Overclock.net banner
Cancel
Create thread New Forums

[Official] NVIDIA RTX 5090 Owner's Club

Tags
5090 blackwell nvidia rtx
44 reading
Jump to Latest
1,141 - 1,160 of 18,210 Posts
#1,141 ·
lol a lot of self argumentations to buy a 5090 lol that for itself it’s a clear signal of this launch.
 
Save
Reply Quote Rep+ Rep+
#1,142 ·
Andre.X86 said:
I have a custom 900w cable for my Cosair SF1000L & 4090, BRING IT ON BABY.
Click to expand...
I'm wondering what is the watt threshold for the connector on the card itself.
 
Save
Reply Quote Rep+ Rep+
#1,144 · (Edited)
zhrooms said:
.. this was exhausting to compile, gave up half way, but I've got enough to make my point, maybe I'll fill in the rest at a later date.

Club386 (7590X3D + 64GB 6000C36)

3DMark - Steel Nomad
5090: 13,870 Graphics Score (+48.1%)
4090: 9,368 Graphics Score

3DMark - 4K DLSS + FG
5090: 200 FPS (+42.3%)
4090: 140 FPS

Alan Wake 2 - 4K DLSS + FG
5090: 244 FPS (+251.5%) with MFG 4x reaching full 240 FPS on 4K 240Hz
5090: 134 FPS (+38.1%)
4090: 97 FPS

Cyberpunk - 4K DLSS + FG
5090: 219 FPS (+267.1%) with MFG 4x almost reaching 240 FPS on 4K 240Hz
5090: 114 FPS (+39.0%)
4090: 82 FPS

Cowcotland (7800X3D + 32GB 6000CL30-36)

3DMark - Steel Nomad
5090: 14,086 Graphics Score (+50.8%)
4090: 9,339 Graphics Score

Alan Wake 2 - 4K PT
5090: 48 FPS (+65.5%)
4090: 29 FPS

Cyberpunk - 4K PT
5090: 28 FPS (+40.0%)
4090: 20 FPS

God of War - 4K
5090: 199 FPS (+42.1%)
4090: 140 FPS

Hogwarts Legacy - 4K
5090: 169 FPS (+45.7%)
4090: 116 FPS

Watch Dogs: Legion - 4K RT
5090: 89 FPS (+50.8%)
4090: 59 FPS

El Chapuzas Informático (285K + 48GB 8400C40)

Cyberpunk - 4K Ultra
5090: 125 FPS (+58.2%)
4090: 79 FPS

God of War - 4K Ultra
5090: 211 FPS (+38.8%)
4090: 152 FPS

Hogwarts Legacy - 4K Ultra
5090: 160 FPS (+39.1%)
4090: 115 FPS

The Witcher 3 - 4K Ultra
5090: 140 FPS (+42.9%)
4090: 98 FPS

Eteknix (9800X3D + 32GB 6000C30)

Cyberpunk - 4K Ultra
5090: 90 FPS (+40.6%)
4090: 64 FPS

Black Myth Wukong - 4K High + RT
5090: 50 FPS (+38.9%)
4090: 36 FPS

Eurogamer (9800X3D + 32GB 6000C30)

A Plague Tale Requiem - 4K Max (Highest 1%)
5090: 110.7 FPS (+35.1%)
4090: 81.9 FPS

Alan Wake 2 - 4K High (Highest 1%)
5090: 53.5 FPS (+37.5%)
4090: 38.9 FPS

Cyberpunk - 4K
5090: 95 FPS (+55.7%)
4090: 61 FPS

Cyberpunk - 4K Psycho
5090: 62.8 FPS (+39.9%)
4090: 44.9 FPS

Dying Light 2 - 4K High RT
5090: 79 FPS (+38.7%)
4090: 57 FPS

Hellblade 2 - 4K High (Highest 5%)
5090: 121.6 FPS (+34.0%)
4090: 90.7 FPS

GamersNexus (9800X3D + 32GB 6000C28 and 4090 Cybertank)

Dragon's Dogma 2 - 4K Max
5090: 132.5 FPS (+35.2%)
4090: 98 FPS

Cyberpunk - 4K Ultra
5090: 95.4 FPS (+50.2%)
4090: 63.5 FPS

Dying Light 2 - 4K High
5090: 126.4 FPS (+38.0%)
4090: 91.6 FPS

Resident Evil 4 - 4K Prio
5090: 206.7 FPS (+36.6%)
4090: 151.3 FPS

Black Myth Wukong - 4K High + FSR
5090: 88.3 FPS (+35.8%)
4090: 65.0 FPS

GDM (9800X3D + 48GB 6000C48)

