Keep in mind that on the 6000 series Radeons, AMD reduced the Memory bandwidth a lot and then added the infinity fabric to compensate. This means that each GPU has a pixel trhougtput "ceiling" when it comes to resolution. The top of the line 6950XT had a chip very well capable of being faster than the 3090 Ti in 1080p and 1440p, but it TANKED in 4K. This goes for all the lineup. The 6700 series could do 1080p very well, but at 1440p it was already saturating it's smaller infinity fabric. So the smaller the resolution the faster the AMD card is compared to Nvidia. When you go triple screens then it only shows more and more it's limitations. Notice that AMD has a 192 bit bus while Nvidia has 256, AMD is only using VRAM chips that are twice as dense as Nvidia's, otherwise it'd be at 6GB. This was a PLANNED move to reduce cost, and Nvidia followed MAD and went the same route with the 4000 series GPUs.
Remember that screen space effects also run at full resolution and also contribute to this bottleneck, reducing their quality will increase longevity for the GPU.
Great video, thanks for all information, really useful. Could I know what monitors are you using on triples? I'm using RTX3060Ti with a 34" Ultra wide 2K monitor but I'd like to change to triples in future.Thanks again!
Thank you :) Gigabyte M28U, it's one of the lesser expensive high-refresh 4K monitors. It definitely makes more sense to have 4K on a bigger screen size, but running three 32" or bigger is a lot of space. I move my gear around too much, it would just be too cumbersome.
I'm very surprised because I thought run triples with a 3060Ti was madness. Obviously it also depends on what CPU are you using to run it. Very interesting!!! Thanks friend. @@BenchmarkOdysseys
Reasonable idea. The under-performance at Daytona at Triple 1080p in iRacing, however, remains the same with max 2700Mhz to GPU core (min 2500Mhz, -50mV), +100Mhz to VRAM with fast timings, and 15% power. What's strange about this performance (with or without settings above) is the card refusing to consume more power, averaging between 93W and 110W during the bench. This 6700XT consistently delivers these poor results during this scenario. iRacing does reward well tuned memory timings, especially on Intel platform, and this XMP profile is well above the 'stock' settings. I'm not familiar with RAM timings specific to systems running Radeon graphics cards.
6700XT owner here, coupled with a 5600X, i can relate on strange behavior, on my tripple 1080p I go from 60 to 140 fps depending on tracks. What is important thought, is that it is still smooth, and on my 144hz monitors I can barely tell the difference if fps meter is hidden. I went full red team for the first time, completely fed up by Nvidia approach to gpu selling/branding the last 3 years, and I’m glad I did. I mean, what is next ? 5070 ti super ti dragon ? A meme became real this year…
I'm not using eyefinity, i'm using the multi projector setting in Iracing. On asseto corsa, FPS are a bit better usually but SOL and all can really make FPS vanish.
You're welcome :) But that's a loaded question! A significant hurdle to overcome is the frustration that nothing is really 'plug-n-play' on PC for sim racing. Yes, everything is customizable... but that's by necessity and not luxury. If you have the patience to learn and tenacity to problem solve, the sim racing experience and community is a huge upgrade over Forza and console racers. When it comes to 'What PC should I buy?', that's specific to your budget, display(s), and racing desires. I'm planning to offer an inexpensive service to help new users, so if you're interested, drop me an email (on my YT info tab).
@13:13 those commenters dont know their arse from their elbow 🤣nvidia shills come out to play. for their information, iracing like most sim games are more cpu bound than gpu bound. get a better processor, the gpu isnt the issue. i need to cap my frames to 161, max settings on my 6000 seriers
Traditional gaming benchmarks suggest the 3080ti is in the ballpark of the 4070 series: a bit faster than vanilla 4070 but slower than 4070ti Super. So any benchmarks of mine using the 3080ti, I would expect 4070 series to be +/- 10%. 7800xt seems really compelling because it's closest to that sweet spot of $500 USD.
