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

MSI Radeon RX 6950 XT Gaming X Trio

W1zzard

Administrator
Staff member
Joined
May 14, 2004
Messages
27,971 (3.71/day)
Processor Ryzen 7 5700X
Memory 48 GB
Video Card(s) RTX 4080
Storage 2x HDD RAID 1, 3x M.2 NVMe
Display(s) 30" 2560x1600 + 19" 1280x1024
Software Windows 10 64-bit
MSI's Radeon RX 6950 XT Gaming X Trio is the fastest RX 6950 XT we tested today, making it the fastest Radeon we ever tested. It beats the NVIDIA RTX 3090 and almost matches the much more expensive RTX 3090 Ti, with very little noise and good temperatures.

Show full review
 
Joined
May 2, 2017
Messages
7,762 (2.78/day)
Location
Back in Norway
System Name Hotbox
Processor AMD Ryzen 7 5800X, 110/95/110, PBO +150Mhz, CO -7,-7,-20(x6),
Motherboard ASRock Phantom Gaming B550 ITX/ax
Cooling LOBO + Laing DDC 1T Plus PWM + Corsair XR5 280mm + 2x Arctic P14
Memory 32GB G.Skill FlareX 3200c14 @3800c15
Video Card(s) PowerColor Radeon 6900XT Liquid Devil Ultimate, UC@2250MHz max @~200W
Storage 2TB Adata SX8200 Pro
Display(s) Dell U2711 main, AOC 24P2C secondary
Case SSUPD Meshlicious
Audio Device(s) Optoma Nuforce μDAC 3
Power Supply Corsair SF750 Platinum
Mouse Logitech G603
Keyboard Keychron K3/Cooler Master MasterKeys Pro M w/DSA profile caps
Software Windows 10 Pro
So, what in the sweaty pits of hell is going on with the power draw of this card? This is ridiculous. 70W more than competing SKUs? Screw that.
 
Joined
Feb 20, 2022
Messages
175 (0.17/day)
System Name Custom Watercooled
Processor 10900k 5.1GHz SSE 5.0GHz AVX
Motherboard Asus Maximus XIII hero z590
Cooling XSPC Raystorm Pro, XSPC D5 Vario, EK Water Blocks EK-CoolStream XE 360 (Triple Fan) Radiator
Memory Team Group 8Pack RIPPED Edition TDPPD416G3600HC14CDC01 @ DDR4-4000 CL15 Dual Rank 4x8GB (32GB)
Video Card(s) KFA2 GeForce RTX 3080 Ti SG 1-Click OC 12GB LHR GDDR6X PCI-Express Graphics Card
Storage WD Blue SN550 1TB NVME M.2 2280 PCIe Gen3 Solid State Drive (WDS100T2B0C)
Display(s) LG 3D TV 32LW450U-ZB and Samsung U28D590
Case Full Tower Case
Audio Device(s) ROG SupremeFX 7.1 Surround Sound High Definition Audio CODEC ALC4082, ESS SABRE9018Q2C DAC/AMP
Power Supply Corsair AX1000 Titanium 80 Plus Titanium Power Supply
Mouse Logitech G502SE
Keyboard Logitech Y-BP62a
Software Windows 11 Pro
Benchmark Scores https://valid.x86.fr/2rdbdl https://www.3dmark.com/spy/27927340 https://ibb.co/YjQFw5t
Most of the games benchmarked are all the ones known to perform well on AMD cards. Also to many are AMD featured Games.

AMD Featured Games examples.
HALO INFINITE
God of War
Far Cry 6
DEATHLOOP
World War Z: Aftermath
Resident Evil Village
Warframe
The Riftbreaker
GODFALL
World of Warcraft: Shadowlands
Assassin’s Creed Valhalla
Horizon Zero Dawn Complete Edition for PC
Dirt 5
Terminator: Resistance
Resident Evil 3
Monster Hunter World: Iceborne
The Outer Worlds
Tom Clancy’s Ghost Recon Breakpoint
Gears 5
Borderlands 3
Red Dead Redemption 2
Forza Horizon 4

Most games in the benhmark are on this list and others are known AMD good performing games.

Also DXR or ray tracing games are tested in lagacy mode, i.e. raster. Metro Exodus Ehnanced Edition is ignored and replaced with Metro Exodus which is then tested in raster mode. As we all know Metro Exodus was upgraded to Metro Exodus Enhanced Edition for free and that does not support raster graphics. Thus you have to go out of your way to run its as a raster game, ignoring all the upgrades and free upgrade the Enhanced Edition.

We can see the issue here with the game picked.



Take out Metro Exodus Enhanced Edition



Now at 4k the 6900xt vs 3080 12GB



Note the games that go AMD's way in this review graph above and are expected to have this result with any AMD card. Most will be AMD featured games.
Assassin’s Creed Valhalla
Battlefield V
Borderlands 3
DOOM Eternal
F1 2021
Far Cry 6
Forza Horizon 5
HALO INFINITE
Hitman 3
Metro Exodus raster. This is a ray tracing show case title. Was updated for free to Metro Exodus Enhanced Edition and gets path tracing. AMD card performance drop towards a 3070 ti being faster than a 6900xt.
Red Dead Redemption 2

AMD Featured
Deathloop
God of War

NVidia benchmarks
Cyberpunk 2077
Total War: Warhammer III
Metro Exodus if you run it as ray tracing. Afterall this is a DXR game, raster is legacy. Metro Exodus Enhanced Edition is DXR only and was a free upgrade to owners of Metro Exodus. It has a massive NVidia performance lead because it is a pathtraced game.

Odd results
Dying Light 2 as we can see the 3080 12GB beats the 6900xt in this benchmark on another website at 4k by 16%.
Guardians of the Galaxy results are different here with NVidia winning. 3080 10GB beats a 6900xt. Game support RT and DLSS and its not used.
Resident Evil Village 6900xt beats a 3090 but cant on another website beat the 3080 12GB.
Watch Dogs Legion

Pointless benchmarks because FPS does not matter. The game is an RPG with light graphics today or something out a very long time ago. These games are used to control the result.
Civilization VI
Divinity Original Sin II
The Witcher 3: Wild Hunt

Benchmarks that are new to me.
Days Gone
Elden Ring

Pointless benchmarks because raster performance is not the point of the games graphics. Ignores DLSS. If it supports DXR/dlss it should be run in DXR/dlss. As this is what you get a 3080 10GB or 12GB/3080ti/3090/3090 ti or 6800xt/6900xt or 6950xt card for.
Control
Metro Exodus
Watch Dogs Legion
F1 2020
Guardians of the Galaxy
DOOM Eternal
Far Cry 6
Resident Evil Village
Watch Dogs: Legion
Deethloop
Cyberpunk 2077

Which company has poor DXR performance and why would treating DXR titles as raster games do for their performance? If most of the games in the sample support DXR way not run them at maximum setting and show the result? There is no logical reason to treat DXR support differently if most of the sample size are DXR supporting games.

