• 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.

Nvidia RTX 30XX |OT|

BluRayHiDef

Banned
Maybe windows is on some power saving mode, that cpu shouldn't be such a bottleneck. Maybe check that, update bios, try other games.
Edit
Didn't know it's from '14 that only supports DDR3 ram, there's significant hardware age difference, could be the time for you to upgrade the whole platform.
No, it supports DDR4 RAM.
 
i7-5820k @ 4GHz via boost.
You are probably a bit CPU limited. But I wouldn't think that much. Why not keep a stable OC though? I'm sure you could easily handle 4.2Ghz at a minimum. I have my i7-9700k at 4.5 Ghz and it's stable and has mad a huge difference over the "boost" clocks that Intel will do.
 

BluRayHiDef

Banned
I bought an entirely new computer halfway through the present life of my 2080Ti, and was astonished at the results - not only were my frames higher, but everything was just way smoother. A computer really is the sum of its parts.... memory, cpu, motherboard.... it all matters. You can only keep on buying new GPUs to a certain point, after which you’re wasting your money.
I guess I'll be getting a 5950x.
 

BluRayHiDef

Banned
Not necessary if your goal is 4k gaming, 3700x is good enough, different story if productivity work is involved.

Why would I not get a chip from the latest generation? It doesn't make sense to do otherwise if my intention is to maximize longevity.
 

Kenpachii

Member
Shoot, maybe you guys were right all along. Getting the RTX 3080 was perhaps a blessing in disguise since the additional money that I would have spent on an RTX 3090 can be used to get a new CPU (and motherboard), which I appear to need.

I just ran the benchmark for GTAV at the same settings that I used when my 1080 Ti was installed (max settings at 4K); the performance is better...but it's not locked at 60 frames per second (it's in the 40s and 50s for the most part), which may be because my CPU is too slow to convey instructions and data to the card.

However, perhaps I'm drawing conclusions prematurely; I have to run other benchmarks and redo the test run of Control.

Lower the resolution to 1080p if still not 60 fps your cpu is bottlenecking the game.
 

J3nga

Member
Why would I not get a chip from the latest generation? It doesn't make sense to do otherwise if my intention is to maximize longevity.
Didn't say you shouldn't, just not necessary for gaming. Longevity is under the question, am4 platform is nearing it's ending, in '22 new socket will be out along with DDR5 memory.
 

regawdless

Banned
Very happy so far. Great thermals and just brutal performance.


RPfa8BE.jpg



Kt6thZm.jpg


Will turn off that lightshow though, PC is closed and dark, so no need for flashy lights.
 

BluRayHiDef

Banned
Lower the resolution to 1080p if still not 60 fps your cpu is bottlenecking the game.
Didn't say you shouldn't, just not necessary for gaming. Longevity is under the question, am4 platform is nearing it's ending, in '22 new socket will be out along with DDR5 memory.
Try and lower your settings too.

My PC is currently rendering a comparison-benchmark video. Hopefully I have it up within the next thirty minutes.
 
I really hope that EVGA auto-notify queue really comes through and I can get a 3090 this year. I'm not even that fussed about playing Cyberpunk day 1 because I know it will need many patches to work out the bugs but I'm hoping I can eventually play it? In 2020 even? We'll see.
 

BluRayHiDef

Banned
Lower the resolution to 1080p if still not 60 fps your cpu is bottlenecking the game.
Didn't say you shouldn't, just not necessary for gaming. Longevity is under the question, am4 platform is nearing it's ending, in '22 new socket will be out along with DDR5 memory.
Try and lower your settings too.

Here are two videos of the same benchmark runs; one shows the data displayed in the upper left corner of the screen via MSI Afterburner and the other shows the frame rate displayed in the center of the screen via the GTAV application itself. Keep in mind that the videos are still processing.




 
So, the shop where i bought my EVGA (kabum.com.br) was supposed to ship it yesterday. But now they say their shipment didn't arrive or something.
Dammit :messenger_grinning_sweat:
 

BluRayHiDef

Banned
In addition, here's the backplate.