Wow! This was great!! Thanks for make and share it! Your work is superb! Thansk also to include AMS2, it gives some interesting information to analize. If you need more topics for new videos I hope some day you could make something referred to VR in Simracing. Again, thanks and congrats for your video! 👏🏼👏🏼👏🏼👏🏼👏🏼👏🏼
Glad you enjoyed! I've only done one video with VR using an Oculus Quest 3. I've got big plans for a more in-depth VR analysis. Stay tuned by subscribing! :D
@@BenchmarkOdysseys heh thanks, although my videos are just simple game benchmarks on my hardware. You put great amount of effort into your content, i know how long it takes to bench everything also keep track of consistent testing methodology, its especially great because sim racing benchmarks are really scarce👍
Image quality is difficult to measure, and my subjective opinion would probably be inconsistent. I only used FSR with ACC because it's a more GPU demanding title, and I figured anyone with these cards would be using it. I'd like include results with it in the future, however, I'd prefer to also show the artifacts that it creates and I dont have a high speed camera for that.
Hi, I was just curious what would you recommend out of these GPU's for the Meta Quest 3 for IRacing and Assetto corsa? Or would you say something like the RX 6800 and 3070 would be a better choice?
Hello :) I have yet to test VR with AC, however, with iRacing there is a clear advantage with GeForce and VR. Therefore, I recommend RTX 3070 over RX 6800!
*Benchmark Odysseys* I just posted your "(unannounced) Liquid Freezer III from Arctic Cooling" vod in both the Hardware Unboxed and PCMR discords. You are welcome =>
@@BenchmarkOdysseys Just this past Monday, I ordered the ARCTIC Liquid Freezer II 420, which was scheduled to be delivered this coming Monday. But I've now cancelled the order, because I'd rather wait for the newest model. Do you have an Amazon affiliate link?
@@rcald-gz5jd Oh for sure I'd go for the new model too! I have not (yet) set up an affiliate connection with any online stores. But I'll look into this weekend. Thank you for asking, that's super cool 😎
Awesome! Thanks for following up on testing cheaper GPUs. How do you think these benchmarks translate to say a 5800x3D CPU? I have a 5800x3D + 5700XT combo on triple 1440p and it's hurting for frames in some instances.
I have not used 5700XT, but from other user experiences I've read, seems like it can struggle with triple screens. Your 5800X3D is plenty quick and keep close with Ryzen 7000 series and Intel 14th Gen.
iRacing is pretty good at working with 8GB of VRAM. You'd probably see bigger performance gains by upgrading to an Ampere GPU like 3070 even though it just has 8GB. Also, many RTX 20 series users have experienced crashes with Tempest. iRacing's July1st patch is trying to address that. 3070 is 2 years newer architecture compared to 2080ti (2018 release).
Yes, I think Triple 1440p is a difficult challenge for an RX 6800. By the time you reduced quality and increased FSR, I think the image on 1080p may look better. But image quality is a subjective experience.
You're welcome 😁 If you enjoy multiple simulators, it seems Intel 'wins' more often then AMD, but I recommend AM5 because of future CPU support. I'm currently making a video for 4080 Super, and it's very fast and power efficient.
Seems like X3D V-Cache does well in iRacing, especially around more complicated tracks. And the tracks are not getting more simple! If you don't mind waiting a little bit, rain will be launching soon and I will be doing some performance analysis with it.@@CharlesDRACOULIDES
@@BenchmarkOdysseys thank you ! Today i have i7 7700k and 2080 ti on 34 4k screen but o want switch on triple 2k screen and i look what cpu and gpu is better for iracing. 4090 seems very powerfull but expensive :( The 7900xtx can be also the good choice ? Yes, you are right for the Rain !
@@CharlesDRACOULIDES definitely upgrade the CPU first! Your 2080ti performance is between the 3060ti and 3080ti results. I think you will be surprised at the gains provided by a current-gen CPU.
Yup, SPS was enabled on the 3060ti. The note dashed across the screen on the first bench. Maybe I'll do a follow-up test... for example, in the benchies where the Radeon performance significantly changes from 4K to triple 1080p... does the 3060ti perform significantly worse when SPS is disabled?