Now from the information I have I could imply that the 6950xt will be a top performing GPU. Remember before were the 6900xt = 3090 but it cant really match a 3080 12GB. Now I am going to do the same for the 6950xt, it is faster than a 3090 but slower than a 3090 ti. You could say its almost as fast as RTX 3090 Ti. Afterall stating it above the 3090 ti would beg investigation. Now I have just read the conclusion of the review and started laughing.
 
Last edited:
Joined
Feb 20, 2019
Messages
8,341 (3.91/day)
System Name Bragging Rights
Processor Atom Z3735F 1.33GHz
Motherboard It has no markings but it's green
Cooling No, it's a 2.2W processor
Memory 2GB DDR3L-1333
Video Card(s) Gen7 Intel HD (4EU @ 311MHz)
Storage 32GB eMMC and 128GB Sandisk Extreme U3
Display(s) 10" IPS 1280x800 60Hz
Case Veddha T2
Audio Device(s) Apparently, yes
Power Supply Samsung 18W 5V fast-charger
Mouse MX Anywhere 2
Keyboard Logitech MX Keys (not Cherry MX at all)
VR HMD Samsung Oddyssey, not that I'd plug it into this though....
Software W10 21H1, barely
Benchmark Scores I once clocked a Celeron-300A to 564MHz on an Abit BE6 and it scored over 9000.
Almost 10% improvement over the 6900XT is impressive, but is that because the results in the chart are from the reference 6900XT which struggled severely with cooling and power limits compared to the partner models?

Given that most evidence points at the 6900XT not being bandwidth limited (1080p improvements over the 6900XT are similar to the 4K improvements) I was honestly expecting closer to a 3-5% improvement over the 6900XT.

AMD say three things:
  • 4% clock bump is close to 4% more performance, at a cost - obviously, and that cost is 70W.
  • 12% mem bump that *shouldn't* improve performance in any meaningful way, especially at 1080p.
  • ??% magic sauce in the drivers, which one could read as being an intentional sabotage of the older 6900XT by denying it improvements. This is the same silicon, after all, so why would improvements in drivers for the 6950XT not also apply to the 6900XT?
 

ARF

Joined
Jan 28, 2020
Messages
4,670 (2.61/day)
Location
Ex-usa | slava the trolls
This looks the fastest RX 6950 XT 16 GB.

Does anyone know why its performance is so strong at 1920x1080? Something makes the overall system overturn the CPU bottleneck?

1652197783372.png

1652197857905.png

1652197887336.png

1652197754980.png


The performance is all over the place. It is like there is a problem in the drivers, or simply the game optimisation doesn't recognise properly the architecture of the GPU?

1652197958933.png


1652197935880.png
 
Joined
Feb 20, 2022
Messages
175 (0.17/day)
System Name Custom Watercooled
Processor 10900k 5.1GHz SSE 5.0GHz AVX
Motherboard Asus Maximus XIII hero z590
Cooling XSPC Raystorm Pro, XSPC D5 Vario, EK Water Blocks EK-CoolStream XE 360 (Triple Fan) Radiator
Memory Team Group 8Pack RIPPED Edition TDPPD416G3600HC14CDC01 @ DDR4-4000 CL15 Dual Rank 4x8GB (32GB)
Video Card(s) KFA2 GeForce RTX 3080 Ti SG 1-Click OC 12GB LHR GDDR6X PCI-Express Graphics Card
Storage WD Blue SN550 1TB NVME M.2 2280 PCIe Gen3 Solid State Drive (WDS100T2B0C)
Display(s) LG 3D TV 32LW450U-ZB and Samsung U28D590
Case Full Tower Case
Audio Device(s) ROG SupremeFX 7.1 Surround Sound High Definition Audio CODEC ALC4082, ESS SABRE9018Q2C DAC/AMP
Power Supply Corsair AX1000 Titanium 80 Plus Titanium Power Supply
Mouse Logitech G502SE
Keyboard Logitech Y-BP62a
Software Windows 11 Pro
Benchmark Scores https://valid.x86.fr/2rdbdl https://www.3dmark.com/spy/27927340 https://ibb.co/YjQFw5t
Almost 10% improvement over the 6900XT is impressive, but is that because the results in the chart are from the reference 6900XT which struggled severely with cooling and power limits compared to the partner models?

Given that most evidence points at the 6900XT not being bandwidth limited (1080p improvements over the 6900XT are similar to the 4K improvements) I was honestly expecting closer to a 3-5% improvement over the 6900XT.

AMD say three things:
  • 4% clock bump is close to 4% more performance, at a cost - obviously, and that cost is 70W.
  • 12% mem bump that *shouldn't* improve performance in any meaningful way, especially at 1080p.
  • ??% magic sauce in the drivers, which one could read as being an intentional sabotage of the older 6900XT by denying it improvements. This is the same silicon, after all, so why would improvements in drivers for the 6950XT not also apply to the 6900XT?
The tested games are the reason for the performance. That and ignoring ray tracing in the games that support it. Its not like this site is different from many others. Nearly half of the games support RT but RT is treated as special off category. Then almost all the games are the ones that result in better performance on AMD cards. DX12u makes Ray Tracing mainstream in games, yet benchmark review ignores every game that supports it. When most of the games you are testing in a review have RT support at maximum settings. You should be doing maximum settings and poor performers be damned. We all know what happens to AMD cards in RT, so we all know this ignoring and treating RT as a special off category is to protect one company from bad reviews. Its not like all the RT games will lead to a nvidia win, its just amd will be cut down to size and the true performance wont likely match nvidia. The 6950xt wont be better than a 3080 12GB in RT and will never match a 3080 ti.

At some point all the new games will be RT, I get the feeling RT like DLSS wont be a thing in reviews until AMD get it together. Have performance matching or beating NVidia. I know 100% that if nvidia was behind in RT and AMD had DLSS. Every review would show this and every game would run at maximum setting with maximum ray tracing.

It funny how RT games were AMD equal nvidia get accepted but Metro Exodus Enhanced Edition gets special treatment or completely ignored. AMD are pushing raster performance and reviews are focusing on raster and its 100% deliberate. 100% biased.

Basically DX has now moved to Ray Tracing, raster performance is legacy. The focus is DXR performance for the now/future and games that support this feature should be front and center. So taking a game that has Ray Tracing and running it in raster so one manufacture can look more competitive. Well it is what it is and its all reviews that are doing this. Picking every game that runs well on AMD is not how you do a review. Ignoring Ray Tracing in every game that has it and making it another special category. This is 100% designed to make AMD cards look on power with nvidia.

People are going to turn on ray tracing in current AAA games and find their amd DXR performance is complete garbage. That they should have got a 3080 12GB. That no amount of overclocking makes their AMD card faster in DXR so that it matches a 3080 12GB or above. God forbid they startup a pathtraced game like Metro Exodus Enhanced Edition. Or try maximum RT in Cyberpunk 2077. The counter for this has been spot the DXR, you cant notice it and its not there. 100% is Don't Look Up.

You can see it with the sea of garbage about FSR 1 how DLSS was more acceptable now FSR 1 beat it. People started to accept DLSS because AMD had FSR. Still no one is running DLSS in benchmarks as the result, its close to native and does look better is some small ways. Not to ignore its problems.
 