YB5QvJw.jpg




It's a 3080. I shortly thought about the 3090, but I'm mostly playing at 1440p. So it wasn't worth the premium for me.

Where'd you get it from? Did you get it in person or online? I wanted this card (under PNY's branding, since I'm in the United States), but my order of it was canceled by Office Depot because PNY didn't receive the stock that they expected.
 

regawdless

Banned
Where'd you get it from? Did you get it in person or online? I wanted this card (under PNY's branding, since I'm in the United States), but my order of it was canceled by Office Depot because PNY didn't receive the stock that they expected.

Germany, from an online retailer. Ordered shortly after it went online, one day after launch. It was then postponed indefinitely, the retailer had no idea when it would be available. And now, it suddenly arrived.
 

regawdless

Banned
Finally played my first raytraced game ever yesterday before going to sleep. Quake 2 RTX of course.
It's kinda weird to look at a huge display in game, just to see the damn realistic reflection of an enemy approaching from behind.

Overclocking wise, there seems to be some headroom on my 3080, easily boosts to 2100mhz. Haven't tried to go higher yet. Don't know if it'll be worth it though, because the actual performance gains aren't that big. May just leave it at 109 power limit and the rest at stock settings. If a game runs at 130 or 133 fps isn't really significant.
 

Kenpachii

Member
Here are two videos of the same benchmark runs; one shows the data displayed in the upper left corner of the screen via MSI Afterburner and the other shows the frame rate displayed in the center of the screen via the GTAV application itself. Keep in mind that the videos are still processing.






I have no clue what the numbers at the top mean as u didn't name them. I got the feeling that 90+% number is your power % usage and not GPU usage. So u are better of naming the numbers.

Basically open msi afterburner and do the following: ( got a gif that showcases this below )

1) propperty's ( some chain wheel under fan speed mostly placed )
2) monitoring tab
3) scroll down until u find something called GPU Usage click on it
4) enable ( below window ) "show in on-screen display"
5) then enable "overwrite group name" and fill in on the right 1080ti usage or 3080 usage

I can basically name here more.

Fast example of how to name stuff ( this memory as the program can only record 7 seconds and i cant find gpu usage fast enough. ).


e58681e5567a95303208e101883ccb66.gif


However don't pay attention towards the 2x names as i have stuff on the same graph line which makes it less clunky in your game, u can do that by making both override group names the same.

bac0b9601da089317bc712ed35b5d2b6.png


Anyway.

I tested your settings with my 1080ti and 9900k at both stock speeds and i was sitting at 24 fps as lowest with mostly around 30.
Tested it with the video i linked with reduced settings lowest i saw it go was 49 fps with mostly around 60. so that's basically double the performance gain.

Frankly gta 5 settings are absolutely brutal at max settings from what i remember. stuff like grass / msaa can destroy any gpu / cpu combination on ultra. This is why barely anybody really maxes all the settings in even youtube benchmark video's. Watched a few and all of them reduce the settings.
 
Last edited:

BluRayHiDef

Banned
I have no clue what the numbers at the top mean as u didn't name them. I got the feeling that 90+% number is your power % usage and not GPU usage. So u are better of naming the numbers.

Basically open msi afterburner and do the following: ( got a gif that showcases this below )

1) propperty's ( some chain wheel under fan speed mostly placed )
2) monitoring tab
3) scroll down until u find something called GPU Usage click on it
4) enable ( below window ) "show in on-screen display"
5) then enable "overwrite group name" and fill in on the right 1080ti usage or 3080 usage

I can basically name here more.

Fast example of how to name stuff ( this memory as the program can only record 7 seconds and i cant find gpu usage fast enough. ).


e58681e5567a95303208e101883ccb66.gif


However don't pay attention towards the 2x names as i have stuff on the same graph line which makes it less clunky in your game, u can do that by making both override group names the same.

bac0b9601da089317bc712ed35b5d2b6.png


Anyway.