@@BenchmarkOdysseys Would definitely be interesting to see that tested along with using a lower end CPU that might be more typically used with this class of GPU. Is the nVidia CPU advantage only down to SPS, or is there something else at play? Is there any difference on a single screen when using a lower end CPU?
Hello, and you are welcome :) Just XMP profile for this RAM: G.SKILL Trident Z5 RGB DDR5 7200 CL34-45-45-115 I'll admit it is not stable with a memory torture test (it is at 6800), but I want to compare these results to running at a slower speed. When the 3060ti runs out of VRAM, the simulator should use system RAM... and I want to see if there's a measurable difference. Subscribe and stay tuned!
Thank you. I am since your first video. I happen to have exactly same setup but with 4090 and run a 6800 c32 Kingston ram using the buildzoid settings shared with hardware unboxed. I asked because I was wondering whether I should try a 7200 kit but it should run at c32 in order to make a difference I think
Oh that's great! The little 'subscribed' icon wasn't showing in my chat window, perhaps that's a bug with UA-cam Studio. Thanks for telling me! I don't want to bug people who have already signed up heh
I'm really interested in AMS2, thanks for the benchmark, It would be great if you did a CPU benchmark like you did for ACC, I'm trying to check if my 10400f is holding me back.
Really interesting video for sim racers. Guessing that the results using the more current 6750xt would be around 7% - 10% higher fps than the 6700xt used in the video.
You say you're going to bench with AMS 2 & iRacing, but later mention Asseto Corsa gives a distinct advantage to vRam. So what are we trying to evaluate here? Also it's important to note that the 6700 series runs at 110c as "normal" based on feedback from Radeon dev team. No idea why, but it's very temp dependant.
I have a 5800X3D and an 6800XT and 32” triples. With I racing, big fields at night I have a lot of trouble, way below 60FPS. Even some day time races are tough, I thought this system would be fine but it’s not.
You have lots of options for Triple 1080p! Something from the 4070 lineup is more than enough, even a used 3080 would be plenty fast. The 4060Ti is also a fine card, it's just overpriced by $50 or $100. I haven't done any testing with the Radeon 7000 series (yet), so I cannot recommend until I do.
bruh....thank you for reaffirming my 6700XT buy last month, its the best bang for buck card when it comes to vram. its a 1440p card, but I game at 1080 locked at 60fps so its an absolute beast per my requirements. the 12 gb vram should keep this card relevant for at least the next 2 or so years. testing it at 4k is a bit unfair because its a 1440p card. it was never advertised at 4k.
VRAM is not as important as you think, you can find a good explanation as to why that is on the gamersnexus channel, the 6700 xt video to be precise, but it doesn't only apply to this card, it applies to all medium budget gpus that have more than 8gb
I'm upgrading to a 4080 super coming from a 2080 it's time. Installing with a older the week. My older CPU is running at 55% with the 2080 at 95% to benchmark( triple 32's @ 2k) , then pairing with a new build with a 78003dx. I just want to see the jumps. Thanks for all your work.
Right so the question will Radeon run triples in iRacing seems to be answered. They will. So that brings Team Red back on to the table as an upgrade path. Where the Radeon dips inexplicably and all the performance is not being utilised throws the thought into my mind would that have a different result if the CPU and GPU were both AMD? Perhaps the Radeon might play better with a Ryzen? Or z670 or b550 motherboard, or might it be the drivers on the Radeon? If it’s a +/-5% up or down from circuit to circuit, I think I could live with that. Great work again.
Thanks! I doubt AMD will explicitly help the Radeon dips, however, 3D V-Cache might have an influence. Unfortunately I do not have another Radeon to test and see if the same dips repeat. It could even be the case that the dips experienced by the 6700XT represent a 'performance floor' of 24.1.1 drivers and faster Radeons will also dip down to. I still have so many questions! lol
Fair game pitting them to eachother based on the similar price you payed. I would expect the 3060Ti to blow the 6700XT away. I know Tflops are not everything, but that Nvidia card has 30% raw power and a raytracing advantage. It's the better deal. Normally the 6700XT is pitted against the 3060 non-Ti.