3DMark Steel Nomad
5090: 14138 Graphics Score (+55.2%)
4090: 9111 Graphics Score

3DMark Time Spy Extreme
5090: 25551 Graphics Score (+35.2%)
4090: 18894 Graphics Score

3DMark Fire Strike Ultra
5090: 33115 Graphics Score (+34.6%)
4090: 24599 Graphics Score

FFXIV Benchmark - 4K
5090: 24424 Score (+35.0%)
4090: 18087 Score

Rainbow Six Siege - 4K
5090: 344 FPS (+33.9%)
4090: 257 FPS

Hardware & Co (9800X3D + 64GB 6000C30)

A Plague Tale Requiem - 4K RT + DLSS
5090: 216.9 FPS (+41.0%)
4090: 153.8 FPS

Black Myth Wukong - 4K RT
5090: 96.8 FPS (+36.3%)
4090: 71.0 FPS

Cyberpunk - 4K Psycho
5090: 68.4 FPS (+51.7%)
4090: 45.1 FPS

Dying Light 2 - 4K RT + DLSS
5090: 223.1 FPS (+42.4%)
4090: 156.7 FPS

F1 2024 - 4K RT + DLSS
5090: 233.5 FPS (+34.8%)
4090: 173.2 FPS

Ratchet and Clank Rift Apart - 4K RT + DLSS
5090: 213.3 FPS (+34.2%)
4090: 158.9 FPS

Total War Pharaoh - 4K Extreme
5090: 128.7 FPS (+35.0%)
4090: 95.3 FPS

Warhammer Space Marine II - 4K Ultra
5090: 128.1 FPS (+39.5%)
4090: 91.8 FPS

Hardware Unboxed (9800X3D + 32GB 6000C30)

A Plague Tale Requiem - 4K Ultra
5090: 128 FPS (+42.2%)
4090: 90 FPS

Black Myth Wukong - 4K Very High RT
5090: 34 FPS (+48%)
4090: 23 FPS

Dying Light 2 - 4K High
5090: 122 FPS (+40.2%)
4090: 87 FPS

God of War Ragnarök - 4K Ultra
5090: 195 FPS (+36.4%)
4090: 143 FPS

Marvel Rivals - 4K Ultra
5090: 101 FPS (+33%)
4090: 76 FPS

STALKER 2 - 4K Epic
5090: 71 FPS (+42%)
4090: 50 FPS

The Last of Us Part I - 4K Ultra
5090: 125 FPS (+40.4%)
4090: 89 FPS

TechPowerUp (9800X3D + 32GB 6200C28)

