An amazing and well thought out test. Results I wasn't quite expecting, but now very clear to me why and makes perfect sense. I've watched all 3 of your videos twice now, just in case I missed anything. Now I'm hungry for more. Keep em coming. 💯👊🏻
Great video/benchmark ! i just hit subscribe ! i only play iRacing (started 6 weeks ago) and i really love it !! 5600x + 3070 + 16gb ram right now (32'' 2k 165hz monitor) and i could upgrade for 12600k + 3070ti for like 100$ (buy a friend's pc and resell mine). I see the 12600k is better for iracing and the 3070ti will be slightly better so i think ill go for it ! i'll build a new beast next year for triple 32's in 2k !! Thanks man keep up the great vids!
Thanks for this extensive effort! Just finished upgrading my simrig with 8channel haptic feedback and the 49" monitor is gone ... Hopefully for christmas a quest 3 will be here. I know my pc (AMD 5600x, nvidia 3070) won't be happy with it though. 😄
Thanks for great video! Now I'm really tempted to upgrade my 12600KF to 13600K or 13700K just to get some performance headroom. Not too costly of an upgrade.
Your videos seem to be helpful on my path to simracing. I recently build a 7800x3d+4080s combo and your vids were helpful in that process. Now I'm thinking between Quest3 vs 3x27" screen setup.
Happy to help :) As for VR vs Triples, that's personal preference. Monitors have better resolution to see the dips and bumps in the track and it's easier to load custom overlays for track position, lap timings, and chat. However, I prefer VR because I want to move my head within the car and lock my eyes to the next apex, because that makes me feel like I'm driving a race car :)
Great info. I had a PC built a couple months ago with a 7700x, 4070ti, and 32gb 6000hz RAM. (shout out to RichD-PC for the amazing build and awesome price) I'm incredibly happy with it. I would have loved to go with a 7800x3d, but I wanted to stay within a certain budget. It worked out because it seems to be very well balanced with the 4070ti on iRacing. I'm running a Rift S right now, and this PC has no problem running it with pretty high settings and 1.5 SS at a very stable 80 FPS. I'm looking to upgrade to a Reverb G2 v2 when I find a good Christmas deal. I'm excited to see how it runs.
I hear the G2 is a great headset. I have not used one, however, it uses fresnel lenses like the Valve Index, Quest 2, and Rift which I have used. The Quest 3 uses 'pancake' lenses, however, and those do not feature the circular ridges of a fresnel design. This means 'God rays' are mostly gone, and you can dart your eyes around the lenses with perfect clarity... so you do not need to move your head! It has amazed me how that makes the headset feel more natural, and the experience more 'real' compared to headsets with fresnel lenses. Just something to consider before you purchase!
If you ever want to test another track that’s like Daytona do Long Beach. Kills FPS when passing the pits and going down the back straight. The water is a huge FPS killer
Hi much appreciated work, I have watched all your videos. My take based on all of them is that intels Loews are much tighter which is beneficial. May I suggest that you compare the two high end processors 7800x3d vs the 13900 or ks so it is a more of an apples to apples comparison. In addition I would also look at optimizing both to their Maximum’s like OC the CPU and memory and /or disabling the ecores . I hope that a lot of people support you because we simracers need this kind of analysis as it saving work and money over the long term. And you need to get a 4090 to really stretch these CPU’s. Thanks again for the effort.
Thank you for the thorough reply, and for watching the videos! The lows do seem better with Intel, and I wonder if that's a 'negative' to the 3D V-Cache from AMD. Perhaps it's worthy of a detailed analysis, including some real-world examples. 13900K is about 40% more expensive than the 13700K and 7800X3D, that is a very hefty premium. for higher boost clocks and 33% more L2 Cache. As you point out, those 16 Ecores are usually wasted with simulators. OCing can certainly bring up the performance of Intel processors, which I why I recommend the K models. I think it's a great topic to specifically explore with some of the processors. I wonder about the 4090 as well, however, it's very expensive! I understand the thought process to using it for CPU testing, but an assumption is made that the CPU will still perform admirably when constrained by a slower GPU. The video I just published for ACC suggests there may be undesirably consequences when pairing the 7800X3D with a 'slower' card (assuming my results are accurate!)
