RTX 4080 Peers or close I need help

[Wanted]

What kind of frames are you seeing in the game, mainly like tower FPS. I have 4080 super and 7800x3d and I am barely only getting 170fps in the tower at 1440p max settings. I say "only" because I'm only at like 40% GPU utilization, and 20% for CPU. Is this normal? I saw a YouTube video of a guy with a 4080 + 15-13600k getting near 240fps with the same settings.

[DCR]

i get 200 on the smallest single fan 4060, on a rx 7800 now it goes to about 900-1000 fps in orbit, around 700 on 4060.. Resolution matters little in d2, its almost the same on both 1080p/4k here. 3733c14 with twcl12 / 4000c16 optimized timings. Ram and motherboard settings matters more then your gpu.

[Chillin]

Originally posted by DCR:

i get 200 on the smallest single fan 4060, on a rx 7800 now it goes to about 900-1000 fps in orbit, around 700 on 4060.. Resolution matters little in d2, its almost the same on both 1080p/4k here. 3733c14 with twcl12 / 4000c16 optimized timings. Ram and motherboard settings matters more then your gpu.

what motherboard settings.

[DCR]

Originally posted by Bing_Chilling:

Originally posted by DCR:

i get 200 on the smallest single fan 4060, on a rx 7800 now it goes to about 900-1000 fps in orbit, around 700 on 4060.. Resolution matters little in d2, its almost the same on both 1080p/4k here. 3733c14 with twcl12 / 4000c16 optimized timings. Ram and motherboard settings matters more then your gpu.

what motherboard settings.

Everything in relation to lower latency, efficiency, voltage consumption/limits. So quite a lot. You can easily configure an intel system to draw max power inefficiently and continously draw max power and thus also continously cause system to throttle.

edit: The most important thing is to check before and after performance, lowering a setting isn't always better real world performance, same is true for pushing voltage higher.. For both amd and intel it's ideal to figure out pressure point for tRCDRD, on ddr4 where stability occurs at 13 is typically the best, most systems don't push below 14 read. And its no point in increasing tRP if it means increasing tRC past tRAS+tRP. Doesn't mean that can't be stable, but in most cases you're giving up performance just for having a lower value on a single setting. The only real world use cases would typically be specific benchmarks like superpi which loves low TRP. In all other cases basically, 100% pointless. Same should be true for ddr5. For gaming, and practically everything - almost, lowest tRC is ideal. If u find the right iod voltage value for the lowest tRCDRD you can do, you have then also find the correct value for the highest bandwidth the ram chips can do. Doesnt mean you have to use that exact value, but it will be pretty close. In the end stability will depend on what stresstests are done. It's entirely possible to get 100% stresstestapp test even running for hours without fail, and system will still start producing WHEA typical errors, as in hardware errors, just because system can't cope with running so fast/overheating, but ram could still run just fine, motherboard/cpu imc just can't handle stressing it over time. This is especially true on ddr4 pushing past 3800mhz. With newer dimm's it's usually not a problem, less voltage that drains out, higher efficiency silicon and all that. In the end everything has to work well together.

[Wanted]

Originally posted by DCR:

i get 200 on the smallest single fan 4060, on a rx 7800 now it goes to about 900-1000 fps in orbit, around 700 on 4060.. Resolution matters little in d2, its almost the same on both 1080p/4k here. 3733c14 with twcl12 / 4000c16 optimized timings. Ram and motherboard settings matters more then your gpu.

I mean I am running DDR5 6000mhz cl 30, cant remember the other timings. I get around 3-500 in orbit, dont really care about orbit lol. Trying to figure out why my gpu isnt being utilized while in game

[DCR]

6000c30 is quite a bit of from ddr5 potential, should be aiming for 6400/6600 cl28 or 30. And even then you wouldn't push into ddr4 potential today outside of bandwidth. D2 isn't optimized for newer gpu's nor capable of fully utilizing them to their max potential on that end. Same is true for cpu's. Hence why going all in on lowest latency and optimizing cpu core for fastest turbo core is, as per usual, ideal for gaming. Anyway ddr5 is still relatively young, it's going to increasive massively over the next iterations. And even then i'd still hope somebody would just create a board with HBM onboard or a chip with it rather. But things are happening, there's some new cpu's archs out there but not mainstream and way too expensive for it now anyway. Orbit fps is just an indicator. Anyway quite horrible for your setup. I just have an 5800x3d but great mobo. Hoping for a 10gb nic itx am5 board on next-gen, that be worth upgrading to. DDR5 today has most benefit on going with just 2 dimms and not 4. And preferrably 2 dimm slot board for increased efficiency in latency.

