Quantum Break PC Performance Review

[WYP]

Quantum Break had a bad reputation on PC even before it was released, but is that hatred justified? Let's have a look at Quantum Break's Performance on PC.

Read more on Quantum Break's PC Performance.

 

[AngryGoldfish]

For every GTA V, Shadow of Mordor, and Mad Max, with great gameplay, limited bugs and superb performance scaling, there is a Syndicate, Quantum Break, and Gears of War.

 

[Mysterae]

Nice write up.

I'm convinced this is a game with a couple of purposes. Firstly, to get Xbone users to try it on their computers as well, to either convince them that PC's are crap or to show them what's better on the PC. They failed at the latter.
Secondly, it's an experiment with UWP on the masses, to see if we will swallow this ****.


You have an important typo on the third page that gave me a glimmer of hope, but alas it's not true - ".. which is a shame since this game has now SLI or Crossfire support".

 

[WYP]

Nice write up.

I'm convinced this is a game with a couple of purposes. Firstly, to get Xbone users to try it on their computers as well, to either convince them that PC's are crap or to show them what's better on the PC. They failed at the latter.
Secondly, it's an experiment with UWP on the masses, to see if we will swallow this ****.


You have an important typo on the third page that gave me a glimmer of hope, but alas it's not true - ".. which is a shame since this game has now SLI or Crossfire support".

Typo fixed, thanks for the spot.

A lot of extra work went into this one, especially when it came to getting benchmarking data. I like the idea of percentile based framerate data, which is something that I will be looking more into for the future.

 

[SimplexPL]

While there has been a lot of talk on the internet about the "terrible" PC performance of Quantum Break I can say that these reports have been exaggerated
(...)
Sadly on my system I was unable to get the game to run at a solid 60FPS, regardless of the GPU that I used (...) which is a very strange prospect given that I was playing Quantum Break from an SSD on a system with an overclocked i7 6700K CPU and 16GB of 3200MHz DDR4 memory.

So you were unable to run at solid 60fps at 1080p (which is effectively 720p if NeoGAF is to be believed) at lowest settings on "overclocked i7 6700K CPU and 16GB of 3200MHz DDR4 memory" and a 980Ti/Fury, yet reports that the game runs terribly are "exaggerated"?
This game runs terribly, period.

Not being able to sustain 60fps at the lowest graphical settings at 1080p (reprojected 720p) on a fastest possible CPU, memory and GPU means, that the talk on the internet about the "terrible" PC performance is accurate, not exaggerated. What is the last AAA game that run that badly?

 

[Primey]

What a bloody mess of a game

 

[SimplexPL]

While there has been a lot of talk on the internet about the "terrible" PC performance of Quantum Break I can say that these reports have been exaggerated

No, they are absolutely not exaggerated, as evidenced here: http://www.eurogamer.net/articles/digitalfoundry-2016-what-went-wrong-with-quantum-break-pc

Quantum Break has arrived on the PC and to suggest that the results are disappointing would be a massive understatement. PC gamers are left out in the cold with another high profile release failing to deliver expected levels of performance, features and customisation. As things stand, it simply isn't possible to achieve a smooth frame-rate on any PC hardware configuration and thanks to the profound limitations imposed on gamers by the Universal Windows Platform, there's no way to fix it

 

[Doothe]

What monitor did Overclock use to test Quantum Break?

Nooooooooooooooooooooooooooo double posting please

What monitor did Overclock use to test Quantum Break?

DigitalFoundry showed that Quantum Break has some 5/6th V-sync implementation.The only way for either 960/380 to hit 55 or 56 would be to use a 66 or 67.2 refresh rate monitor. This new dx12 API is asking a lot out of benchmarkers.

 

[Mysterae]

Typo fixed, thanks for the spot.

A lot of extra work went into this one, especially when it came to getting benchmarking data. I like the idea of percentile based framerate data, which is something that I will be looking more into for the future.

No problem.

I can imagine the work that goes in to the review. When we (or I) criticise the results, it's game we are disappointed with, not the review.

If you are thinking of using statistics and displaying the data, have a look at box and whisker plot:

Caution, if you're not in to statistics the video below will be a bit dry!



Excel has this as a function so they are easy to produce, although I'm not sure if your audience would get it, and the b&w charts don't tell the whole story.

 

[WYP]

What monitor did Overclock use to test Quantum Break?

Nooooooooooooooooooooooooooo double posting please



DigitalFoundry showed that Quantum Break has some 5/6th V-sync implementation.The only way for either 960/380 to hit 55 or 56 would be to use a 66 or 67.2 refresh rate monitor. This new dx12 API is asking a lot out of benchmarkers.

I can't think of the monitor model number off the top of my head, but we are using a Samsung 4K 60Hz monitor for testing. Right now I am looking into acquiring a 144Hz monitor in the future for testing purposes, but right now all I have is 60.

