Testing frame rates on 4 different machines. Will update.

[Valorin]

Ryzen 5 5600
RTX 3070
32gb Ram
SSD
Ultra settings (Ultra for capture for terrain) and supersampling at 1.0.
2560x1440

Overall framerate couldn't be smoother.
Sometimes I get a few stutters on planets mostly when there is terrain drawing a lot of shadows. Happens very rarely though.
When scavenging I noticed FPS going down a bit when scavenging larger ships with npcs around.
In settlements framerate is constantly a bit worse (not using a counter though) at daytime.
At nighttime it's worse as soon the power is on and there's a lot of light sources.

All these things only apply when I am looking directly at them. Looking away into the sky or elsewhere and it goes back to smooth again.

None of these issues make the game unplayable for me.

 

[Agrar Therenus]

Ryzen 5 5600
RTX 3070
32gb Ram
SSD
Ultra settings (Ultra for capture for terrain) and supersampling at 1.0.
2560x1440

Overall framerate couldn't be smoother.
Sometimes I get a few stutters on planets mostly when there is terrain drawing a lot of shadows. Happens very rarely though.
When scavenging I noticed FPS going down a bit when scavenging larger ships with npcs around.
In settlements framerate is constantly a bit worse (not using a counter though) at daytime.
At nighttime it's worse as soon the power is on and there's a lot of light sources.

All these things only apply when I am looking directly at them. Looking away into the sky or elsewhere and it goes back to smooth again.

None of these issues make the game unplayable for me.

You really ought to be providing metrics.

"couldn't be smoother" is very much a subjective statement, as it could actually be smoother in numbers. Only you run the risk of vindicating FD's statement about how it runs 'fine' on some machines.

I mean for instance I can enjoy the game at 1080p 30fps (capped because it often can't get much higher, and of course one should cap at 30, 60, 120 or other multiples of this) but that's still pretty bad performance given the graphics quality.

 

[TheMindBuddah]

Running Odyssey on a i7 4790k, 32gb and RTX 3070 running on high

Single screen in space 120-144 single screen @ 2560x1440
Triple screen in space 110-130 triple screen @ 7680x1440

Single screen sat in Carrier 90-120 single screen@ 2560x1440
Triple screen sat in Carrier 60-70 triple screen @ 7680x1440

Single screen at settlement 28-40 single screen @ 2560x1440
Triple screen at settlement 19-35 triple screen @ 7680x1440

One of the strangest things I found was switching from triple 1440p to single 1440p doesn't really have the performance difference I would have expected, and even that can be totally inconsistent. While each frame rate was taken while stationery (it was a quick test) and I get bigger dips with triple than single (expected at games present state). Elite is the only game I play triple screen and the whole system is purely for playing games.

 

[Agrar Therenus]

It seems to me it's heavily GPU intensive, which to me seems the opposite of what would be expected on the ground, what with all the CPU stuff like NPCs and having to calculate all the space stuff when you look outside in the dock or ground base.

 

[Ambient.Impact]

Watch the performance issue be caused by something insanely weird, like which brand of Memory your Graphics card is using. I know my card has GDDR6 from Micron. I honestly can't think of anything that would cause such consistently inconsistent performance across such a wide range of hardware. What about Antivirus, is anyone who is NOT having problems running Bitdefender?

Alas, my perfomance is pretty bad but I'm also running Bitdefender.

 

[ebbrell]

i had a play with my system last night. when in a station on foot, CPU was hitting 20% usage and my 2080ti was 100% being used and RAM was 10.5gb of the 11gb on the GPU.
im getting 43 - 45 FPS with my 2080ti, its obviously GPU power which is needed in my case.
weird, depending which direction i was facing he performance chances, its like the card is busy doing stuff I cannot see.

 

[Comtedeloach]

Are you going to show your calculations to back up that assertion?

For the record, my AMD 5800 with 6800 machine runs E: D poorly. There isn't anywhere near enough feedback on these forums to make the kind of assertion you are making.

No, I am not going to go back over 18 pages of posts and count the number of Nvidia cards vs the number of AMD cards in each post, and half of the first posts don’t even mention their hardware. If you can’t see there is a huge number of people with problems using Nvidia vs AMD from just reading the posts that’s your reading comprehension problem, not mine.

oh, BTW, I worked in IT for 20+ years. I can spot trends quite easily. Sure, my analysis is off the cuff, and doesn’t consider every possible variable, but from the information we got here, you can’t say AMD is the problem when so many more people with Nvidia cards are complaining about frame rate problems. I would suggest that there is something wrong with your set up if you are having problems with such high end equipment from AMD. Might want to go through all your settings and see if something isn’t set correctly.

