profiler and it doesn't seem very bandwidth limited at all. This is especially true at 60fps where my system only averages ~13GB/s of its ~35GB/s realistic maximum.

Here are a couple of shots that show this rather well. These are captured on a 9900k + 2080Ti at 640x360 with the first running with an unconstrained framerate (at ~120fps) and the other with a 60fps cap.

Uncapped
60fps cap

I don't know what's causing your frame spikes but you could try running at 640x360 just to see how it performs. Even my 2080 Ti is often maxed out at just 720p in this game.
 
Top Bottom