Joined
Apr 30, 2020
Messages
999 (0.59/day)
System Name S.L.I + RTX research rig
Processor Ryzen 7 5800X 3D.
Motherboard MSI MEG ACE X570
Cooling Corsair H150i Cappellx
Memory Corsair Vengeance pro RGB 3200mhz 32Gbs
Video Card(s) 2x Dell RTX 2080 Ti in S.L.I
Storage Western digital Sata 6.0 SDD 500gb + fanxiang S660 4TB PCIe 4.0 NVMe M.2
Display(s) HP X24i
Case Corsair 7000D Airflow
Power Supply EVGA G+1600watts
Mouse Corsair Scimitar
Keyboard Cosair K55 Pro RGB
  • ??% magic sauce in the drivers, which one could read as being an intentional sabotage of the older 6900XT by denying it improvements. This is the same silicon, after all, so why would improvements in drivers for the 6950XT not also apply to the 6900XT?
It could have been due to very minor error in manufacturing that original RX 6900xt wasn't hitting its true full potential too. Maybe they found out what was happening and fixed it.
like say the original was hitting 95% performance of what they hit in labs and said well we can still sell it. Later then went back and seen they would 98% with the manufacturing process maturing
The Raytracing Benches sort of show this card has a minor increase over the original that's just not clock speed and power from what it looks like to me.
 
Last edited:
Joined
Feb 20, 2022
Messages
175 (0.17/day)
System Name Custom Watercooled
Processor 10900k 5.1GHz SSE 5.0GHz AVX
Motherboard Asus Maximus XIII hero z590
Cooling XSPC Raystorm Pro, XSPC D5 Vario, EK Water Blocks EK-CoolStream XE 360 (Triple Fan) Radiator
Memory Team Group 8Pack RIPPED Edition TDPPD416G3600HC14CDC01 @ DDR4-4000 CL15 Dual Rank 4x8GB (32GB)
Video Card(s) KFA2 GeForce RTX 3080 Ti SG 1-Click OC 12GB LHR GDDR6X PCI-Express Graphics Card
Storage WD Blue SN550 1TB NVME M.2 2280 PCIe Gen3 Solid State Drive (WDS100T2B0C)
Display(s) LG 3D TV 32LW450U-ZB and Samsung U28D590
Case Full Tower Case
Audio Device(s) ROG SupremeFX 7.1 Surround Sound High Definition Audio CODEC ALC4082, ESS SABRE9018Q2C DAC/AMP
Power Supply Corsair AX1000 Titanium 80 Plus Titanium Power Supply
Mouse Logitech G502SE
Keyboard Logitech Y-BP62a
Software Windows 11 Pro
Benchmark Scores https://valid.x86.fr/2rdbdl https://www.3dmark.com/spy/27927340 https://ibb.co/YjQFw5t
The card is excellent. If only can be sold for $800, would be a bestseller..
But this is the real performance and the 6950xt should be priced at that level. Basically all raster games at 4k are over 60fps but some titles like Cyberpunk 2077 are not in these reviews. Turn on DXR and the performance in the latest titles become clear.

Ray Tracing benchmarks here cant post them here. 6950xt behind the 3080 10GB at all resolutions.

If the benchmarks showed the truth of gaming with DXR which is 11 games in the review. The 6950xt would look less like a 3090 performance wise. Just because raster as lagecy is available does not mean that DXR should be ignored. Given the gap between NVidia and AMD is so large. Ignoring DXR has only one reason. Changing image quality so one manufacture can look better performance wise. The games should be tested only in DXR mode if supported. Raster performance should take a back seat and treated as lagecy support. Not DXR taking a back seat, its more or less replaced raster at this point. DXR performance has not changed. The 6950xt is still behind the 3080.

When you balance both sizes of the coin, raster and DXR AMD cards are not a good buy at their current prices. A focus on raster lets AMD charge the earth for cards nowhere near the performance of nvidia.
 
Joined
May 2, 2017
Messages
7,762 (2.78/day)
Location
Back in Norway
System Name Hotbox
Processor AMD Ryzen 7 5800X, 110/95/110, PBO +150Mhz, CO -7,-7,-20(x6),
Motherboard ASRock Phantom Gaming B550 ITX/ax
Cooling LOBO + Laing DDC 1T Plus PWM + Corsair XR5 280mm + 2x Arctic P14
Memory 32GB G.Skill FlareX 3200c14 @3800c15
Video Card(s) PowerColor Radeon 6900XT Liquid Devil Ultimate, UC@2250MHz max @~200W
Storage 2TB Adata SX8200 Pro
Display(s) Dell U2711 main, AOC 24P2C secondary
Case SSUPD Meshlicious
Audio Device(s) Optoma Nuforce μDAC 3
Power Supply Corsair SF750 Platinum
Mouse Logitech G603
Keyboard Keychron K3/Cooler Master MasterKeys Pro M w/DSA profile caps
Software Windows 10 Pro
But this is the real performance and the 6950xt should be priced at that level. Basically all raster games at 4k are over 60fps but some titles like Cyberpunk 2077 are not in these reviews. Turn on DXR and the performance in the latest titles become clear.

Ray Tracing benchmarks here cant post them here. 6950xt behind the 3080 10GB at all resolutions.

If the benchmarks showed the truth of gaming with DXR which is 11 games in the review. The 6950xt would look less like a 3090 performance wise. Just because raster as lagecy is available does not mean that DXR should be ignored. Given the gap between NVidia and AMD is so large. Ignoring DXR has only one reason. Changing image quality so one manufacture can look better performance wise. The games should be tested only in DXR mode if supported. Raster performance should take a back seat and treated as lagecy support. Not DXR taking a back seat, its more or less replaced raster at this point. DXR performance has not changed. The 6950xt is still behind the 3080.

When you balance both sizes of the coin, raster and DXR AMD cards are not a good buy at their current prices. A focus on raster lets AMD charge the earth for cards nowhere near the performance of nvidia.
I think your perspective here is a bit... off. Yes, RT is absolutely relevant. But has it eclipsed rasterization, rendering it (no pun intended) a legacy option? Not even remotely close. RT is still a niche thing, relegated to select high budget AAA games and certain showcases/toys (Quake, Minecraft). Outside of that, rasterized graphics are still massively dominant (and still the basis for essentially all RT games too), and will most likely continue to be so for the foreseeable future, even if RT use will likely increase noticeably.

You're also mischaracterizing quite a few things here: The main benchmark suite of TPU is a rasterization test suite, because up until quite recently, rasterization was the only graphics mode available, and many GPUs still don't support it. A list full of "DNF" entries is not a useful performance comparison, and would unfairly skew comparisons (or render summary graphs useless). Until the two modes reach some form of parity, keeping them separate is thus reasonable. Your claims that DXR has "more or less replaced raster at this point" is just nowhere near reality. And given the relative scarcity of RT titles, it also makes sense to test a smaller subset of those. Unless I'm mixing something up you also said that Metro Exodus EE was RT only, which is just plain untrue. Having saved it for when I got an RT capable GPU and then switching between modes quite a bit while playing, I'm quite familiar with it supporting both RT and rasterization. Also, nobody is ignoring DXR - there's a whole test suite dedicated to testing it! Just because the reviewer leaves it to the reader to judge just how important it is to them - as it should be with all features! - doesn't mean it's being ignored.