I see your point, but unfortunately every little helps and it costs . Furthermore in regards to iracing triple settings you might need to have a custom setting based on ultra high but a bit optimized excluding high intensity ones like foliage blur etc
Bundle deals that have a CPU sold with a motherboard, memory, and/or CPU cooler are usually the best deals when buying new. Going into Black Friday and Holiday sales, I would keep your eyes open for such combos that fit within your budget. Hopefully the results shown in the video provide some insight on what you should expect when powering a Quest 3.
Wow i guess my 5600x and 6750xt are just a terrible vr combo, im litteraly struggeling to get it barely playable on lowest graphics with fsr and vrperfkit on quest 2. Seems like swapping my gpu for a nvidia would make it playable atleast. Cheers for the video!
Hello and Thank You for the hard work! Curious about the 5800x3d now that Tempest has reared its head! Initial thoughts? I realize you may not be able to abstractly speak to this. Appreciate in advance! - edit- I just found more recent content from you that already covers my question- you’re on top of it!
Under volt the GPU and reduce its power target. I have a 3080ti as well but limit it to around 220-260w and it’s much cooler. I’d say only around a 10% loss in performance for a ~40% reduction in power usage.
The 3080ti reminds me of Intel 14th generation, Nvidia pushed the power consumption too far to gain just a few more points of performance. Out of the box, it will recklessly consume voltage and chase high clocks. I'm a bit torn, however, when it comes to testing my equipment 'tweaked' instead of Stock. I imagine most sim racers just use the card as it was configured from the factory.
Thanks dude :) Unfortunately I do not (yet) have experience with Radeon and VR like Quest 3. On paper, your 6800XT should have enough grunt to render iRacing for the headset.
Running a 4090 with a Ryzen 9 5950x. Is the CPU outdated ? I am fine i dont run and need to get AM5. But would bring a better CPU mutch more perfomance? Nice Video.. Thx for the effort.
Bluntly speaking, you've got the fastest video card available! So yes... pairing it with anything short of the fastest CPU means you are leaving performance on the table ;) Easiest option is 5800X3D... next-gen AM5 chips should be Q3 this year but rumors of performance gains are not conclusive yet.
Vr is so weird to figure out. I have a 4080, 7800x3d, and quest3, and for the last few weeks I run it with the oculus slider maxed, debug set to 1.3 supersampling, and then the clarity is so great that it's finally like a monitor and the settings IN iracing are essentially irrelevant. I can get 80 fps out of that setup in a open wheeler and 75ish in a gt3 or what have you. I set it to 72 and turned off the asynchronous spacewarp in the debug1
got myself Pimax Crystal Light with local dimming and Display Port Cable, end of May delivery expected. Till then I have to get PC as I never had any. 4080 super and 7800x3D is what I am getting. Watched lot of PC builder videos and I believe it must be 4080 super and 7800x3D or 17 14700. If anyone got suggestion what will suit my VR better 7800x3d or 17 14700 let me know!
That's a difficult question to answer, because as you know, there are a lot of settings that can be adjusted! There is a 'Quest 3' thread in the official iRacing forums and a few others that discuss settings. Unfortunately I have to point you there until I cover the topic in more detail.
Oh totally. I tried AMS2, ACC, and RF2 in VR. The problem is that race replays in those sims do not accurately reflect the real-time experience, or I cannot watch the replay from a VR perspective. I'm down to manually do the laps for data collection, but I need to create demanding circumstances to properly evaluate the hardware. And that means VR racing with low FPS, which is nauseating. I tried doing that with ACC for an afternoon and had to quit all computing for a day, that's how sick it made me :S Anyways, I've got a few other ideas on how to tackle VR benchmarking. I'm working on another analysis with iRacing and VR, but hopefully I can apply this new technique to the other sims.
I do own some RaceRoom content, but I'll admit I haven't launched that title in a while. My content this month will be focused on iRacing rain, VR, and adding another racing title. I was leaning towards Le Mans Ultimate just because it's the 'newest' title on the market.
Can I ask one more question. The screen resolution in game.. does that help boost VR if you turn it down? Mines set to 4K as it’s a 4K monitor. But in VR obviously it’s not 4K
"screen resolution in game" do you mean the resolution displayed within iRacing's graphic options? I've never noticed iRacing actually force that resolution to the VR preview window, because I think that setting only applies to the Monitor ini file.
