• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Ratchet & Clank Rift Apart Benchmark Test & Performance Analysis

lol that statement is not placing blame on devs they found a bug and both sides are working to fix it.

do you always just make stuff up and run with it?

So how often did you read exactly "working with the developers" in release notes? :wtf: The game has obviously quite some issues, not only with AMD cards but also with Nvidia cards.

If they could fix it themselfs they would do so without mentioning the developers. Pretty clear. 1+1 and some basic common sense isn't your thing, eh?
 
Working with developers does not equal "Pretty much blame the game developers for the problems"

seems english isn't your thing.

They worded it like that for a reason when you have a working relationship finger pointing solves nothing.

Your statement says they are 100% blaming the dev for the issue.

These are the quotes from the 23.7.2 drivers

"Application crash or driver timeout may be observed while playing Ratchet & Clank™: Rift Apart with Ray-Tracing and Dynamic Resolution Scaling enabled on some AMD Graphics Products, such as the Radeon™ RX 7900 XTX."

"AMD is working with the game developers of Ratchet & Clank™: Rift Apart to resolve some stability issues when Ray-Tracing is enabled."
 
Last edited:
The performance results look great. Nixxis is getting pretty good at porting these games. Gonna look forward to playing it monday
 
The new x60 series GPUs lag behind their predecessors(1080p). I'm not the only one who noticed this disgrace, right?
 
That is a very nice video from DF, I expect the focus at end on the portal sequence is likely at my guess is that the differential is that the PC is still doing some stuff in RAM affecting it, the PS5 only has VRAM, so it has much more bandwidth available. So the solution would probably be moving more stuff to GPU but in turn that increases VRAM requirement and I expect the porting team have had to make compromises.

Also good in pointing out the issue on low VRAM PC's is to shutdown anything using the GPU for acceleration, typically all game launchers do, discord does by default, telegram does, browsers do by default. Although with browsers and discord you can disable GPU acceleration to significantly reduce system VRAM usage, then the only issue would be the processes stealing cycles.

I plan to use my iGPU for desktop use (still testing this on my 9900k before I sell it), so all system apps use system RAM regardless of GPU acceleration, and then my GPU is exclusive to games. This will be my setup until I get something with decent VRAM capacity. Discrete GPUs do support shared memory same as iGPU's so would be good if Microsoft made a DWM update which uses shared memory for desktop GPU apps or an option where you can choose on a per app basis..

The only thing I noticed is that some textures appeared to fade in instead of loading in (almost) instantaneously when I hadn't yet removed the DLLs. But I'll tell you that the frametime issues were a LOT more noticeable with DS enabled, especially with very high textures

DS is worth it, playing at 200fps instead of 160, is no meaningful benefit on a platform game, I play games at 30 or 60fps capped anyway, but I would most definitely be noticing texture issues.
 
Last edited:
Fwiw, dsogaming had better 1% lows at 4K on the 4090 than the 7900 xtx (but not at 1080p) during their testing so ymmv.
 
"AMD is working with the game developers of Ratchet & Clank™: Rift Apart to resolve some stability issues when Ray-Tracing is enabled."

By the wording they pretty much blame the game developers for the problems. :cool: Guess they are not very experienced yet with the implementation of RT for PC games.
It's probably because they want the devs to run some tests AMD can't. If this wasn't a confirmed driver issue, it wouldn't be mentioned in driver's known issue section, would it?
 
It would be great if you could include the PS5 in the comparison, thank you for your hard work and stats, you are the best
 
My impressions from a combination of this review and DF video.

Lowest quality setting has absolutely horrid textures, so I can only assume the high VRAM usage is again down to unified memory architecture memory management which I think is going to dominate many new AAA titles moving forward.

Lack of stutters is likely DS, so its good thats doing what it is supposed to do, however DS logically surely will consume VRAM, so will also be a contributor to the VRAM usage, as using system RAM would require the CPU.

Even the lowest resolution setting at your max settings is going to be too much for my 10 gig 3080, although hoping the setting below that is manageable.

VRAM usage might go up further if they fix the problem with not all high res textures loading on PC port.

Think I might buy this game just as a technical test on my PC.


@W1zzard did state the 8 gig cards suffered. 6700XT has 12 gigs VRAM.
DF is the most partisan half-technical site there is, evidently pro console.

A patch came out the day after the first test, yet the final analysis contains parts (photos, references, etc...) of the problems solved, even if it is said in the whole text, but half-heartedly.

RTAO looks great, but they look in odd places to make the difference seem insignificant.
Shadows look for mistakes, in the first analysis they considered natural shading differences to be flaws.