It's generally and widely established that Nvidia's Ampere cards are overall stronger at 2160p and in RT than AMD's, while AMD is generally stronger at lower resolutions and rasterization. If you have a problem with a review presenting that balance so that a possible buyer can make an informed decision based on their own desires and needs, then you're just opposed to letting people judge for themselves which performance results matter the most to them. Luckily most reviewers don't share that attitude.
 

Cheeseball

Not a Potato
Supporter
Joined
Jan 2, 2009
Messages
2,045 (0.35/day)
Location
Pittsburgh, PA
System Name Titan
Processor AMD Ryzen™ 7 7950X3D
Motherboard ASRock X870 Taichi Lite
Cooling Thermalright Phantom Spirit 120 EVO CPU
Memory TEAMGROUP T-Force Delta RGB 2x16GB DDR5-6000 CL30
Video Card(s) ASRock Radeon RX 7900 XTX 24 GB GDDR6 (MBA)
Storage Crucial T500 2TB x 3
Display(s) LG 32GS95UE-B, ASUS ROG Swift OLED (PG27AQDP), LG C4 42" (OLED42C4PUA)
Case Cooler Master QUBE 500 Flatpack Macaron
Audio Device(s) Kanto Audio YU2 and SUB8 Desktop Speakers and Subwoofer, Cloud Alpha Wireless
Power Supply Corsair SF1000
Mouse Logitech Pro Superlight 2 (White), G303 Shroud Edition
Keyboard Keychron K2 HE Wireless / 8BitDo Retro Mechanical Keyboard (N Edition) / NuPhy Air75 v2
VR HMD Meta Quest 3 512GB
Software Windows 11 Pro 64-bit 24H2 Build 26100.2605
It's simple really.

The RX 6900 XT and 6950 XT (128 ROPs) have more rasterization operators than the RTX 3090 and RTX 3090 Ti (112 ROPs), thats why it can challenge or even beat the latter in many titles.

It's generally and widely established that Nvidia's Ampere cards are overall stronger at 2160p and in RT than AMD's, while AMD is generally stronger at lower resolutions and rasterization.

This is the truth. ^^^

If you're playing games that are more rasterization-oriented (not ray-tracing heavy), don't need more than 16 GB of VRAM and are under 3840x2160p/5120x1440p, then go with the RX 6900 XT/6950 XT.

TBH, a RX 6900 XT at $950 is a pretty good deal nowadays.
 
Joined
Feb 20, 2019
Messages
8,341 (3.91/day)
System Name Bragging Rights
Processor Atom Z3735F 1.33GHz
Motherboard It has no markings but it's green
Cooling No, it's a 2.2W processor
Memory 2GB DDR3L-1333
Video Card(s) Gen7 Intel HD (4EU @ 311MHz)
Storage 32GB eMMC and 128GB Sandisk Extreme U3
Display(s) 10" IPS 1280x800 60Hz
Case Veddha T2
Audio Device(s) Apparently, yes
Power Supply Samsung 18W 5V fast-charger
Mouse MX Anywhere 2
Keyboard Logitech MX Keys (not Cherry MX at all)
VR HMD Samsung Oddyssey, not that I'd plug it into this though....
Software W10 21H1, barely
Benchmark Scores I once clocked a Celeron-300A to 564MHz on an Abit BE6 and it scored over 9000.
But this is the real performance and the 6950xt should be priced at that level. Basically all raster games at 4k are over 60fps but some titles like Cyberpunk 2077 are not in these reviews. Turn on DXR and the performance in the latest titles become clear.

Ray Tracing benchmarks here cant post them here. 6950xt behind the 3080 10GB at all resolutions.

If the benchmarks showed the truth of gaming with DXR which is 11 games in the review. The 6950xt would look less like a 3090 performance wise. Just because raster as lagecy is available does not mean that DXR should be ignored. Given the gap between NVidia and AMD is so large. Ignoring DXR has only one reason. Changing image quality so one manufacture can look better performance wise. The games should be tested only in DXR mode if supported. Raster performance should take a back seat and treated as lagecy support. Not DXR taking a back seat, its more or less replaced raster at this point. DXR performance has not changed. The 6950xt is still behind the 3080.

When you balance both sizes of the coin, raster and DXR AMD cards are not a good buy at their current prices. A focus on raster lets AMD charge the earth for cards nowhere near the performance of nvidia.
Many other sites and popular Youtube channels also downplay RTX/DXR though - not just TPU.
You'll also find many RTX-card owners (like myself) admitting that they prefer to disable raytracing because high-framerates and crisp native resolution are preferable to a lower framerate and softer, upscaled image. The biggest problem with raytracing, even on Ampere, is that it's too damn expensive for what are subjectively minor image quality improvements a lot of the time.

If you are truly in the market for DXR gaming, then Nvidia is your best option but you'd need to have been living under a rock to not know that by now. AMD's DXR hardware isn't even as good as Nvidia's 2000-series and if you genuinely want to enable raytracing in current games that you need not just an Ampere card, but a high-end one (3070Ti and up). CP2077 may be the most demanding RTX titleright now and my poor little 3060Ti simply can't do it. 1440p medium + RT medium + DLSS balanced doesn't even average 60fps, it's like 55-70fps but most of the time is in the high 50s. Disable RT and I can run 1440p high + DLSS quality at 80-100fps. RT medium settings aren't even that nice to look at - they're grainy, imprecise and then upscaled a lot with DLSS balanced. I know SSAO a fake raster shadows/AO aren't perfect but compared to RT medium I think I prefer them - they're cleaner, sharper, and the vast improvements in image quality, graphics settings, and framerate more than make up for it.

My personal opinion (and you're entitled to disagree) is that if a game makes heavy use of multiple DXR aspects, not even a 3090Ti can really pull it off without making horrible sacrifices.
 
Joined
Feb 20, 2022
Messages
175 (0.17/day)
System Name Custom Watercooled
Processor 10900k 5.1GHz SSE 5.0GHz AVX
Motherboard Asus Maximus XIII hero z590
Cooling XSPC Raystorm Pro, XSPC D5 Vario, EK Water Blocks EK-CoolStream XE 360 (Triple Fan) Radiator
Memory Team Group 8Pack RIPPED Edition TDPPD416G3600HC14CDC01 @ DDR4-4000 CL15 Dual Rank 4x8GB (32GB)
Video Card(s) KFA2 GeForce RTX 3080 Ti SG 1-Click OC 12GB LHR GDDR6X PCI-Express Graphics Card
Storage WD Blue SN550 1TB NVME M.2 2280 PCIe Gen3 Solid State Drive (WDS100T2B0C)
Display(s) LG 3D TV 32LW450U-ZB and Samsung U28D590
Case Full Tower Case
Audio Device(s) ROG SupremeFX 7.1 Surround Sound High Definition Audio CODEC ALC4082, ESS SABRE9018Q2C DAC/AMP
Power Supply Corsair AX1000 Titanium 80 Plus Titanium Power Supply
Mouse Logitech G502SE
Keyboard Logitech Y-BP62a
Software Windows 11 Pro
Benchmark Scores https://valid.x86.fr/2rdbdl https://www.3dmark.com/spy/27927340 https://ibb.co/YjQFw5t
Many other sites and popular Youtube channels also downplay RTX/DXR though - not just TPU.
You'll also find many RTX-card owners (like myself) admitting that they prefer to disable raytracing because high-framerates and crisp native resolution are preferable to a lower framerate and softer, upscaled image. The biggest problem with raytracing, even on Ampere, is that it's too damn expensive for what are subjectively minor image quality improve
ments a lot of the time.