Yeah that. Didn’t know if in game resolution mattered. Wanted to turn down in game resolution just to maybe give an extra help to gpu as it’s not using that screen. I turned off Asynchronous space warp in oculus debug tool, to stop frame rate from dropping in game. Turned it up to 90hz and 1.5 ss from the oculus app - won’t harm the GPU will it?
@@FruitLoop2020 Which GPU do you have? The primary thing that can damage a video card over time is heat saturation. For example, if your computer chassis cannot exhaust hot air and bring in cooler air. Or dust collects within the card's heatsink fins, preventing airflow. Even in those scenarios, most modern cards can automatically decrease voltage draw and clock speeds so the card produces less heat.
@@FruitLoop2020 1.5SS is pretty high for a 3070, I think you'd have to reduce a lot of visual graphic options to get it stable at 90 fps. Keep in mind that distant objects will always be blurry with a Quest 3 because your computer renders the sim, encodes it to video, and streams that to the headset. The cockpit looks great on Q3 and super sampling helps there, but as you look farther and farther away, all that processing power needed for SS is 'wasted' on pixels that just get washed out via the encoding. Considering how little time you spend looking at your cockpit, decrease SS until you notice it impacts your ability to read the gauge cluster. Then tune graphic options to find stable 90fps with your 3070.
An amazing and well thought out test. Results I wasn't quite expecting, but now very clear to me why and makes perfect sense. I've watched all 3 of your videos twice now, just in case I missed anything. Now I'm hungry for more. Keep em coming. 💯👊🏻
Man, thank you for the real test's. I hope your channel can grow. Really got work here
Great video/benchmark ! i just hit subscribe ! i only play iRacing (started 6 weeks ago) and i really love it !! 5600x + 3070 + 16gb ram right now (32'' 2k 165hz monitor) and i could upgrade for 12600k + 3070ti for like 100$ (buy a friend's pc and resell mine). I see the 12600k is better for iracing and the 3070ti will be slightly better so i think ill go for it ! i'll build a new beast next year for triple 32's in 2k !! Thanks man keep up the great vids!
Sweet vid! Thanks for the info 👍
Thanks for this extensive effort! Just finished upgrading my simrig with 8channel haptic feedback and the 49" monitor is gone ... Hopefully for christmas a quest 3 will be here. I know my pc (AMD 5600x, nvidia 3070) won't be happy with it though. 😄
Thanks for great video! Now I'm really tempted to upgrade my 12600KF to 13600K or 13700K just to get some performance headroom. Not too costly of an upgrade.
Amazing video!! VERY VERY GOOD
Would be great to see tests on HP Reverb G2. Thanks for the video. 😊
You're welcome :) I have a Valve Index which should provide relevant OpenXR results to HP G2 users. I can also conduct more interesting data analysis.
Your videos seem to be helpful on my path to simracing. I recently build a 7800x3d+4080s combo and your vids were helpful in that process. Now I'm thinking between Quest3 vs 3x27" screen setup.
Happy to help :) As for VR vs Triples, that's personal preference. Monitors have better resolution to see the dips and bumps in the track and it's easier to load custom overlays for track position, lap timings, and chat. However, I prefer VR because I want to move my head within the car and lock my eyes to the next apex, because that makes me feel like I'm driving a race car :)
You are a legend!
I am so happy I chose the 5800X3D a year ago. It is still a beast today
Thank you. Very informative video. Might grab a 13900KS for a whirl
Great info. I had a PC built a couple months ago with a 7700x, 4070ti, and 32gb 6000hz RAM. (shout out to RichD-PC for the amazing build and awesome price) I'm incredibly happy with it. I would have loved to go with a 7800x3d, but I wanted to stay within a certain budget. It worked out because it seems to be very well balanced with the 4070ti on iRacing. I'm running a Rift S right now, and this PC has no problem running it with pretty high settings and 1.5 SS at a very stable 80 FPS. I'm looking to upgrade to a Reverb G2 v2 when I find a good Christmas deal. I'm excited to see how it runs.
I hear the G2 is a great headset. I have not used one, however, it uses fresnel lenses like the Valve Index, Quest 2, and Rift which I have used. The Quest 3 uses 'pancake' lenses, however, and those do not feature the circular ridges of a fresnel design. This means 'God rays' are mostly gone, and you can dart your eyes around the lenses with perfect clarity... so you do not need to move your head! It has amazed me how that makes the headset feel more natural, and the experience more 'real' compared to headsets with fresnel lenses. Just something to consider before you purchase!