I tested your settings with my 1080ti and 9900k at both stock speeds and i was sitting at 24 fps as lowest with mostly around 30.
Tested it with the video i linked with reduced settings lowest i saw it go was 49 fps with mostly around 60. so that's basically double the performance gain.

Frankly gta 5 settings are absolutely brutal at max settings from what i remember. stuff like grass / msaa can destroy any gpu / cpu combination on ultra. This is why barely anybody really maxes all the settings in even youtube benchmark video's. Watched a few and all of them reduce the settings.

Well, I honestly don't want to reinstall the 1080Ti just to redo the benchmark with labeled data. That would be tedious.

With default settings at 4K, I can maintain 60 frames per second on the 1080Ti. However, I chose to max out the settings for the benchmark in order to compare the two cards under the most stressful conditions.

So, you don't think that my CPU is bottlenecking the 3080?
 

BluRayHiDef

Banned
LOL, I'm back at Micro Center to see if they have any 3090 FTW3s despite getting a 3080 FTW3 yesterday (since the former is what I really want)...and there's a line of 6 people here. I'm surprised because every other day that I've come here at this time, there has been only one other person here by the time that I arrived or sometimes none.

EDIT:

They had only two 3080s, so I've left. Also, they have a policy by which you can buy only one of each chip per month (e.g. one 3080 per month and one 3090 per month, regardless of brand). Hence, one person can buy one 3080 AND one 3090 within a month's time.

a74aeWl.jpg
 
Last edited:

McHuj

Member
I can't believe Austin, Tx doesn't have a Microcenter. Otherwise, I would have gone to get in line. We have a shitty Fry's.
 

BluRayHiDef

Banned
They had only two 3080s, so I've left. Also, they have a policy by which you can buy only one of each chip per month (e.g. one 3080 per month and one 3090 per month, regardless of brand). Hence, one person can buy one 3080 AND one 3090 within a month's time.
 

BluRayHiDef

Banned
Eagerly awaiting the ROG Strix 3090 :p

Glad I got an order in for it


It's barely faster than the non-OC RTX 3090s from other brands; check out Guru3D's review here. I'm not trying to burst your bubble, especially considering that I still want an RTX 3090 despite having an RTX 3080; however, I'm just surprised since it needs an additional 50 watts or more of power for a negligibly higher level of performance.

index.php


index.php


index.php


index.php


index.php
 
Last edited:

BluRayHiDef

Banned
I can't believe Austin, Tx doesn't have a Microcenter. Otherwise, I would have gone to get in line. We have a shitty Fry's.

Well, at least you have a brick-and-mortar store. Some people don't have that and subsequently must rely exclusively on online retailers.
 
Last edited:

BluRayHiDef

Banned
I'm mainly getting it for the 2 HDMI ports and the Asus brand. :lollipop_grinning:

In the video they show a 7% improvement with manual overclock in Death Stranding. <.<

To each there own. I'm not trying to burst your bubble, especially considering that I still want an RTX 3090 despite having an RTX 3080; however, I'm just surprised since it needs an additional 50 watts or more of power for a negligibly higher level of performance.
 

Ulysses 31

Gold Member
To each there own. I'm not trying to burst your bubble, especially considering that I still want an RTX 3090 despite having an RTX 3080; however, I'm just surprised since it needs an additional 50 watts or more of power for a negligibly higher level of performance.
With the 2 HDMIs I can hook up the LG CX and the Q950TS at the same time without having to pull cables. I'm not expecting noticeable performance gains and I'm not planning to OC.

I pre-ordered at amazon for 1650€ so I'm not paying scalper prices for it either like I did for the MSI Ventus 3090. :lollipop_grinning:
 

BluRayHiDef

Banned
With the 2 HDMIs I can hook up the LG CX and the Q950TS at the same time without having to pull cables. I'm not expecting noticeable performance gains and I'm not planning to OC.