If you are truly in the market for DXR gaming, then Nvidia is your best option but you'd need to have been living under a rock to not know that by now. AMD's DXR hardware isn't even as good as Nvidia's 2000-series and if you genuinely want to enable raytracing in current games that you need not just an Ampere card, but a high-end one (3070Ti and up). CP2077 may be the most demanding RTX titleright now and my poor little 3060Ti simply can't do it. 1440p medium + RT medium + DLSS balanced doesn't even average 60fps, it's like 55-70fps but most of the time is in the high 50s. Disable RT and I can run 1440p high + DLSS quality at 80-100fps. RT medium settings aren't even that nice to look at - they're grainy, imprecise and then upscaled a lot with DLSS balanced. I know SSAO a fake raster shadows/AO aren't perfect but compared to RT medium I think I prefer them - they're cleaner, sharper, and the vast improvements in image quality, graphics settings, and framerate more than make up for it.

My personal opinion (and you're entitled to disagree) is that if a game makes heavy use of multiple DXR aspects, not even a 3090Ti can really pull it off without making horrible sacrifices.
Thats the thing about Ray Tracing Unreal Engine 5 can do it with Lumen and they use TSR for upscaling. Every game I like is under development using UE5.

This are what the software will do once its ready.


Games can all run as fully RT. Stop the difference between Full RT and just RTGI.


All the AAA games are going to be fully Ray Traced or look dated. Ignoring the Ray Tracing feature is pointless when approx. 11 of the games you test support DXR. If you just want to play with raster graphics, then get a mid range GPU. Stick to 1440p. A top end gpu, the main feature is RT performance. All the top end gpus, for the most part are overkill for raster games. You get >100fps at most resolutions. You dont need more than 8GB of vram.
 
Joined
Feb 20, 2019
Messages
8,341 (3.91/day)
System Name Bragging Rights
Processor Atom Z3735F 1.33GHz
Motherboard It has no markings but it's green
Cooling No, it's a 2.2W processor
Memory 2GB DDR3L-1333
Video Card(s) Gen7 Intel HD (4EU @ 311MHz)
Storage 32GB eMMC and 128GB Sandisk Extreme U3
Display(s) 10" IPS 1280x800 60Hz
Case Veddha T2
Audio Device(s) Apparently, yes
Power Supply Samsung 18W 5V fast-charger
Mouse MX Anywhere 2
Keyboard Logitech MX Keys (not Cherry MX at all)
VR HMD Samsung Oddyssey, not that I'd plug it into this though....
Software W10 21H1, barely
Benchmark Scores I once clocked a Celeron-300A to 564MHz on an Abit BE6 and it scored over 9000.
Thats the thing about Ray Tracing Unreal Engine 5 can do it with Lumen and they use TSR for upscaling. Every game I like is under development using UE5.

This are what the software will do once its ready.


Games can all run as fully RT. Stop the difference between Full RT and just RTGI.


All the AAA games are going to be fully Ray Traced or look dated. Ignoring the Ray Tracing feature is pointless when approx. 11 of the games you test support DXR. If you just want to play with raster graphics, then get a mid range GPU. Stick to 1440p. A top end gpu, the main feature is RT performance. All the top end gpus, for the most part are overkill for raster games. You get >100fps at most resolutions. You dont need more than 8GB of vram.
While you're not wrong, the full-DXR games we have right now don't look dramatically different to the baked lighting games of yore. Metro Exodus looks pretty damn similar between PS4 and PS5, one being raytraced, the other not.

Yes, in future, AAA games will likely be DXR-only and will replicate GI, shadows, and reflections well enough that we'll be seeing games that look like those UE5 demos. But we're not there yet and by the time the first real wave of those arrives on the scene, the Nvidia RTX4000-series and AMD RX7000-series will be on the way out. UE5 has only just left beta, and outside of test scenes we're nowhere near getting those carefully-curated visuals yet. By the time RTX 4080 is launched, games from the 2000-series launch are finally running acceptably with their extremely limited DXR features. Making the most of UE5 in AAA games is half a decade away, if I was forced to bet on a date.

You're right to say that in the future, when a significant portion of games people play are DXR-exclusive, sites should focus more on DXR performance. That is the future. We are talking about reviews happening now.
 
Joined
Apr 6, 2019
Messages
164 (0.08/day)
Most of the games benchmarked are all the ones known to perform well on AMD cards. Also to many are AMD featured Games.

AMD Featured Games examples.
HALO INFINITE
God of War
Far Cry 6
DEATHLOOP
World War Z: Aftermath
Resident Evil Village
Warframe
The Riftbreaker
GODFALL
World of Warcraft: Shadowlands
Assassin’s Creed Valhalla
Horizon Zero Dawn Complete Edition for PC
Dirt 5
Terminator: Resistance
Resident Evil 3
Monster Hunter World: Iceborne
The Outer Worlds
Tom Clancy’s Ghost Recon Breakpoint
Gears 5
Borderlands 3
Red Dead Redemption 2
Forza Horizon 4

Most games in the benhmark are on this list and others are known AMD good performing games.

Also DXR or ray tracing games are tested in lagacy mode, i.e. raster. Metro Exodus Ehnanced Edition is ignored and replaced with Metro Exodus which is then tested in raster mode. As we all know Metro Exodus was upgraded to Metro Exodus Enhanced Edition for free and that does not support raster graphics. Thus you have to go out of your way to run its as a raster game, ignoring all the upgrades and free upgrade the Enhanced Edition.

We can see the issue here with the game picked.



Take out Metro Exodus Enhanced Edition



Now at 4k the 6900xt vs 3080 12GB



Note the games that go AMD's way in this review graph above and are expected to have this result with any AMD card. Most will be AMD featured games.
Assassin’s Creed Valhalla
Battlefield V
Borderlands 3
DOOM Eternal
F1 2021
Far Cry 6
Forza Horizon 5
HALO INFINITE
Hitman 3
Metro Exodus raster. This is a ray tracing show case title. Was updated for free to Metro Exodus Enhanced Edition and gets path tracing. AMD card performance drop towards a 3070 ti being faster than a 6900xt.
Red Dead Redemption 2

AMD Featured
Deathloop
God of War

NVidia benchmarks
Cyberpunk 2077
Total War: Warhammer III
Metro Exodus if you run it as ray tracing. Afterall this is a DXR game, raster is legacy. Metro Exodus Enhanced Edition is DXR only and was a free upgrade to owners of Metro Exodus. It has a massive NVidia performance lead because it is a pathtraced game.