The comparison of the last article is deliberately set with the 30fps settings, while the RT performance ones are decidedly lower.
In the preview video it was immediately noticeable that there were 4 vehicles in the air bazaar, much less than the PC ones, perhaps below low.

On the SSD they showed the 5400rpm HD of PS4, not only that, they also installed Windows.
It didn't help, because SONY said that only PS5 SSD was good, it didn't speak di HDD.

It's not the first time they've mixed as they prefer, and their tendency is now clear: On PC, big titles with bad ports (that's not them, it's the excuse to look for flaws, making the console version look like the original one, so better optimized).

I remember GTA-5 PS5 vs PC, PS5 won even though it lacked many things and was at 1440p.
There the RT in the shadows - which suck at every opportunity on PC - tipped the balance.
On PS5 Low AF as usual, low LOD, sparse grass etc... no problem.

I would have many other examples of their modus: On PC, a defect rejects the porting which is not, on consoles it's always good, even when it drops to 720p at 40fps and with FSR-1, as on FF-XVI, which from the title seemed the most stable job in the world...

Forspoken idem, a disaster that became half good and half bad.
In Returnal they were looking for the difference with 'AO at a minimum on PS5 by zooming in on a meaningless detail, when it was enough to see the preview in the settings to understand the big difference.

In this regard, Kena for 2 a little more occlusion than PS4 seemed to change the world, while textures mixed with low, aniso always low, shadows worse, low res VS. PCs were not noteworthy differences.

It's so obvious to me that it bothers me to point it out.
Probably many follow without having a historical memory of their values.
If you pay attention, you always notice the low-level tricks useful to convince noobs.
Then above it becomes less easy to notice them, It serves to defend oneself from the partisanship that is seen.

Someone pushes from behind, and pushes hard.
 
Last edited:
Working with developers does not equal "Pretty much blame the game developers for the problems"

seems english isn't your thing.

They worded it like that for a reason when you have a working relationship finger pointing solves nothing.

Your statement says they are 100% blaming the dev for the issue.

These are the quotes from the 23.7.2 drivers

"Application crash or driver timeout may be observed while playing Ratchet & Clank™: Rift Apart with Ray-Tracing and Dynamic Resolution Scaling enabled on some AMD Graphics Products, such as the Radeon™ RX 7900 XTX."

"AMD is working with the game developers of Ratchet & Clank™: Rift Apart to resolve some stability issues when Ray-Tracing is enabled."
I think every company releasing a product with drivers are obligated to tell consumer that there is a problem with the game. it is a must despite whose fault is it. The company would put it in the list that there is a problem. But I kinda agree with @MarsM4N. If it was only AMD or NV, they would have not talk to the developer if it was only AMD or NV driver issue. You have to remember the game works fine on PS5 why it doesn't on PC? It kinda makes not much sense.
The lows are staggering thus It would seem the game is optimized well for AMD and yet you can't run it with FSR nor RT. That is weird I'd say.
 
Last edited:
You have to remember the game works fine on PS5 why it doesn't on PC? It kinda makes not much sense.
The game running fine on PS5 even if it has AMD hardware does not equal it will run well on AMD PC hardware as the PS5 is a custom unit with dedicated hardware blocks for decompression etc.

And the only people that know why are AMD and the dev's so its wise not to speculate and point fingers. All of Nixxies PS5 ports have all required patching on release no different than any other PC games.

I own some of their Ports of PS5 games.

HZD
Spiderman remastered
GOD of War

None of these were perfect on launch regardless of them running well on PS5.
 
Last edited:
You are missing some obvious reflections I guess, for example in the 2nd shot up, right...
But I gotta say while playing none of this matters much. That vibe the game gives off at the highest settings - that you're playing some pixar cgi film - remains untouched ^^
That’s probably the only comparo where there’s a big difference, and maybe the floor reflections on the last one too. Even though the cynic in me thinks that those could be made without RT (screen space reflections?), but I don’t know how lighting is done in games.
Pixar-like vibes is the best way to describe it, 250% agree with that!
 
The game running fine on PS5 even if it has AMD hardware does not equal it will run well on AMD PC hardware as the PS5 is a custom unit with dedicated hardware blocks for decompression etc.

And the only people that know why are AMD and the dev's so its wise not to speculate and point fingers. All of Nixxies PS5 ports have all required patching on release no different than any other PC games.

I own some of their Ports of PS5 games.