 

[Koma Pahu]

No, I am not going to go back over 18 pages of posts and count the number of Nvidia cards vs the number of AMD cards in each post, and half of the first posts don’t even mention their hardware. If you can’t see there is a huge number of people with problems using Nvidia vs AMD from just reading the posts that’s your reading comprehension problem, not mine.

What proportion of ED players use Nvidia compared to AMD?

When the blood clots were first linked to the COVID vaccine a higher number of women were affected than men. But when you looked at vaccinations a higher proportion of women had been vaccinated. Women live longer and more women work in the NHS and care work. Now the numbers have evened out, women are not considered at greater risk than men.

There's far more to statistics than raw numbers...

 

[Comtedeloach]

No, I am not going to go back over 18 pages of posts and count the number of Nvidia cards vs the number of AMD cards in each post, and half of the first posts don’t even mention their hardware. If you can’t see there is a huge number of people with problems using Nvidia vs AMD from just reading the posts that’s your reading comprehension problem, not mine.

What proportion of ED players use Nvidia compared to AMD?

When the blood clots were first linked to the COVID vaccine a higher number of women were affected than men. But when you looked at vaccinations a higher proportion of women had been vaccinated. Women live longer and more women work in the NHS and care work. Now the numbers have evened out, women are not considered at greater risk than men.

There's far more to statistics than raw numbers...

I am guessing that’s why the Fdev developers wanted the hardware survey done. You will note that the whole idea of a hardware survey was met with derision by the vast minority of the community that’s so upset over Odyssey. They seems to think that Fdev is only trying to find someone to blame the problems on vs using it as a tool to figure out the where some of the problems are.

oh, but they did study the problems of blood clots in women being higher to find out if there might be a reason, they didn’t just say “ it, let’s see what the numbers are in 6 months, no reason investigate this further..”

 

[Crank Larson]

No, I am not going to go back over 18 pages of posts and count the number of Nvidia cards vs the number of AMD cards in each post, and half of the first posts don’t even mention their hardware. If you can’t see there is a huge number of people with problems using Nvidia vs AMD from just reading the posts that’s your reading comprehension problem, not mine.

oh, BTW, I worked in IT for 20+ years. I can spot trends quite easily. Sure, my analysis is off the cuff, and doesn’t consider every possible variable, but from the information we got here, you can’t say AMD is the problem when so many more people with Nvidia cards are complaining about frame rate problems. I would suggest that there is something wrong with your set up if you are having problems with such high end equipment from AMD. Might want to go through all your settings and see if something isn’t set correctly.

Working in IT for any number of years doesn't give your analysis any more weight - it is missing facts and figures. More to the point, as has been proved on numerous occasions, a tiny percentage of the player base even visit the forums.

You're welcome to your opinion, though.

 

[MCCART]

i5 3570k
980 ti
16gb RAM

It seems to me that looking through glasses tanks the framerate. Can anyone else confirm that?

If I look away I get on average 10-12 FPS more-- even when the scene is clearly more complex.

Looking through windows of shops I get about 30, but if I look elsewhere, including the massive window that looks out to the docking area (they don't have a reflective surface it seems so they might be why), I get 40-45. I'm pretty sure in some docks I get lower FPS.

This is at 1080p with mostly high settings, and ground details like terrain and materials on ultra. The only thing that's set to medium are shadows.

I don't know about glass but there are assets that, appear, to tank performance. I say "appear" because, in reality, you can't tell if it's the asset, an anchor for the asset, something behind the asset, etc. The ones I've seen are animated panels on consoles or walls and some of the animated signs in stations and hangers.

As for general performance, I see:

1080P x 1.0 SS Ultra at the moment.
Space 60-100+ fps
Stations 25-55 fps
General on-foot 25-50 fps
CZ on-foot 11-35 fps

AMD Ryzen 5 2600X
AMD Radeon RX 5600 XT
32 GB DDR-3000 (running dual channel)
1TB SSD

 

[LoneTech]

45 FPS is not acceptable for hardware that is "recommended". If the tutorial alone already shows weaknesses, it's sufficient as a benchmark to compare systems.

You can increase the stress and expand the test basis once the tutorial runs smoothly.