Odd results
Dying Light 2 as we can see the 3080 12GB beats the 6900xt in this benchmark on another website at 4k by 16%.
Guardians of the Galaxy results are different here with NVidia winning. 3080 10GB beats a 6900xt. Game support RT and DLSS and its not used.
Resident Evil Village 6900xt beats a 3090 but cant on another website beat the 3080 12GB.
Watch Dogs Legion

Pointless benchmarks because FPS does not matter. The game is an RPG with light graphics today or something out a very long time ago. These games are used to control the result.
Civilization VI
Divinity Original Sin II
The Witcher 3: Wild Hunt

Benchmarks that are new to me.
Days Gone
Elden Ring

Pointless benchmarks because raster performance is not the point of the games graphics. Ignores DLSS. If it supports DXR/dlss it should be run in DXR/dlss. As this is what you get a 3080 10GB or 12GB/3080ti/3090/3090 ti or 6800xt/6900xt or 6950xt card for.
Control
Metro Exodus
Watch Dogs Legion
F1 2020
Guardians of the Galaxy
DOOM Eternal
Far Cry 6
Resident Evil Village
Watch Dogs: Legion
Deethloop
Cyberpunk 2077

Which company has poor DXR performance and why would treating DXR titles as raster games do for their performance? If most of the games in the sample support DXR way not run them at maximum setting and show the result? There is no logical reason to treat DXR support differently if most of the sample size are DXR supporting games.

Now from the information I have I could imply that the 6950xt will be a top performing GPU. Remember before were the 6900xt = 3090 but it cant really match a 3080 12GB. Now I am going to do the same for the 6950xt, it is faster than a 3090 but slower than a 3090 ti. You could say its almost as fast as RTX 3090 Ti. Afterall stating it above the 3090 ti would beg investigation. Now I have just read the conclusion of the review and started laughing.
Hahaha, wow. Most of that is fanboy nonsense.
 
Joined
Mar 1, 2021
Messages
496 (0.36/day)
Location
Germany
System Name Homebase
Processor Ryzen 5 5600
Motherboard Gigabyte Aorus X570S UD
Cooling Scythe Mugen 5 RGB
Memory 2*16 Kingston Fury DDR4-3600 double ranked
Video Card(s) AMD Radeon RX 6800 16 GB
Storage 1*512 WD Red SN700, 1*2TB Curcial P5, 1*2TB Sandisk Plus (TLC), 1*14TB Toshiba MG
Display(s) Philips E-line 275E1S
Case Fractal Design Torrent Compact
Power Supply Corsair RM850 2019
Mouse Sharkoon Sharkforce Pro
Keyboard Fujitsu KB955
There are some “rumors” out there that MSI changed something on the inline filtering to reduce coil whine. Are there any hints that this might be true?
 
Joined
Jun 5, 2021
Messages
284 (0.22/day)
4000 series will get a massive l2 cache.. ampere is bandwidth starved at 1080p and 1440p.. infinity cache is a game changer for amd the 6950xt is well fed
 
Joined
May 2, 2017
Messages
7,762 (2.78/day)
Location
Back in Norway
System Name Hotbox
Processor AMD Ryzen 7 5800X, 110/95/110, PBO +150Mhz, CO -7,-7,-20(x6),
Motherboard ASRock Phantom Gaming B550 ITX/ax
Cooling LOBO + Laing DDC 1T Plus PWM + Corsair XR5 280mm + 2x Arctic P14
Memory 32GB G.Skill FlareX 3200c14 @3800c15
Video Card(s) PowerColor Radeon 6900XT Liquid Devil Ultimate, UC@2250MHz max @~200W
Storage 2TB Adata SX8200 Pro
Display(s) Dell U2711 main, AOC 24P2C secondary
Case SSUPD Meshlicious
Audio Device(s) Optoma Nuforce μDAC 3
Power Supply Corsair SF750 Platinum
Mouse Logitech G603
Keyboard Keychron K3/Cooler Master MasterKeys Pro M w/DSA profile caps
Software Windows 10 Pro
Thats the thing about Ray Tracing Unreal Engine 5 can do it with Lumen and they use TSR for upscaling. Every game I like is under development using UE5.

This are what the software will do once its ready.


Games can all run as fully RT. Stop the difference between Full RT and just RTGI.


All the AAA games are going to be fully Ray Traced or look dated. Ignoring the Ray Tracing feature is pointless when approx. 11 of the games you test support DXR. If you just want to play with raster graphics, then get a mid range GPU. Stick to 1440p. A top end gpu, the main feature is RT performance. All the top end gpus, for the most part are overkill for raster games. You get >100fps at most resolutions. You dont need more than 8GB of vram.
You understand that what you're drawing up is in the quite distant future, right? I mean, that path tracing demo isn't even close to real time. It looks amazing, but it would render at sub-1fps on the fastest GPUs today. So unless we get a >60x increase in RT performance with the next generation, we're not getting games looking like that any time soon. UE5 Path Tracing isn't meant to be real time, isn't, and won't be.

The rest of what you're saying are just vague platitudes with no time frame. Nobody is disputing that games can run as fully RT - we're just saying that neither hardware nor software are anywhere near where they need to be for that to happen.

You're also fundamentally ignoring market realities and how game developers are reliant on sales to survive. Despite existing for two generations now, RT-enabled GPUs are still overall quite rare. 8 out of the top 20 GPUs on the most recent Steam Survey have RTRT support, representing a grand total of <20% of GPUs in use. Of course the total number is higher than that - I only counted the top 20 - but even if we're quite generous, no more than 33% of GPUs actively used for gaming on Steam have RTRT support. There are extremely few game developers who have the financial security to make a game that is only playable at all for 1/3rd of their potential customer base. That just isn't economically feasible. For that reason alone, rasterized graphics will be dominant for quite a few years still. We might start seeing RT-only AAA titles in ... oh, half a decade? Maybe more? That is of course also largely dependent on consoles, and whether current-gen consoles with their install bases in the tens of millions are able to run these RT-only titles. If not, then they will also have rasterized modes. Nobody can afford to exclude that large a proportion of their customer base.
 
Joined
May 31, 2016
Messages
4,443 (1.42/day)
Location
Currently Norway
System Name Bro2
Processor Ryzen 5800X
Motherboard Gigabyte X570 Aorus Elite
Cooling Corsair h115i pro rgb
Memory 32GB G.Skill Flare X 3200 CL14 @3800Mhz CL16
Video Card(s) Powercolor 6900 XT Red Devil 1.1v@2400Mhz
Storage M.2 Samsung 970 Evo Plus 500MB/ Samsung 860 Evo 1TB
Display(s) LG 27UD69 UHD / LG 27GN950
Case Fractal Design G
Audio Device(s) Realtec 5.1
Power Supply Seasonic 750W GOLD
Mouse Logitech G402
Keyboard Logitech slim
Software Windows 10 64 bit
Thats the thing about Ray Tracing Unreal Engine 5 can do it with Lumen and they use TSR for upscaling. Every game I like is under development using UE5.

This are what the software will do once its ready.


Games can all run as fully RT. Stop the difference between Full RT and just RTGI.