HZD
Spiderman remastered
GOD of War

None of these were perfect on launch regardless of them running well on PS5.
Fwiw, Nixxes did not work on the God of War pc port. And they were not involved in the HZD port until much later on after release.
 
Fwiw, Nixxes did not work on the God of War pc port. And they were not involved in the HZD port until much later on after release.

You are right done by Jet Pack supervised by Nixxes

"Although this port was mainly handled by Canadian studio Jetpack Interactive, the game’s original creators at Santa Monica Studio helped supervise the project."

HZD done by Guerrilla Games and Virtuos and patching by Nixxes.

 
You are right done by Jet Pack supervised by Nixxes

"Although this port was mainly handled by Canadian studio Jetpack Interactive, the game’s original creators at Santa Monica Studio helped supervise the project."

HZD done by Guerrilla Games and Virtuos and patching by Nixxes.


*You are right done by Jet Pack supervised by Santa Monica Studio

Fixed that for you.
 
*You are right done by Jet Pack supervised by Santa Monica Studio

Fixed that for you.
Episode 1 Slow Clap GIF by One Chicago
 
Man you can see the moment the 10GB on a 3080 becomes a limiting factor next to the 12GB 4070. They're both neck and neck in all resolutions without RT, then with RT the 4070 is just that tiny bit ahead in 1080p and 1440p, then boom, 2160p, and miles behind.
You can see in the VRAM chart that 1440p RT can use 11455MB maximum, so I'm guessing in the 3080 10GB test it was pretty much running nigh on full, then the jump to 2160p pushed it way over and it tanked.

And next to the 3090 with its massive 24GB, the 3080 is about 8% behind in 1080p and 1440p, then a gigantic 46% behind in 2160p.

Obviously when playing on a 3080 in 2160p you could lower various settings and use DLSS to get a solid 60. But still, just seeing the differences there on max is dramatic.
 
Seemed to work fine on AMD console's.

Before Nvidia sponsored it.
A very selective take here.

It ran fine ........ on a PS5 which has bespoke silicon and only one variant of it, the console uses it's own custom and bespoke API, not DX12U or Vulkan, so they've completely overhauled and ported the game to an engine that runs on a PC compatible API, they've also added two other effects that aren't on the console version that require more work for the GFX card and CPU to process, and AMD have since launched a new generation of cards. So perhaps it even works on RDNA2 but not RDNA3 but they want to have it working across all before patching it in.

But hey Nvidia sponsored it so lets point at that, excellent reasoning indeed. Well it got the likes from the usual crowd so that's nice.

This isn't even close to the same kind of situation as blocking competitor upscaling (something a lot more RDNA2 and 3 users should be upset about considering how excellent XeSS 1.1 DP4a is).
 
The game running fine on PS5 even if it has AMD hardware does not equal it will run well on AMD PC hardware as the PS5 is a custom unit with dedicated hardware blocks for decompression etc.

And the only people that know why are AMD and the dev's so its wise not to speculate and point fingers. All of Nixxies PS5 ports have all required patching on release no different than any other PC games.

I own some of their Ports of PS5 games.

HZD
Spiderman remastered
GOD of War

None of these were perfect on launch regardless of them running well on PS5.
I'm not pointing any fingers but I sure look at the posts and other people are. If it is so bad on a PC and there is a ton of problems, keeping in mind PS5 is using AMD hardware and on top of that the min FPS for the cards are pretty high. It does seem the game has been tweaked for the hardware. Not saying there is nothing to improve though.
The game runs well on AMD hardware and you can see it clearly. Tweaks are required always. For the record, I'm not speculating anything. Even though it is PS5 you still have to understand it is an AMD hardware and the driver is not that much different.
 
Last edited:
A very selective take here.

It ran fine ........ on a PS5 which has bespoke silicon and only one variant of it, the console uses it's own custom and bespoke API, not DX12U or Vulkan, so they've completely overhauled and ported the game to an engine that runs on a PC compatible API, they've also added two other effects that aren't on the console version that require more work for the GFX card and CPU to process, and AMD have since launched a new generation of cards. So perhaps it even works on RDNA2 but not RDNA3 but they want to have it working across all before patching it in.

But hey Nvidia sponsored it so lets point at that, excellent reasoning indeed. Well it got the likes from the usual crowd so that's nice.

This isn't even close to the same kind of situation as blocking competitor upscaling (something a lot more RDNA2 and 3 users should be upset about considering how excellent XeSS 1.1 DP4a is).
You read a whole shit load into one factual statement, I'll leave you to the mental gymnastics session.


Ps just noticed the whataboutisn I get it you love Nvidia.
 
Last edited:
Back
Top