The camera POV is probably misleading. I drive with a 90 degree bend between forearm and bicep, that's the most comfortable for me and the cars I drive at moderate FFB. I appreciate you lookin' out for me tho ;)
The benchmarks that no one else has done. Really nice! For the iRacing results, I think you are seeing the higher AMD CPU overhead that only seems to affect iRacing. Could be that is fixed now for single screen, but for triples that definitely seems to be the case, with the relatively powerful CPU used for testing masking it in some of the benchmarks.
Thanks. Nice to have this on sim racing as opposed to just games in general!
Thanks! I'm also planning to do more 'deep dive' videos into specific topics,so stay tuned!
Keep in mind that on the 6000 series Radeons, AMD reduced the Memory bandwidth a lot and then added the infinity fabric to compensate. This means that each GPU has a pixel trhougtput "ceiling" when it comes to resolution. The top of the line 6950XT had a chip very well capable of being faster than the 3090 Ti in 1080p and 1440p, but it TANKED in 4K. This goes for all the lineup. The 6700 series could do 1080p very well, but at 1440p it was already saturating it's smaller infinity fabric. So the smaller the resolution the faster the AMD card is compared to Nvidia. When you go triple screens then it only shows more and more it's limitations. Notice that AMD has a 192 bit bus while Nvidia has 256, AMD is only using VRAM chips that are twice as dense as Nvidia's, otherwise it'd be at 6GB. This was a PLANNED move to reduce cost, and Nvidia followed MAD and went the same route with the 4000 series GPUs.
Remember that screen space effects also run at full resolution and also contribute to this bottleneck, reducing their quality will increase longevity for the GPU.
Really great technical information. Thank you for sharing!
@@BenchmarkOdysseysand all completely wrong and misinformed
Based on the amount of work in this video alone, i had to subscribe and leave a comment. And also liked the video. Thank you for the video. Peace✌️
Great video, thanks for all information, really useful. Could I know what monitors are you using on triples? I'm using RTX3060Ti with a 34" Ultra wide 2K monitor but I'd like to change to triples in future.Thanks again!
Thank you :) Gigabyte M28U, it's one of the lesser expensive high-refresh 4K monitors. It definitely makes more sense to have 4K on a bigger screen size, but running three 32" or bigger is a lot of space. I move my gear around too much, it would just be too cumbersome.
I'm very surprised because I thought run triples with a 3060Ti was madness. Obviously it also depends on what CPU are you using to run it. Very interesting!!! Thanks friend. @@BenchmarkOdysseys
undervolting the 6700xt and tuning your system memory will get rid of the variance
Reasonable idea. The under-performance at Daytona at Triple 1080p in iRacing, however, remains the same with max 2700Mhz to GPU core (min 2500Mhz, -50mV), +100Mhz to VRAM with fast timings, and 15% power. What's strange about this performance (with or without settings above) is the card refusing to consume more power, averaging between 93W and 110W during the bench. This 6700XT consistently delivers these poor results during this scenario.
iRacing does reward well tuned memory timings, especially on Intel platform, and this XMP profile is well above the 'stock' settings. I'm not familiar with RAM timings specific to systems running Radeon graphics cards.
6700XT owner here, coupled with a 5600X, i can relate on strange behavior, on my tripple 1080p I go from 60 to 140 fps depending on tracks.
What is important thought, is that it is still smooth, and on my 144hz monitors I can barely tell the difference if fps meter is hidden.
I went full red team for the first time, completely fed up by Nvidia approach to gpu selling/branding the last 3 years, and I’m glad I did. I mean, what is next ? 5070 ti super ti dragon ? A meme became real this year…
Thank you for sharing :) Do you run Eyefinity as well? I plan to do more analysis with 6700XT so I encourage you to subscribe!
I'm not using eyefinity, i'm using the multi projector setting in Iracing. On asseto corsa, FPS are a bit better usually but SOL and all can really make FPS vanish.
Thank you for your info, I'm thinking to change fro. Xbox Series X to pc, any comments? Thank you.