Counter-Strike 2 - 4K
5090: 347.9 FPS (+52.0%)
4090: 228.8 FPS

Cyberpunk - 4K
5090: 108.8 FPS (+46.4%)
4090: 74.3 FPS

Dragon Age Veilguard - 4K
5090: 110.7 FPS (+40.0%)
4090: 79.1 FPS

Elden Ring - 4K
5090: 184.7 FPS (+37.7%)
4090: 134.1 FPS

Ghost of Tsushima - 4K
5090: 129.1 FPS (+45.4%)
4090: 88.8 FPS

God of War Ragnarök - 4K
5090: 151.2 FPS (+34%)
4090: 112.9 FPS

Hogwarts Legacy - 4K
5090: 97.1 FPS (+41.5%)
4090: 68.6 FPS

Like a Dragon 8 - 4K
5090: 134.2 FPS (+50.3%)
4090: 89.3 FPS

No Rest for the Wicked - 4K
5090: 150.4 FPS (+40.6%)
4090: 107.0 FPS

Resident Evil 4 - 4K
5090: 156.8 FPS (+39.4%)
4090: 112.5 FPS

The Last of Us Part 1 - 4K
5090: 108.4 FPS (+42.6%)
4090: 76.0 FPS

The Witcher 3 - 4K
5090: 299.3 FPS (+40.0%)
4090: 213.8 FPS

Warhammer Space Marine 2 - 4K
5090: 125.4 FPS (+41.5%)
4090: 88.6 FPS

SUMMARY

3DMark - Steel Nomad
+55.2%

3DMark - DLSS
+42.3%

Alan Wake 2
+37.5%

Alan Wake 2 - Ray Tracing
+65.5%

A Plague Tale Requiem
+42.2%

Black Myth Wukong
+48%

Counter-Strike 2
+52%

Cyberpunk
+58.2%

Dying Light 2
+42.4%

Dragon Age Veilguard
+40%

Elden Ring
+37.7%

Ghost of Tsushima
+45.4%

God of War
+42.1%

Hogwarts Legacy
+45.7%

Like a Dragon 8
+50.3%

No Rest for the Wicked
+40.6%

Resident Evil 4
+39.4%

STALKER 2
+42%

The Last of Us Part 1
+42.6%

The Witcher 3
+42.9%

Warhammer Space Marine 2
+41.5%

Watch Dogs Legion
+50.8%

Out of these 20 games (hand picked to try and get rid of bottlenecked benchmarks and poorly conducted benchmarks)
+45.65%

If we include a few other games that "only" reached 33-34%, then that average gets closer to 40%.

It's safe to say that RTX 5090 at 575W is at least 40% faster than the RTX 4090 at 450W.

If we move to higher resolutions like 5120x2160 (4K 21:9) and 7680x2160 (4K 32:9), the percentage increase is highly likely to be up to 50%. I plan to test this myself as soon as I get my hands on a card.

Is it accurate to claim that the RTX 5090 is 50% faster than the RTX 4090? That's still up for debate. I'm about 70% sure it will reach that percentage at those higher resolutions, especially with DLSS at 5K and 4K 32:9, so I would probably say yes.

Is it 40% faster? Absolutely.

Is it 50% faster? More likely than not.
In 1080p? No. I don't care, I don't run such a low resolution.
In 1440p? No. I don't care, I don't run such a low resolution.
In 4K 16:9? No. I don't care, I don't run such a narrow aspect ratio.
In 4K 21:9? More likely than not, with the improved optimization when FG and DLSS is enabled.
In 4K 32:9? Almost definitely.

(And that's not mentioning DLSS 4 MFG 4x)

We still don't have a clear understanding of why some reviewers report significantly lower performance with the RTX 5090. For example, one review might show the RTX 5090 as being only 28% faster, while another, for the same game, claims it's 38% faster. I haven't come across any reviewer providing an explanation for these discrepancies—perhaps because they see no need to, and they certainly don't coordinate with each other. Given that most use similar hardware like the 9800X3D with 32GB of 6000C30 RAM, it's unlikely to be a hardware issue. Therefore, the most plausible explanation is flawed benchmarking methodology; many reviewers seem to lack expertise in this area, which I've come to expect. To get a comprehensive view, you really have to comb through numerous reviews, piecing together details here and there until the full picture emerges.
Click to expand...
"We still don't have a clear understanding of why some reviewers report significantly lower performance with the RTX 5090. For example, one review might show the RTX 5090 as being only 28% faster, while another, for the same game, claims it's 38% faster".

My guess is it's a combination of drivers, BIOS, and CPU's.

Red Text Screenshot Font Number
 
Save
Reply Quote Rep+ Rep+
#1,145 ·
zzztopzzz said:
"We still don't have a clear understanding of why some reviewers report significantly lower performance with the RTX 5090. For example, one review might show the RTX 5090 as being only 28% faster, while another, for the same game, claims it's 38% faster".

My guess is it's a combination of drivers, BIOS, and CPU's.

View attachment 2692425
Click to expand...
The Tom's Hardware review is truly bizarre - with the 9800x3d at times beating a 13900k, at times losing to it, and with the 4090 sometimes beating a 5090. Either their methodology is completely FUBAR, or what they've mentioned about driver overhead has some credence? Could it be that somehow these early drivers behave differently in intel systems vs amd?

Also - anyone else find it odd that there was almost ZERO overclocking performed? I've seen reviews from almost all of these publications before, and they explicitly have whole sections of their review dedicated to OC potential, performance, and conclusions. I have to assume this was part of the NDA/review guide for the FE? Hope that changes tomorrow with the AIB reviews.
 
#1,146 ·
zhrooms said:
.. this was exhausting to compile, gave up half way, but I've got enough to make my point, maybe I'll fill in the rest at a later date.