I can confirm that main RAM timings are a dominant factor in my case (i7-4790K, 32GiB DDR3-2400, Radeon RX 6900 XT, 2560x1440@120Hz with high settings, blur and depth of field off). Reloading XMP profile #2 in BIOS caused it to jump from rarely going above 50 during tutorial, to sitting around 70 and rarely dipping below 60. This doesn't cure the slow resource leak I was hunting, but raises the baseline considerably. Also not confirmed is whether it's mainly latency or bandwidth constrained.

During gameplay I did encounter a rather large settlement that dropped the frame rate all the way to 30 (I had also adjusted to Ultra as base settings). However, the GPU was not busy; it was CPU/RAM limited still.

 

[KapitanKugel]

I can confirm that main RAM timings are a dominant factor in my case (i7-4790K, 32GiB DDR3-2400, Radeon RX 6900 XT, 2560x1440@120Hz with high settings, blur and depth of field off). Reloading XMP profile #2 in BIOS caused it to jump from rarely going above 50 during tutorial, to sitting around 70 and rarely dipping below 60. This doesn't cure the slow resource leak I was hunting, but raises the baseline considerably. Also not confirmed is whether it's mainly latency or bandwidth constrained.

During gameplay I did encounter a rather large settlement that dropped the frame rate all the way to 30 (I had also adjusted to Ultra as base settings). However, the GPU was not busy; it was CPU/RAM limited still.

in my case my xeon e3-1231 v3 is about 60% in odyssey- Only GPU is giving fire on 100% (gtx 970)

 

[sneakyc4]

Ryzen7 3750h / AMD Rx 5500m / 24GB DDR4 2600 / Low settings

120 fps in space
24 in starports
19 in conflict zones

 

[Jack Dowd]

My laptop has a GTX1660ti with 6gb vram, 16 gb ram, i7-9750h

In horizons i could dual log a steam and an epic account with no discernible frame drops

I tried the same in EDO and i get like 15 fps in concourse
A single client can do 30 on average (between 27 and 32 after patch 1, before i was getting 35-45)

I have this same exact setup in my laptop and I'm getting 50+ FPS in the concourse, except planetary ports. This is with a mix of High/Ultra settings in 1080. Weird to see such variance between seemingly identical machines.

 

[Darth Ender]

the difference in similar hardware reeks of a performance related problem triggered by an action being done/taken by some players and not others (On top of certain settings varying between people that may be important in how they are combined with others in triggering the problem).

Like some common behavior some players may do such as viewing the system map vs others who aren't before visiting a planet. Or some other activity. Triggering a bug in a shader or the pipeline in general leading to stupid things like consuming all of the video card's texture memory. Leading to vram thrashing with system ram (which also brings system ram speed into play - where quad channel high frequency ram will hide the issue better than dual or single channel system ram). So basically we all get the overdrawing issues with the pipeline caused by occlusion failures and poorly optimized shaders, but only some of us are really being slammed by the texture memory thrashing.

All things that shouldn't be happening in the first place.

I am guessing that’s why the Fdev developers wanted the hardware survey done. You will note that the whole idea of a hardware survey was met with derision by the vast minority of the community that’s so upset over Odyssey. They seems to think that Fdev is only trying to find someone to blame the problems on vs using it as a tool to figure out the where some of the problems are.

oh, but they did study the problems of blood clots in women being higher to find out if there might be a reason, they didn’t just say “ it, let’s see what the numbers are in 6 months, no reason investigate this further..”

Fdev was given reports during their "alpha" regarding these issues (all of which existed during the alpha). That's why the request was met with derision. Hardware reports can also be generated by the client but they wanted users to manually create and send reports to a certain email address. Effectively pretending that the alpha didn't happen and any work players did towards that was a waste. Also when 70% of the playerbase has similar issues, the idea that your internal QA isn't having a similar problem anyway (and so you dont need additional info from players) doesn't make any sense. The fact that the console was delayed due to performance issues also points to the fact that they're well aware of the problems and have knowledge of these issues already (even before the alpha) and none of it is news to them.

so yea, the idea that they need more info like this was all a surprise to them and it's such a small problem that they didn't experience it until players did after release is not just insulting to us players but just a ridiculously obvious stall and blame shifting tactic.