You're welcome :) But that's a loaded question! A significant hurdle to overcome is the frustration that nothing is really 'plug-n-play' on PC for sim racing. Yes, everything is customizable... but that's by necessity and not luxury. If you have the patience to learn and tenacity to problem solve, the sim racing experience and community is a huge upgrade over Forza and console racers.
When it comes to 'What PC should I buy?', that's specific to your budget, display(s), and racing desires. I'm planning to offer an inexpensive service to help new users, so if you're interested, drop me an email (on my YT info tab).
@13:13 those commenters dont know their arse from their elbow 🤣nvidia shills come out to play. for their information, iracing like most sim games are more cpu bound than gpu bound. get a better processor, the gpu isnt the issue. i need to cap my frames to 161, max settings on my 6000 seriers
I think a 3060 Ti is more like a 5700XT than a 6700XT. At least the 3060 (no Ti) is... A "vanilla" 3070 is more similar to a 6700XT I guess.
3060 ti in Sim Racing @ Triple Screens = equal to 24 / 27 inches x 3 @ 1080p = 90+ fps all stock
will be nice to find somewhere online comparision 7800xt, 7900xt to 4070 super , 4070ti super
Traditional gaming benchmarks suggest the 3080ti is in the ballpark of the 4070 series: a bit faster than vanilla 4070 but slower than 4070ti Super. So any benchmarks of mine using the 3080ti, I would expect 4070 series to be +/- 10%.
7800xt seems really compelling because it's closest to that sweet spot of $500 USD.
Wow! This was great!! Thanks for make and share it! Your work is superb! Thansk also to include AMS2, it gives some interesting information to analize. If you need more topics for new videos I hope some day you could make something referred to VR in Simracing. Again, thanks and congrats for your video! 👏🏼👏🏼👏🏼👏🏼👏🏼👏🏼
Glad you enjoyed! I've only done one video with VR using an Oculus Quest 3. I've got big plans for a more in-depth VR analysis. Stay tuned by subscribing! :D
awesome content man, these bencharks are very good!
Thank you! I dig your channel too, and just sub'd :D
@@BenchmarkOdysseys heh thanks, although my videos are just simple game benchmarks on my hardware.
You put great amount of effort into your content, i know how long it takes to bench everything also keep track of consistent testing methodology, its especially great because sim racing benchmarks are really scarce👍
Could some of the odd results be different in Win11? isn't windowed mode optimized in that?
You did not mention it but did you use FSR in iRacing? if not, why not?
Image quality is difficult to measure, and my subjective opinion would probably be inconsistent. I only used FSR with ACC because it's a more GPU demanding title, and I figured anyone with these cards would be using it. I'd like include results with it in the future, however, I'd prefer to also show the artifacts that it creates and I dont have a high speed camera for that.
Hi, I was just curious what would you recommend out of these GPU's for the Meta Quest 3 for IRacing and Assetto corsa?
Or would you say something like the RX 6800 and 3070 would be a better choice?
Hello :) I have yet to test VR with AC, however, with iRacing there is a clear advantage with GeForce and VR. Therefore, I recommend RTX 3070 over RX 6800!
*Benchmark Odysseys* I just posted your "(unannounced) Liquid Freezer III from Arctic Cooling" vod in both the Hardware Unboxed and PCMR discords. You are welcome =>
I appreciate that :) I think the official press release is just around the corner, other leaks have popped up as well. Cheers mate!
@@BenchmarkOdysseys Just this past Monday, I ordered the ARCTIC Liquid Freezer II 420, which was scheduled to be delivered this coming Monday. But I've now cancelled the order, because I'd rather wait for the newest model. Do you have an Amazon affiliate link?
@@rcald-gz5jd Oh for sure I'd go for the new model too! I have not (yet) set up an affiliate connection with any online stores. But I'll look into this weekend. Thank you for asking, that's super cool 😎
fantastic, just the type of info i needed, i dont mind going used market, got yourself a sub, keep up the good work.
Awesome! Thanks for following up on testing cheaper GPUs. How do you think these benchmarks translate to say a 5800x3D CPU? I have a 5800x3D + 5700XT combo on triple 1440p and it's hurting for frames in some instances.