Club386 (7590X3D + 64GB 6000C36)

3DMark - Steel Nomad
5090: 13,870 Graphics Score (+48.1%)
4090: 9,368 Graphics Score

3DMark - 4K DLSS + FG
5090: 200 FPS (+42.3%)
4090: 140 FPS

Alan Wake 2 - 4K DLSS + FG
5090: 244 FPS (+251.5%) with MFG 4x reaching full 240 FPS on 4K 240Hz
5090: 134 FPS (+38.1%)
4090: 97 FPS

Cyberpunk - 4K DLSS + FG
5090: 219 FPS (+267.1%) with MFG 4x almost reaching 240 FPS on 4K 240Hz
5090: 114 FPS (+39.0%)
4090: 82 FPS

Cowcotland (7800X3D + 32GB 6000CL30-36)

3DMark - Steel Nomad
5090: 14,086 Graphics Score (+50.8%)
4090: 9,339 Graphics Score

Alan Wake 2 - 4K PT
5090: 48 FPS (+65.5%)
4090: 29 FPS

Cyberpunk - 4K PT
5090: 28 FPS (+40.0%)
4090: 20 FPS

God of War - 4K
5090: 199 FPS (+42.1%)
4090: 140 FPS

Hogwarts Legacy - 4K
5090: 169 FPS (+45.7%)
4090: 116 FPS

Watch Dogs: Legion - 4K RT
5090: 89 FPS (+50.8%)
4090: 59 FPS

El Chapuzas Informático (285K + 48GB 8400C40)

Cyberpunk - 4K Ultra
5090: 125 FPS (+58.2%)
4090: 79 FPS

God of War - 4K Ultra
5090: 211 FPS (+38.8%)
4090: 152 FPS

Hogwarts Legacy - 4K Ultra
5090: 160 FPS (+39.1%)
4090: 115 FPS

The Witcher 3 - 4K Ultra
5090: 140 FPS (+42.9%)
4090: 98 FPS

Eteknix (9800X3D + 32GB 6000C30)

Cyberpunk - 4K Ultra
5090: 90 FPS (+40.6%)
4090: 64 FPS

Black Myth Wukong - 4K High + RT
5090: 50 FPS (+38.9%)
4090: 36 FPS

Eurogamer (9800X3D + 32GB 6000C30)

A Plague Tale Requiem - 4K Max (Highest 1%)
5090: 110.7 FPS (+35.1%)
4090: 81.9 FPS

Alan Wake 2 - 4K High (Highest 1%)
5090: 53.5 FPS (+37.5%)
4090: 38.9 FPS

Cyberpunk - 4K
5090: 95 FPS (+55.7%)
4090: 61 FPS

Cyberpunk - 4K Psycho
5090: 62.8 FPS (+39.9%)
4090: 44.9 FPS

Dying Light 2 - 4K High RT
5090: 79 FPS (+38.7%)
4090: 57 FPS

Hellblade 2 - 4K High (Highest 5%)
5090: 121.6 FPS (+34.0%)
4090: 90.7 FPS

GamersNexus (9800X3D + 32GB 6000C28 and 4090 Cybertank)

Dragon's Dogma 2 - 4K Max
5090: 132.5 FPS (+35.2%)
4090: 98 FPS

Cyberpunk - 4K Ultra
5090: 95.4 FPS (+50.2%)
4090: 63.5 FPS

Dying Light 2 - 4K High
5090: 126.4 FPS (+38.0%)
4090: 91.6 FPS

Resident Evil 4 - 4K Prio
5090: 206.7 FPS (+36.6%)
4090: 151.3 FPS

Black Myth Wukong - 4K High + FSR
5090: 88.3 FPS (+35.8%)
4090: 65.0 FPS

GDM (9800X3D + 48GB 6000C48)

3DMark Steel Nomad
5090: 14138 Graphics Score (+55.2%)
4090: 9111 Graphics Score

3DMark Time Spy Extreme
5090: 25551 Graphics Score (+35.2%)
4090: 18894 Graphics Score

3DMark Fire Strike Ultra
5090: 33115 Graphics Score (+34.6%)
4090: 24599 Graphics Score

FFXIV Benchmark - 4K
5090: 24424 Score (+35.0%)
4090: 18087 Score

Rainbow Six Siege - 4K
5090: 344 FPS (+33.9%)
4090: 257 FPS

Hardware & Co (9800X3D + 64GB 6000C30)