[Wanted]

Originally posted by DCR:

6000c30 is quite a bit of from ddr5 potential, should be aiming for 6400/6600 cl28 or 30. And even then you wouldn't push into ddr4 potential today outside of bandwidth. D2 isn't optimized for newer gpu's nor capable of fully utilizing them to their max potential on that end. Same is true for cpu's. Hence why going all in on lowest latency and optimizing cpu core for fastest turbo core is, as per usual, ideal for gaming. Anyway ddr5 is still relatively young, it's going to increasive massively over the next iterations. And even then i'd still hope somebody would just create a board with HBM onboard or a chip with it rather. But things are happening, there's some new cpu's archs out there but not mainstream and way too expensive for it now anyway. Orbit fps is just an indicator. Anyway quite horrible for your setup. I just have an 5800x3d but great mobo. Hoping for a 10gb nic itx am5 board on next-gen, that be worth upgrading to. DDR5 today has most benefit on going with just 2 dimms and not 4. And preferrably 2 dimm slot board for increased efficiency in latency.

6000mhz and CL 30 is what is recommended for most AMD chips. I cant imagine me not having faster Ram is what is holding my GPU down from being used. I have no problems in other game and my gpu is being used like it should be. So does this come down to the game being ♥♥♥♥♥♥? If that is so, how is someone with a similar setup getting 70 more fps than me, also he is only running 5600mhz DDR5 ram

[DCR]

do you have stuff like shadowplay, reflex, or other things utilizing video enabled, either ingame or otherwise? I'd start with disabling everything, set everything to max, and disable one by one to see what makes a difference.

[Wanted]

Originally posted by DCR:

do you have stuff like shadowplay, reflex, or other things utilizing video enabled, either ingame or otherwise? I'd start with disabling everything, set everything to max, and disable one by one to see what makes a difference.

Nah i have reflex in the background, but nothing active while playing

[DCR]

Well for me difference between default and optimized bios is about 20-80 fps all depending on what im running. I can't troubleshoot for you, you got to start somewhere.

[sometimes headshoot]

I have a 4080 Super + 7800X3D and I am getting worse frames on high compared to my wife who has a 3080ti + 5800X3D, did you ever figure this out?

[Chalchiutlicue]

The game just uses an old engine from halo 2 times.
We won't get any technology upgrade in destiny 2.
Management decision.

[RetroJohn92]

Originally posted by Wanted:

What kind of frames are you seeing in the game, mainly like tower FPS. I have 4080 super and 7800x3d and I am barely only getting 170fps in the tower at 1440p max settings. I say "only" because I'm only at like 40% GPU utilization, and 20% for CPU. Is this normal? I saw a YouTube video of a guy with a 4080 + 15-13600k getting near 240fps with the same settings.

i have the same cpu and gpu, but my frames are capped by my current monitor's refresh rate, but i have to ask what you have to make it use 40%, because mine tries uses almost 100% all the time at 1080p high settings and in higher level content i see it hit 68-70c at times. it does drop all the way back down quickly when not in game or in orbit

edit: apparently the render resolution at 200 and turning back to default fixed the utilization/ heating issue

[illuser]

I'm at a capped 200 fps on a 4080 and a i9. I rarely have framerate issues unless I'm running a number of other processes at the same time.

[_galaxy]

Originally posted by Wanted:

What kind of frames are you seeing in the game, mainly like tower FPS. I have 4080 super and 7800x3d and I am barely only getting 170fps in the tower at 1440p max settings. I say "only" because I'm only at like 40% GPU utilization, and 20% for CPU. Is this normal? I saw a YouTube video of a guy with a 4080 + 15-13600k getting near 240fps with the same settings.

rtx 2060 ryzen 7 1700
graphics: medium (anti-aliasing smaa)
vsync: 60fps
storage: hdd


Never stuttered (as long as I don't have too many tabs open in my browser), with these exceptions:
Inside crucible because of the anticheat. I need to restart my pc and have my browser closed to not feel stutters in pvp.
Excision (still manageable, no stutters on second phase)


system is debloated with unshitter
I use chris tools to set updates to only minimum and nvinstall to install drivers
didn't check utilization but I think it uses mostly the GPU? This is a console game so cpu shouldn't be affected too much however bungie goes crazy sometimes with stuff like excision


It's a old game so I'd check if X3D processor does anything to the game since it's like 10 years old at this point I know you can disable certain cpu features in BIOS