Thanks for that explanation, make some sense! Has the reverb g2 myself.
But might upgrade my 9600k really soon :)
well done, well done
great benchmarks! can’t wait for hardware to handle 120+ 1%/0.1% low fps
If you ever want to test another track that’s like Daytona do Long Beach. Kills FPS when passing the pits and going down the back straight. The water is a huge FPS killer
Hi much appreciated work, I have watched all your videos. My take based on all of them is that intels Loews are much tighter which is beneficial. May I suggest that you compare the two high end processors 7800x3d vs the 13900 or ks so it is a more of an apples to apples comparison. In addition I would also look at optimizing both to their Maximum’s like OC the CPU and memory and /or disabling the ecores . I hope that a lot of people support you because we simracers need this kind of analysis as it saving work and money over the long term. And you need to get a 4090 to really stretch these CPU’s. Thanks again for the effort.
Thank you for the thorough reply, and for watching the videos!
The lows do seem better with Intel, and I wonder if that's a 'negative' to the 3D V-Cache from AMD. Perhaps it's worthy of a detailed analysis, including some real-world examples.
13900K is about 40% more expensive than the 13700K and 7800X3D, that is a very hefty premium. for higher boost clocks and 33% more L2 Cache. As you point out, those 16 Ecores are usually wasted with simulators.
OCing can certainly bring up the performance of Intel processors, which I why I recommend the K models. I think it's a great topic to specifically explore with some of the processors.
I wonder about the 4090 as well, however, it's very expensive! I understand the thought process to using it for CPU testing, but an assumption is made that the CPU will still perform admirably when constrained by a slower GPU. The video I just published for ACC suggests there may be undesirably consequences when pairing the 7800X3D with a 'slower' card (assuming my results are accurate!)
I see your point, but unfortunately every little helps and it costs . Furthermore in regards to iracing triple settings you might need to have a custom setting based on ultra high but a bit optimized excluding high intensity ones like foliage blur etc
@@BenchmarkOdysseys maybe a chiplet architecture flaw. The extra cache needed to compensate for the chiplet latency penalty.
Thanks for the CPU performance comparison data. Do you have any comments or suggestions wrt costs of the 6 CPUs?
Bundle deals that have a CPU sold with a motherboard, memory, and/or CPU cooler are usually the best deals when buying new. Going into Black Friday and Holiday sales, I would keep your eyes open for such combos that fit within your budget. Hopefully the results shown in the video provide some insight on what you should expect when powering a Quest 3.
Wow i guess my 5600x and 6750xt are just a terrible vr combo, im litteraly struggeling to get it barely playable on lowest graphics with fsr and vrperfkit on quest 2. Seems like swapping my gpu for a nvidia would make it playable atleast. Cheers for the video!
A triple monitor test would be awesome
I have already begun the testing!
Hello and Thank You for the hard work! Curious about the 5800x3d now that Tempest has reared its head! Initial thoughts? I realize you may not be able to abstractly speak to this. Appreciate in advance! - edit- I just found more recent content from you that already covers my question- you’re on top of it!
Under volt the GPU and reduce its power target. I have a 3080ti as well but limit it to around 220-260w and it’s much cooler. I’d say only around a 10% loss in performance for a ~40% reduction in power usage.
The 3080ti reminds me of Intel 14th generation, Nvidia pushed the power consumption too far to gain just a few more points of performance. Out of the box, it will recklessly consume voltage and chase high clocks.
I'm a bit torn, however, when it comes to testing my equipment 'tweaked' instead of Stock. I imagine most sim racers just use the card as it was configured from the factory.
Hey man, amazing work, I have a 5900X and a 6800XT, and thinking about picking a quest 3, do you think it will be able to drive it?
Thanks dude :) Unfortunately I do not (yet) have experience with Radeon and VR like Quest 3. On paper, your 6800XT should have enough grunt to render iRacing for the headset.
@@BenchmarkOdysseys thanks a lot for the answer 👍
Running a 4090 with a Ryzen 9 5950x. Is the CPU outdated ? I am fine i dont run and need to get AM5. But would bring a better CPU mutch more perfomance? Nice Video.. Thx for the effort.
Bluntly speaking, you've got the fastest video card available! So yes... pairing it with anything short of the fastest CPU means you are leaving performance on the table ;) Easiest option is 5800X3D... next-gen AM5 chips should be Q3 this year but rumors of performance gains are not conclusive yet.