A Plague Tale Requiem - 4K RT + DLSS
5090: 216.9 FPS (+41.0%)
4090: 153.8 FPS

Black Myth Wukong - 4K RT
5090: 96.8 FPS (+36.3%)
4090: 71.0 FPS

Cyberpunk - 4K Psycho
5090: 68.4 FPS (+51.7%)
4090: 45.1 FPS

Dying Light 2 - 4K RT + DLSS
5090: 223.1 FPS (+42.4%)
4090: 156.7 FPS

F1 2024 - 4K RT + DLSS
5090: 233.5 FPS (+34.8%)
4090: 173.2 FPS

Ratchet and Clank Rift Apart - 4K RT + DLSS
5090: 213.3 FPS (+34.2%)
4090: 158.9 FPS

Total War Pharaoh - 4K Extreme
5090: 128.7 FPS (+35.0%)
4090: 95.3 FPS

Warhammer Space Marine II - 4K Ultra
5090: 128.1 FPS (+39.5%)
4090: 91.8 FPS

Hardware Unboxed (9800X3D + 32GB 6000C30)

A Plague Tale Requiem - 4K Ultra
5090: 128 FPS (+42.2%)
4090: 90 FPS

Black Myth Wukong - 4K Very High RT
5090: 34 FPS (+48%)
4090: 23 FPS

Dying Light 2 - 4K High
5090: 122 FPS (+40.2%)
4090: 87 FPS

God of War Ragnarök - 4K Ultra
5090: 195 FPS (+36.4%)
4090: 143 FPS

Marvel Rivals - 4K Ultra
5090: 101 FPS (+33%)
4090: 76 FPS

STALKER 2 - 4K Epic
5090: 71 FPS (+42%)
4090: 50 FPS

The Last of Us Part I - 4K Ultra
5090: 125 FPS (+40.4%)
4090: 89 FPS

TechPowerUp (9800X3D + 32GB 6200C28)

Counter-Strike 2 - 4K
5090: 347.9 FPS (+52.0%)
4090: 228.8 FPS

Cyberpunk - 4K
5090: 108.8 FPS (+46.4%)
4090: 74.3 FPS

Dragon Age Veilguard - 4K
5090: 110.7 FPS (+40.0%)
4090: 79.1 FPS

Elden Ring - 4K
5090: 184.7 FPS (+37.7%)
4090: 134.1 FPS

Ghost of Tsushima - 4K
5090: 129.1 FPS (+45.4%)
4090: 88.8 FPS

God of War Ragnarök - 4K
5090: 151.2 FPS (+34%)
4090: 112.9 FPS

Hogwarts Legacy - 4K
5090: 97.1 FPS (+41.5%)
4090: 68.6 FPS

Like a Dragon 8 - 4K
5090: 134.2 FPS (+50.3%)
4090: 89.3 FPS

No Rest for the Wicked - 4K
5090: 150.4 FPS (+40.6%)
4090: 107.0 FPS

Resident Evil 4 - 4K
5090: 156.8 FPS (+39.4%)
4090: 112.5 FPS

The Last of Us Part 1 - 4K
5090: 108.4 FPS (+42.6%)
4090: 76.0 FPS

The Witcher 3 - 4K
5090: 299.3 FPS (+40.0%)
4090: 213.8 FPS

Warhammer Space Marine 2 - 4K
5090: 125.4 FPS (+41.5%)
4090: 88.6 FPS

SUMMARY

3DMark - Steel Nomad
+55.2%

3DMark - DLSS
+42.3%

Alan Wake 2
+37.5%

Alan Wake 2 - Ray Tracing
+65.5%

A Plague Tale Requiem
+42.2%

Black Myth Wukong
+48%

Counter-Strike 2
+52%

Cyberpunk
+58.2%

Dying Light 2
+42.4%

Dragon Age Veilguard
+40%

Elden Ring
+37.7%

Ghost of Tsushima
+45.4%

God of War
+42.1%

Hogwarts Legacy
+45.7%

Like a Dragon 8
+50.3%

No Rest for the Wicked
+40.6%

Resident Evil 4
+39.4%

STALKER 2
+42%

The Last of Us Part 1
+42.6%

The Witcher 3
+42.9%

Warhammer Space Marine 2
+41.5%

Watch Dogs Legion
+50.8%

Out of these 20 games (hand picked to try and get rid of bottlenecked benchmarks and poorly conducted benchmarks)
+45.65%

If we include a few other games that "only" reached 33-34%, then that average gets closer to 40%.