 

[Comtedeloach]

Darth Ender Said:
"Fdev was givDarth Enderen reports during their "alpha" regarding these issues (all of which existed during the alpha). That's why the request was met with derision. Hardware reports can also be generated by the client but they wanted users to manually create and send reports to a certain email address. Effectively pretending that the alpha didn't happen and any work players did towards that was a waste. Also when 70% of the playerbase has similar issues, the idea that your internal QA isn't having a similar problem anyway (and so you dont need additional info from players) doesn't make any sense. The fact that the console was delayed due to performance issues also points to the fact that they're well aware of the problems and have knowledge of these issues already (even before the alpha) and none of it is news to them.

so yea, the idea that they need more info like this was all a surprise to them and it's such a small problem that they didn't experience it until players did after release is not just insulting to us players but just a ridiculously obvious stall and blame shifting tactic."


I will put to you that a lot less players played the alpha than played the release version, so it would be a vastly larger sample size which would make it easier to see trends. I didn't play the alpha, because alpha and beta testing sucks. I beta tested Red Baron 3D back in the 1990's, and it ruined the game for me.

Consoles are always behind the level of technology of PCs so that's why they delayed the consoles. If PCs have problems, then consoles are definitely going to have the same problems or worse.

As far as the stall and blame tactics go, what a crock of crap. They haven't tried to blame it on anyone else, and not because the vast minority of players on the forums called them out on it. they need more data to see if there are any common denominators that could cause problems. They also have to work with Nvidia and AMD to get the drivers working if that's the issue, or even a small part of it.

Oh, and BTW, this forum software sucks.

 

[Darth Ender]

I will put to you that a lot less players played the alpha than played the release version, so it would be a vastly larger sample size which would make it easier to see trends. I didn't play the alpha, because alpha and beta testing sucks. I beta tested Red Baron 3D back in the 1990's, and it ruined the game for me.

It doesn't matter if you did it or not. A lot of people did and a lot of people reported the performance issues and bugs (which were the same then as they were after release).
(plus, many if not all of the performance/graphics related problems seem to be readily identifiable thru industry standard profiling tools that would illustrate the flaws pretty clearly - there's no need to find patterns and acquire more data. This wasn't a niche group of players having these issues, this was roughly 70% and the QA team they have should easily have captured such problems - assuming they have QA)

Consoles are always behind the level of technology of PCs so that's why they delayed the consoles. If PCs have problems, then consoles are definitely going to have the same problems or worse.

Consoles were not scheduled to be delayed until well after the beginning of alpha when it became extremely apparent how screwed up the release version would be. If it was something planned, we would have heard about the delay long before the alpha and release dates were made known.

As far as the stall and blame tactics go, what a crock of crap. They haven't tried to blame it on anyone else, and not because the vast minority of players on the forums called them out on it. they need more data to see if there are any common denominators that could cause problems. They also have to work with Nvidia and AMD to get the drivers working if that's the issue, or even a small part of it.

Oh, and BTW, this forum software sucks.

They're certainly attempting to make it look like it's wasn't something they were well aware of even before alpha began which would shift the blame from them to "we had no way of knowing it would be this way with different hardware". They're not shifting the blame to any other individual, they're shifting it off of themselves. That's a crock of crap. They were aware of these major issues the entire time.

The drivers are fine. They're playing in dx11 - not some bleeding edge stuff - uncovering driver bugs because nobody has used the api the way they happen to be using it before. Plus we get these issues even in linux, which doesn't use the windows drivers. you dont need to make excuses for them, they pretty much burned any chance that any excuse would be believable.

 

[LoneTech]

in my case my xeon e3-1231 v3 is about 60% in odyssey- Only GPU is giving fire on 100% (gtx 970)

That 60% is often misleading. Your processor has hyperthreading, so when one thread is waiting for memory, it counts as 50% of one core - even when it's the system bottleneck. That thread could have counted as 100% if the core didn't have a second thread slot, but it still wouldn't show clearly as the core wouldn't be going into turbo clocks since it's waiting, not working. If your GPU is entirely busy, though, your system may be GPU bound.

 

[Morbad]

Just tested the suit tutorial with a 5GHz Zen 3 part (5800X custom PBO) and even with this setup there were only three areas where I was clearly completely GPU bound at 1440p Ultra with a 6800 XT:

• The power station, which is filled with fog or mist.
• The fire in the command center.
• The hangar after the ship lands at the surface starport at the very end.

Everywhere else was CPU bound to at least some extent. The first two are particle heavy, probably making fill rate the limiting factor.

This is also what the CPU utilization looked like with a 100ms polling rate:

Those spikes are smoothed out considerably by most reporting software at stock settings, which only poll about once per second. This conceals the true extent of utilization.