Forced V-Sync on offers a lot of problems for presenting data and as the traditional average FPS data that we used does not really line up with a particular V-Sync interval.

the Average of 55 etc is because the cards stayed at 60 a lot of the time, but suffered a lot of drops to 30 due to v-sync. If we used median for our data, in that case, it would have been 60, though anyone that is not well acquainted with mathematics could find a lot of additional data confusing.

UWP presents a lot of issues of benchmarking, especially when it comes to the tools that we need to use. VR also presents similar issues, with traditional applications like fraps being completely useless when it comes to giving accurate/usable data.

No problem.

I can imagine the work that goes in to the review. When we (or I) criticise the results, it's game we are disappointed with, not the review.

If you are thinking of using statistics and displaying the data, have a look at box and whisker plot:

Caution, if you're not in to statistics the video below will be a bit dry!



Excel has this as a function so they are easy to produce, although I'm not sure if your audience would get it, and the b&w charts don't tell the whole story.

New ways of presenting data is certainly something worth considering though it will need to be done in a way that the layman can understand.

 

[ImprovizoR]

Who cares? It's just a mediocre 8-12 hour linear shooter that people will forget about in a few months anyway. People wouldn't think twice about skipping it if it wasn't for the massive PR behind it and the fact that it was made by Remedy. People forget that these guys haven't made that many games, and aside from the spooky atmosphere, Alan Wake was also mediocre.

 

[NeverBackDown]

I can't think of the monitor model number off the top of my head, but we are using a Samsung 4K 60Hz monitor for testing. Right now I am looking into acquiring a 144Hz monitor in the future for testing purposes, but right now all I have is 60.

Forced V-Sync on offers a lot of problems for presenting data and as the traditional average FPS data that we used does not really line up with a particular V-Sync interval.

the Average of 55 etc is because the cards stayed at 60 a lot of the time, but suffered a lot of drops to 30 due to v-sync. If we used median for our data, in that case, it would have been 60, though anyone that is not well acquainted with mathematics could find a lot of additional data confusing.

UWP presents a lot of issues of benchmarking, especially when it comes to the tools that we need to use. VR also presents similar issues, with traditional applications like fraps being completely useless when it comes to giving accurate/usable data.



New ways of presenting data is certainly something worth considering though it will need to be done in a way that the layman can understand.

Personally I don't understand any of that percentile crap. I'm more interested in raw numbers. I don't need super specifics, if I know the minimum, then really I can just adjust to that knowing that's the lowest i'll get. I don't need to know it'll be in this framerate range .01% of the time. The only thing I think OC3D as a whole can improve with charts is there readability. Just need a newer format and a compare chart would be cool, like a dedicated hardware page where all the results are uploaded into a single data pool and can be collected and shown indivdually or against another product. Kinda like what other websites have, it's a very cool and useful feature. Don't think we have the staff size or time to run something like that though.

 

[SPS]

Personally I don't understand any of that percentile crap. I'm more interested in raw numbers. I don't need super specifics, if I know the minimum, then really I can just adjust to that knowing that's the lowest i'll get. I don't need to know it'll be in this framerate range .01% of the time. The only thing I think OC3D as a whole can improve with charts is there readability. Just need a newer format and a compare chart would be cool, like a dedicated hardware page where all the results are uploaded into a single data pool and can be collected and shown indivdually or against another product. Kinda like what other websites have, it's a very cool and useful feature. Don't think we have the staff size or time to run something like that though.

Personally, a box plot makes it much easier to interpret this kind of data. A bar chart doesn't perform well for displaying in these instances, IMO. An array of graphical representations would be ideal.

 

[Dicehunter]

Not liking this universal Windows stuff so far, The performance even on decent rigs is horrible -

• No V-Sync option
• No overlay option
• No mGPU support out of the box without a patch
• No Fullscreen option
• No modding option.

Just a whole lot of "Not allowed".

 

[NeverBackDown]

Personally, a box plot makes it much easier to interpret this kind of data. A bar chart doesn't perform well for displaying in these instances, IMO. An array of graphical representations would be ideal.

I agree, bar graphs just get flooded and look out of place once a lot of data goes into them

Not liking this universal Windows stuff so far, The performance even on decent rigs is horrible -

• No V-Sync option
• No overlay option
• No mGPU support out of the box without a patch
• No Fullscreen option
• No modding option.

Just a whole lot of "Not allowed".

and no Adaptive Sync..
But they are working on it, so just gonna have to wait.

 

[Dicehunter]

and no Adaptive Sync..
But they are working on it, so just gonna have to wait.

For "AS" to work all you need to do is have the ability to turn off v-sync which I honestly do not understand why v-sync toggle wasn't an option at launch, On console I can understand it but on PC it seems silly.

 

[RDex73]

Game works fine????

I must have some kind of magic PC, this game runs solid at 1440p with medium settings on my GTX970!!!