It's safe to say that RTX 5090 at 575W is at least 40% faster than the RTX 4090 at 450W.

If we move to higher resolutions like 5120x2160 (4K 21:9) and 7680x2160 (4K 32:9), the percentage increase is highly likely to be up to 50%. I plan to test this myself as soon as I get my hands on a card.

Is it accurate to claim that the RTX 5090 is 50% faster than the RTX 4090? That's still up for debate. I'm about 70% sure it will reach that percentage at those higher resolutions, especially with DLSS at 5K and 4K 32:9, so I would probably say yes.

Is it 40% faster? Absolutely.

Is it 50% faster? More likely than not.
In 1080p? No. I don't care, I don't run such a low resolution.
In 1440p? No. I don't care, I don't run such a low resolution.
In 4K 16:9? No. I don't care, I don't run such a narrow aspect ratio.
In 4K 21:9? More likely than not, with the improved optimization when FG and DLSS is enabled.
In 4K 32:9? Almost definitely.

(And that's not mentioning DLSS 4 MFG 4x)

We still don't have a clear understanding of why some reviewers report significantly lower performance with the RTX 5090. For example, one review might show the RTX 5090 as being only 28% faster, while another, for the same game, claims it's 38% faster. I haven't come across any reviewer providing an explanation for these discrepancies—perhaps because they see no need to, and they certainly don't coordinate with each other. Given that most use similar hardware like the 9800X3D with 32GB of 6000C30 RAM, it's unlikely to be a hardware issue. Therefore, the most plausible explanation is flawed benchmarking methodology; many reviewers seem to lack expertise in this area, which I've come to expect. To get a comprehensive view, you really have to comb through numerous reviews, piecing together details here and there until the full picture emerges.
Click to expand...
This is tremendously helpful. Just wanted to give you recognition for the amount of time and organization this took. Great stuff.
 
#1,147 ·
So are we saying derbauer doesn’t know how to do reviews either ? Because his results are in line with the others …
 
  • Rep+
Reactions: diogogmaio
Save
Reply Quote Rep+ Rep+
#1,148 ·
In for FE if I can score one of the handful Best Buy Canada will have nationwide on launch day. If I can't get one then I'll wait out 6090.

More AI training/inference workloads than gaming for me so the 32GB of memory is worth the upgrade.
 
#1,149 ·
Pedros said:
So are we saying derbauer doesn’t know how to do reviews either ? Because his results are in line with the others …
Click to expand...
or just that drivers are not good enough for all games yet because of the new architecture , for example the 7900XTX was losing to 6900XT on some old games aswell and that was fixed with drivers later on but yeah coming from nvidia that surprising
 
#1,150 ·
Anyone seen reviews when using a intel 200 series (15th gen).
Yes, not the best CPU around - but some of us (foolishly) upgraded to it.
It would also be interesting data @ 4k (anything under 4k is pointless IMO) - It would be interesting if AMDs new CPUs still rock the world at that point or it doesn't matter. (or maybe some of the strong points of the 200 series (ie, memory) help.
My gut tells me @ 4k it doesn't make a difference and any CPU 12th gen or better is about the same.
 
#1,151 ·
Very good review here. He highlights the bottleneck of the 5090 even on the 9800x3d and pci express scaling on some games is an issue small but there it is , best bet is to have a new motherboard with pci express gen 5.0 for the gpu.

 
#1,152 ·
MagmaTET said:
or just that drivers are not good enough for all games yet because of the new architecture , for example the 7900XTX was losing to 6900XT on some old games aswell and that was fixed with drivers later on but yeah coming from nvidia that surprising
Click to expand...
I think this board it is what it is. No need to sugar coat it. If you have the budget it’s the best out there.
But …
 
Save
Reply Quote Rep+ Rep+
#1,153 ·
chispy said:
Very good review here. He highlights the bottleneck of the 5090 even on the 9800x3d and pci express scaling on some games is an issue small but there it is , best bet is to have a new motherboard with pci express gen 5.0 for the gpu.

Click to expand...
For sure hardware canucks are amazing at gpu reviews ehehe …
 
  • Rep+
Reactions: chispy
Save
Reply Quote Rep+ Rep+
#1,154 ·
Looks like the 5090 FE is yet another card that doesn't always reach it's max PL before throttling. You can see that it is downclocking in the shot from der8auer's video at just 550w.