I pre-ordered at amazon for 1650€ so I'm not paying scalper prices for it either like I did for the MSI Ventus 3090. :lollipop_grinning:
A DisplayPort to HDMI cable is not an option? I'm not trying to convince you to reverse your purchase but am curious as to whether you considered this option.
 

Ulysses 31

Gold Member
A DisplayPort to HDMI cable is not an option? I'm not trying to convince you to reverse your purchase but am curious as to whether you considered this option.
Could be but I've read about loss of VRR and periodic flickering or other oddities with adapters. If you know of proven solid adapter then I could take a look at it. :messenger_sunglasses:
 

BluRayHiDef

Banned
Could be but I've read about loss of VRR and periodic flickering or other oddities with adapters. If you know of proven solid adapter then I could take a look at it. :messenger_sunglasses:
That's understandable. I originally wanted an Asus ROG Strix RTX 3090 as well, particularly because I thought that it would significantly outperform other models of the RTX 3090 and because I like the LED indicators on its power-connector ports.
 

Kenpachii

Member
Well, I honestly don't want to reinstall the 1080Ti just to redo the benchmark with labeled data. That would be tedious.

With default settings at 4K, I can maintain 60 frames per second on the 1080Ti. However, I chose to max out the settings for the benchmark in order to compare the two cards under the most stressful conditions.

So, you don't think that my CPU is bottlenecking the 3080?

It reads out any gpu in your system. no need to swap cards.

The problem with default settings is it reduces the demand on the CPU also so u don't know at higher settings if your CPU will bottleneck or not. If you want to know that then do some tests or use msi afterburner correctly with labels so u know if your cpu or gpu bottlenecks at settings u like to play basically.

Don't want to deal with that, just lower the resolution on ultra settings and your gpu demand will go down and u will see if your get higher fps on the low end. if that happens then its gpu bottlenecked like i mentioned before and not your cpu at those settings, if the fps stays the same on the low end its cpu bottleneck. Either reduce settings and if u can't because u need higher settings upgrade.

Anyway good luck.
 
Last edited:

Rentahamster

Rodent Whores
BluRayHiDef BluRayHiDef

If you want 4K at 60 fps, just dial in the settings that allow you to play there. Getting into the weeds about whether or not your CPU is a bottleneck with benchmark after benchmark is pointless if you're not going to get a new one.

Overclock your current CPU and call it a day.
 

BluRayHiDef

Banned
It reads out any gpu in your system. no need to swap cards.

The problem with default settings is it reduces the demand on the CPU also so u don't know at higher settings if your CPU will bottleneck or not. If you want to know that then do some tests or use msi afterburner correctly with labels so u know if your cpu or gpu bottlenecks at settings u like to play basically.

Don't want to deal with that, just lower the resolution on ultra settings and your gpu demand will go down and u will see if your get higher fps on the low end. if that happens then its gpu bottlenecked like i mentioned before and not your cpu at those settings, if the fps stays the same on the low end its cpu bottleneck. Either reduce settings and if u can't because u need higher settings upgrade.

Anyway good luck.

YouTube is currently processing a video of the GTAV benchmark running on my RTX 3080; the settings are the same as in the other videos but anti-aliasing is set to 4X MSAA instead of 8X. Also, the data has been labeled this time.
 
Last edited:
BluRayHiDef BluRayHiDef

If you want 4K at 60 fps, just dial in the settings that allow you to play there. Getting into the weeds about whether or not your CPU is a bottleneck with benchmark after benchmark is pointless if you're not going to get a new one.

Overclock your current CPU and call it a day.

Just dial down the settings? Gross. No way. Don't do it, BluRayHiDef.
 

benno

Member
Anyone with a 30x0 GPU play dragon's dogma on pc still? I would like to know what frames you get @4k with my enb. I could only hit 45fps with a 2080ti.
 

Chiggs

Gold Member
I think March of 2021 is when I'll be able to successfully purchase a 3090 without having to resort to black magic.
 
Top Bottom