All the AAA games are going to be fully Ray Traced or look dated. Ignoring the Ray Tracing feature is pointless when approx. 11 of the games you test support DXR. If you just want to play with raster graphics, then get a mid range GPU. Stick to 1440p. A top end gpu, the main feature is RT performance. All the top end gpus, for the most part are overkill for raster games. You get >100fps at most resolutions. You dont need more than 8GB of vram.
First of all and most important I think. If the UE5 is the future as you expressed, you dont need an RTX for it. Lumen doesn't use RT cores exclusively which means it is not hardware accelerated. Any card with sufficient power aka performance level can pull it off. If that is the future you don't need NV to run it but enough performance. What that performance would be, time will tell.
So your argument about not having RT tested exclusive games for NV hardware is irrelevant here. Of course this will come in a separate article dedicated to RayTracing. 3090Ti struggles with RT in some titles. By the time RayTracing is implemented (NV hardware accelerated games, RT cores) The 3000 series will be already obsolete with lack of power to pull it off.
 
Joined
Feb 20, 2022
Messages
175 (0.17/day)
System Name Custom Watercooled
Processor 10900k 5.1GHz SSE 5.0GHz AVX
Motherboard Asus Maximus XIII hero z590
Cooling XSPC Raystorm Pro, XSPC D5 Vario, EK Water Blocks EK-CoolStream XE 360 (Triple Fan) Radiator
Memory Team Group 8Pack RIPPED Edition TDPPD416G3600HC14CDC01 @ DDR4-4000 CL15 Dual Rank 4x8GB (32GB)
Video Card(s) KFA2 GeForce RTX 3080 Ti SG 1-Click OC 12GB LHR GDDR6X PCI-Express Graphics Card
Storage WD Blue SN550 1TB NVME M.2 2280 PCIe Gen3 Solid State Drive (WDS100T2B0C)
Display(s) LG 3D TV 32LW450U-ZB and Samsung U28D590
Case Full Tower Case
Audio Device(s) ROG SupremeFX 7.1 Surround Sound High Definition Audio CODEC ALC4082, ESS SABRE9018Q2C DAC/AMP
Power Supply Corsair AX1000 Titanium 80 Plus Titanium Power Supply
Mouse Logitech G502SE
Keyboard Logitech Y-BP62a
Software Windows 11 Pro
Benchmark Scores https://valid.x86.fr/2rdbdl https://www.3dmark.com/spy/27927340 https://ibb.co/YjQFw5t
First of all and most important I think. If the UE5 is the future as you expressed, you dont need an RTX for it. Lumen doesn't use RT cores exclusively which means it is not hardware accelerated. Any card with sufficient power aka performance level can pull it off. If that is the future you don't need NV to run it but enough performance. What that performance would be, time will tell.
So your argument about not having RT tested exclusive games for NV hardware is irrelevant here. Of course this will come in a separate article dedicated to RayTracing. 3090Ti struggles with RT in some titles. By the time RayTracing is implemented (NV hardware accelerated games, RT cores) The 3000 series will be already obsolete with lack of power to pull it off.
Lumen does use RT cores, if you had every used UE5 you would know there is a check box for hardware acceleration.

From Lumen Documentation, Hardware Ray Tracing

Hardware Ray Tracing supports a larger range of geometry types than Software Ray Tracing, in particular it supports tracing against skinned meshes. Hardware Ray Tracing also scales up better to higher qualities — it intersects against the actual triangles and has the option to evaluate lighting at the ray hit instead of the lower quality Surface Cache.

To use high quality hit lighting, the following must be enabled in the Project Settings under the Engine > Rendering section:


Support Hardware Ray Tracing

Enables ray tracing from supported operating systems, RHI, and video cards for higher quality results.

Ray Lighting Mode

When using Hardware Ray Tracing with Lumen, this setting controls whether reflections will reuse the Surface Cache for cheap lighting, or calculate lighting at the hit point for higher quality.

Lumen Platform Support​

  • Lumen does not support previous-generation consoles, such as PlayStation 4 and Xbox One.
  • Projects that rely on dynamic lighting can use a combination of Distance Field Ambient Occlusion and Screen Space Global Illumination on those platforms and legacy PC hardware.
  • Lumen is developed for next-generation consoles (PlayStation 5 and Xbox Series S / X) and high-end PCs. Lumen has two ray tracing modes, each with different requirements.
  • Software Ray Tracing Requirements: PlayStation 5 and Xbox Series S / X
  • Video cards using DirectX 11 with support for Shader Model 5 (SM5)
    Requires an NVIDIA GeForce GTX-1070 or higher card.
  • Hardware Ray Tracing Requirements: high-end PC
    • Windows 10 with DirectX 12
    • Video cards must be NVIDIA RTX-2000 series or higher, or AMD RX-6000 series or higher.
  • Lumen does not support mobile platforms. There are no plans to develop a dynamic global illumination system for the mobile renderer. Games using dynamic lighting need to use unshadowed Sky Light on mobile.
  • Lumen does not currently support Virtual Reality (VR) systems. While VR can be supported, the high frame rates and resolutions required by VR make dynamic global illumination a poor fit.

This demo is really nice if your PC can you can play it real time.

 
Joined
May 31, 2016
Messages
4,443 (1.42/day)
Location
Currently Norway
System Name Bro2
Processor Ryzen 5800X
Motherboard Gigabyte X570 Aorus Elite
Cooling Corsair h115i pro rgb
Memory 32GB G.Skill Flare X 3200 CL14 @3800Mhz CL16
Video Card(s) Powercolor 6900 XT Red Devil 1.1v@2400Mhz
Storage M.2 Samsung 970 Evo Plus 500MB/ Samsung 860 Evo 1TB
Display(s) LG 27UD69 UHD / LG 27GN950
Case Fractal Design G
Audio Device(s) Realtec 5.1
Power Supply Seasonic 750W GOLD
Mouse Logitech G402
Keyboard Logitech slim
Software Windows 10 64 bit
Lumen does use RT cores, if you had every used UE5 you would know there is a check box for hardware acceleration.

From Lumen Documentation, Hardware Ray Tracing




Support Hardware Ray Tracing

Enables ray tracing from supported operating systems, RHI, and video cards for higher quality results.

Ray Lighting Mode

When using Hardware Ray Tracing with Lumen, this setting controls whether reflections will reuse the Surface Cache for cheap lighting, or calculate lighting at the hit point for higher quality.

Lumen Platform Support​



This demo is really nice if your PC can you can play it real time.

Maybe I didnt express myself correctly. It does use hardware (5700 XT is not capable of RT you need 6000 series card for instance) but you dont need RT cores as a must since that is proprietary to NV only.
 
Joined
May 2, 2017
Messages
7,762 (2.78/day)
Location
Back in Norway
System Name Hotbox
Processor AMD Ryzen 7 5800X, 110/95/110, PBO +150Mhz, CO -7,-7,-20(x6),
Motherboard ASRock Phantom Gaming B550 ITX/ax
Cooling LOBO + Laing DDC 1T Plus PWM + Corsair XR5 280mm + 2x Arctic P14
Memory 32GB G.Skill FlareX 3200c14 @3800c15
Video Card(s) PowerColor Radeon 6900XT Liquid Devil Ultimate, UC@2250MHz max @~200W
Storage 2TB Adata SX8200 Pro
Display(s) Dell U2711 main, AOC 24P2C secondary
Case SSUPD Meshlicious
Audio Device(s) Optoma Nuforce μDAC 3
Power Supply Corsair SF750 Platinum
Mouse Logitech G603
Keyboard Keychron K3/Cooler Master MasterKeys Pro M w/DSA profile caps
Software Windows 10 Pro
Lumen does use RT cores, if you had every used UE5 you would know there is a check box for hardware acceleration.