@@BenchmarkOdysseys Thx a lot for ur quick assessment. Good to know it for Sure 🤣 Maybe later this year.
Vr is so weird to figure out. I have a 4080, 7800x3d, and quest3, and for the last few weeks I run it with the oculus slider maxed, debug set to 1.3 supersampling, and then the clarity is so great that it's finally like a monitor and the settings IN iracing are essentially irrelevant. I can get 80 fps out of that setup in a open wheeler and 75ish in a gt3 or what have you. I set it to 72 and turned off the asynchronous spacewarp in the debug1
I have a 3070 and 32gb. Going to try your settings
got myself Pimax Crystal Light with local dimming and Display Port Cable, end of May delivery expected. Till then I have to get PC as I never had any. 4080 super and 7800x3D is what I am getting. Watched lot of PC builder videos and I believe it must be 4080 super and 7800x3D or 17 14700. If anyone got suggestion what will suit my VR better 7800x3d or 17 14700 let me know!
What settings would you recommend for quest 3, RTX 3070 and 5800. As my frame are 72 but drop to 35 sometimes
That's a difficult question to answer, because as you know, there are a lot of settings that can be adjusted! There is a 'Quest 3' thread in the official iRacing forums and a few others that discuss settings. Unfortunately I have to point you there until I cover the topic in more detail.
In use 5800x3d 4080 and quest 3 linkcable works great with virtual desktop godlike.
I am surprised to hear this! Which version of Virtual Desktop are you using? I thought the latest versions of the software only work with WiFi Link.
Dowloaded the vdxr version. Latest. Works fine
Why just Iracing? Be interesting to see results for other racing sims
Oh totally. I tried AMS2, ACC, and RF2 in VR. The problem is that race replays in those sims do not accurately reflect the real-time experience, or I cannot watch the replay from a VR perspective.
I'm down to manually do the laps for data collection, but I need to create demanding circumstances to properly evaluate the hardware. And that means VR racing with low FPS, which is nauseating. I tried doing that with ACC for an afternoon and had to quit all computing for a day, that's how sick it made me :S
Anyways, I've got a few other ideas on how to tackle VR benchmarking. I'm working on another analysis with iRacing and VR, but hopefully I can apply this new technique to the other sims.
Can I suggest Raceroom, which very much CPU bound being DX9@@BenchmarkOdysseys
I do own some RaceRoom content, but I'll admit I haven't launched that title in a while. My content this month will be focused on iRacing rain, VR, and adding another racing title. I was leaning towards Le Mans Ultimate just because it's the 'newest' title on the market.
Can I ask one more question. The screen resolution in game.. does that help boost VR if you turn it down? Mines set to 4K as it’s a 4K monitor. But in VR obviously it’s not 4K
"screen resolution in game" do you mean the resolution displayed within iRacing's graphic options? I've never noticed iRacing actually force that resolution to the VR preview window, because I think that setting only applies to the Monitor ini file.
Yeah that. Didn’t know if in game resolution mattered. Wanted to turn down in game resolution just to maybe give an extra help to gpu as it’s not using that screen.
I turned off Asynchronous space warp in oculus debug tool, to stop frame rate from dropping in game. Turned it up to 90hz and 1.5 ss from the oculus app - won’t harm the GPU will it?
@@FruitLoop2020 Which GPU do you have? The primary thing that can damage a video card over time is heat saturation. For example, if your computer chassis cannot exhaust hot air and bring in cooler air. Or dust collects within the card's heatsink fins, preventing airflow. Even in those scenarios, most modern cards can automatically decrease voltage draw and clock speeds so the card produces less heat.
@@BenchmarkOdysseys I have a 3070 (not ti) but I just can’t get it looking good
@@FruitLoop2020 1.5SS is pretty high for a 3070, I think you'd have to reduce a lot of visual graphic options to get it stable at 90 fps.
Keep in mind that distant objects will always be blurry with a Quest 3 because your computer renders the sim, encodes it to video, and streams that to the headset. The cockpit looks great on Q3 and super sampling helps there, but as you look farther and farther away, all that processing power needed for SS is 'wasted' on pixels that just get washed out via the encoding.
Considering how little time you spend looking at your cockpit, decrease SS until you notice it impacts your ability to read the gauge cluster. Then tune graphic options to find stable 90fps with your 3070.