Text Screenshot Font Number Software
 
#1,155 ·
I'm really curious to see how the additional memory bandwidth handles DLDSR 2.25x situations at 4K res compared to the 4090 since I only play games with DLDSR active now. Should remove CPU bottleneck issues and possibly see higher performance bump in my use case scenario. Just praying for an OC bios, as well as active cooled backplate block from Optimus. Though I could go ghetto rigging and get my 6x ram block out and stick it on the back of a normal block like I did with the 3090...

Also not liking the latency increase with MFG but interesting how nobody seems to have shown latency changes using Reflex 2 Frame Warp yet since that could save the feature and make it more universally usable. I guess it wasn't available with the review drivers?

p.s. Since my card was one of the ones with limited memory OC capabilities, maxing at at a +800 mem offset, I think I'll automatically end up with a slightly bigger bump than you guys who could do already +1600-2000.
 
#1,156 ·
HyperMatrix said:
I'm really curious to see how the additional memory bandwidth handles DLDSR 2.25x situations at 4K res compared to the 4090 since I only play games with DLDSR active now. Should remove CPU bottleneck issues and possibly see higher performance bump in my use case scenario. Just praying for an OC bios, as well as active cooled backplate block from Optimus. Though I could go ghetto rigging and get my 6x ram block out and stick it on the back of a normal block like I did with the 3090...

Also not liking the latency increase with MFG but interesting how nobody seems to have shown latency changes using Reflex 2 Frame Warp yet since that could save the feature and make it more universally usable. I guess it wasn't available with the review drivers?
Click to expand...
Everything is very new. I think nvidia will be like amd and it’s fine wine ehehe
 
Save
Reply Quote Rep+ Rep+
#1,158 ·
HyperMatrix said:
I'm really curious to see how the additional memory bandwidth handles DLDSR 2.25x situations at 4K res compared to the 4090 since I only play games with DLDSR active now. Should remove CPU bottleneck issues and possibly see higher performance bump in my use case scenario. Just praying for an OC bios, as well as active cooled backplate block from Optimus. Though I could go ghetto rigging and get my 6x ram block out and stick it on the back of a normal block like I did with the 3090...

Also not liking the latency increase with MFG but interesting how nobody seems to have shown latency changes using Reflex 2 Frame Warp yet since that could save the feature and make it more universally usable. I guess it wasn't available with the review drivers?

p.s. Since my card was one of the ones with limited memory OC capabilities, maxing at at a +800 mem offset, I think I'll automatically end up with a slightly bigger bump than you guys who could do already +1600-2000.
Click to expand...
I didn’t see any reviews commenting on whether men oc preferred higher or lower temps. Active backplate on 4090 gddr6 was actually harmful to performance as the memory performance tanked at low temps.

GN made mention of high mem temps and made it sound like an area of concern, but also, like everyone else, included no OC details, so it’s unclear if it actually is a good or a bad thing that the mem runs hot.
 
#1,159 ·
bmagnien said:
I didn’t see any reviews commenting on whether men oc preferred higher or lower temps. Active backplate on 4090 gddr6 was actually harmful to performance as the memory performance tanked at low temps.

GN made mention of high mem temps and made it sound like an area of concern, but also, like everyone else, included no OC details, so it’s unclear if it actually is a good or a bad thing that the mem runs hot.
Click to expand...
I don’t believe that’s part of the review context … what kind of reviewers have you seen doing that on day one of a launch ?! Eheh
 
Save
Reply Quote Rep+ Rep+
#1,160 ·
Joe Flores said:
What's the temp limit for GDDR7? Anyone know yet
Click to expand...
...need to know the manufacturer (ie. Samsung, Micron etc) to look up spec sheets on GDDR7. I also wondered before if GDDR7 - like GDDR6X - needs a certain minimum temperature for optimal performance and does not like 'cold'.

...speaking of GDDR7, @alasdairvfr will like the 96 GB of it > here
 
1,141 - 1,160 of 18,210 Posts
About this Discussion
18K Replies
579 Participants
Last post:
synergon
Overclock.net
posts
28M
members
562K
Since
2004
The premier forum for overclocking experts and enthusiasts. Discuss hardware optimization, custom builds, benchmarking, cooling solutions, and pushing the boundaries of computing performance. From beginner guides to extreme overclocking, join our technical community to master system tuning.
Full Forum Listing
Explore Our Forums
Intel CPUs Hardware News NVIDIA PC Gaming Water Cooling