From Lumen Documentation, Hardware Ray Tracing




Support Hardware Ray Tracing

Enables ray tracing from supported operating systems, RHI, and video cards for higher quality results.

Ray Lighting Mode

When using Hardware Ray Tracing with Lumen, this setting controls whether reflections will reuse the Surface Cache for cheap lighting, or calculate lighting at the hit point for higher quality.

Lumen Platform Support​



This demo is really nice if your PC can you can play it real time.

That Titanic demo is a nice enough demo, but it looks nowhere near the photorealism of that workshop demo above. You see that, right? It's not all that far from some of the nicer looking games today - though those tend to have more pragmatic concerns in mind (considering actual gameplay) rather than just looking good.

And yes, Lumen can use hardware RT, but it can also be run without it. And even with hardware acceleration it isn't a real-time renderer, making it fundamentally unsuited for games. There is little doubt that there are tons of very cool things that can be done with RT in the future. But that future is still quiat a ways off - 5+ years in my estimation. The games industry moves slowly in adopting new features, and a fundamentally new and different rendering paradigm is going to take a lot of time to become dominant. You're describing it as if it is dominant today, which is just pure nonsense.
 
Joined
Feb 20, 2022
Messages
175 (0.17/day)
System Name Custom Watercooled
Processor 10900k 5.1GHz SSE 5.0GHz AVX
Motherboard Asus Maximus XIII hero z590
Cooling XSPC Raystorm Pro, XSPC D5 Vario, EK Water Blocks EK-CoolStream XE 360 (Triple Fan) Radiator
Memory Team Group 8Pack RIPPED Edition TDPPD416G3600HC14CDC01 @ DDR4-4000 CL15 Dual Rank 4x8GB (32GB)
Video Card(s) KFA2 GeForce RTX 3080 Ti SG 1-Click OC 12GB LHR GDDR6X PCI-Express Graphics Card
Storage WD Blue SN550 1TB NVME M.2 2280 PCIe Gen3 Solid State Drive (WDS100T2B0C)
Display(s) LG 3D TV 32LW450U-ZB and Samsung U28D590
Case Full Tower Case
Audio Device(s) ROG SupremeFX 7.1 Surround Sound High Definition Audio CODEC ALC4082, ESS SABRE9018Q2C DAC/AMP
Power Supply Corsair AX1000 Titanium 80 Plus Titanium Power Supply
Mouse Logitech G502SE
Keyboard Logitech Y-BP62a
Software Windows 11 Pro
Benchmark Scores https://valid.x86.fr/2rdbdl https://www.3dmark.com/spy/27927340 https://ibb.co/YjQFw5t
Maybe I didnt express myself correctly. It does use hardware (5700 XT is not capable of RT you need 6000 series card for instance) but you dont need RT cores as a must since that is proprietary to NV only.
Maybe you dont get Lumen software is for hardware like the new consoles and it states this fact in the UE5 documentation. High end PC's as pointed out before will use HW Ray Tracing as they support that feature. That why a 2000 series nvidia gpu and above and an amd 6000 series as about are listed for HW RT. Note Lumen software mode hammers the cpu and will reduce fps massively.

Lumen is Unreal Engine 5’s new dynamic Global Illumination and reflections system, targeted at next-generation consoles. source
Lumen is Unreal Engine 5's fully dynamic global illumination and reflections system that is designed for next-generation consoles source
https://docs.unrealengine.com/5.0/en-US/lumen-global-illumination-and-reflections-in-unreal-engine/

That Titanic demo is a nice enough demo, but it looks nowhere near the photorealism of that workshop demo above. You see that, right? It's not all that far from some of the nicer looking games today - though those tend to have more pragmatic concerns in mind (considering actual gameplay) rather than just looking good.

And yes, Lumen can use hardware RT, but it can also be run without it. And even with hardware acceleration it isn't a real-time renderer, making it fundamentally unsuited for games. There is little doubt that there are tons of very cool things that can be done with RT in the future. But that future is still quiat a ways off - 5+ years in my estimation. The games industry moves slowly in adopting new features, and a fundamentally new and different rendering paradigm is going to take a lot of time to become dominant. You're describing it as if it is dominant today, which is just pure nonsense.
You can play the titanic demo. Its basiclly created so you can see the inside of the titanic. In Unreal Engine 5 you can 100% use hardware Ray Tracing in real time. Metro Exodus Enhanced Edition is pathtraced with infinite light bounces.

Dynamic Diffuse Global Illumination (DDGI) models the additional interactions from the second bounce of light on to infinity. source

Then you state its slowly adopting, the list of RT games.

AMID EVIL
Aron's Adventure
Battlefield 2042
Battlefield V
BIOHAZARD VILLAGE Z Version
Bright Memory
Bright Memory: Infinite
Call of Duty: Black Ops Cold War
Call of Duty: Modern Warfare
Chernobylite
Chorus
Control
Crysis Remastered
Crysis 2 Remastered
Crysis 3 Remastered
Cyberpunk 2077
Deathloop
Deliver Us The Moon
DiRT 5
Domino Simulator 2020
DOOM Eternal
Dying Light 2 Stay Human
Enscape
Escape From Naraka
F1 2021
F.I.S.T.: Forged In Shadow Torch
Far Cry 6
Five Nights at Freddy's: Security Breach
Fortnite
Forza Horizon 5
Ghostrunner
Ghostwire: Tokyo
Godfall
Helios
Hellblade: Senua's Sacrifice
ICARUS
INDUSTRIA
Justice
JX3 Online RTX Version
LEGO Builder's Journey
Life Is Strange: True Colors
Martha Is Dead
Marvel's Guardians of the Galaxy
Mechwarrior 5: Mercenaries
Mercs Fully Loaded
Metro Exodus PC Enhanced Edition
Minecraft with RTX
Moonlight Blade
Mortal Shell
Myst
Myth of Empires
Observer: System Redux
Paradise Killer
Poker Club
Pumpkin Jack
Quake II RTX
RAZE 2070
Redout: Space Assault
Resident Evil Village
Rune II
Severed Steel
Shadow of the Tomb Raider
Soulmate
Stay in the Light
Sword and Fairy 7
The Ascent
The Fabled Woods
The Medium
The Orville Interactive Fan Experience
The Persistence
The Riftbreaker
The Riftbreaker: Prologue
To Hell With It
Watch Dogs: Legion
Wolfenstein: Youngblood
World of Warcraft: Shadowlands
Wrench
Xuan-Yuan Sword VII

You can hve a 50 game review and never touch a raster game. Accept that I have some degree of a point. Gaslighting me wont work.
 
Last edited:
Top