I have not used 5700XT, but from other user experiences I've read, seems like it can struggle with triple screens. Your 5800X3D is plenty quick and keep close with Ryzen 7000 series and Intel 14th Gen.
I've been looking at getting a 2080TI for it's 11gb of vram. I run triple 1080p, and my 2070 Super isn't cutting it in iRacing.
iRacing is pretty good at working with 8GB of VRAM. You'd probably see bigger performance gains by upgrading to an Ampere GPU like 3070 even though it just has 8GB. Also, many RTX 20 series users have experienced crashes with Tempest. iRacing's July1st patch is trying to address that. 3070 is 2 years newer architecture compared to 2080ti (2018 release).
Thnxs for this video it helps me a lot i was thinking going between rx 6709 xt or 4060ti so now i can got with 4060ti
Glad to help! I'll be posting more testing with this 3060ti, so be sure to sub to find out the results :D
thnx mate if u ever had a chance to do benchmark with 4060ti will be nice and thnx again
@@BenchmarkOdysseys
Will the RTX 4060TI on triples 1080p get good FPS
Yes. It is about 10% faster than my 3060 Ti results.
And for an RX 6800 (non xt) is triple 1440p too much and better go for 1080p ?
Yes, I think Triple 1440p is a difficult challenge for an RX 6800. By the time you reduced quality and increased FSR, I think the image on 1080p may look better. But image quality is a subjective experience.
awesome job, very important info as a simracer looking to buy a GPU in the coming months. Cheers!
Thank you! Im looking to bench other titles, so stayed tuned :D
which one do u recommended for vr sim racing (Dirt Rally 2, Assetto Corsa) 4060ti or 7600 xt?
New GPU? I would go with 4060ti, however, the store MSRP price is too high. You can find Used 3080 on eBay for around $400 USD.
@@BenchmarkOdysseys yes, need upgrade my old 5600xt.. thank a lot for your help!
Thx for this vidéo ! With triple screen 2k, what cobfig is the best ? 7800x3d or 14700k ? 7900xtx or 4800 super ? Thanks !
You're welcome 😁 If you enjoy multiple simulators, it seems Intel 'wins' more often then AMD, but I recommend AM5 because of future CPU support. I'm currently making a video for 4080 Super, and it's very fast and power efficient.
Thanks ! It's only for iRacing
Seems like X3D V-Cache does well in iRacing, especially around more complicated tracks. And the tracks are not getting more simple! If you don't mind waiting a little bit, rain will be launching soon and I will be doing some performance analysis with it.@@CharlesDRACOULIDES
@@BenchmarkOdysseys thank you ! Today i have i7 7700k and 2080 ti on 34 4k screen but o want switch on triple 2k screen and i look what cpu and gpu is better for iracing. 4090 seems very powerfull but expensive :( The 7900xtx can be also the good choice ? Yes, you are right for the Rain !
@@CharlesDRACOULIDES definitely upgrade the CPU first! Your 2080ti performance is between the 3060ti and 3080ti results. I think you will be surprised at the gains provided by a current-gen CPU.
I’m waiting for the rtx5080 to upgrade my 3070 card. It’s doing great for 1080p but eventually 4K would be the goal.
Curious if SMP was enabled in iracing for the 3060ti just to see if it made a difference or not?
Yup, SPS was enabled on the 3060ti. The note dashed across the screen on the first bench. Maybe I'll do a follow-up test... for example, in the benchies where the Radeon performance significantly changes from 4K to triple 1080p... does the 3060ti perform significantly worse when SPS is disabled?
@@BenchmarkOdysseys Would definitely be interesting to see that tested along with using a lower end CPU that might be more typically used with this class of GPU. Is the nVidia CPU advantage only down to SPS, or is there something else at play? Is there any difference on a single screen when using a lower end CPU?
Hi there , thank you for the bench. Did you have the ram oc.d or just xmp?
Hello, and you are welcome :) Just XMP profile for this RAM: G.SKILL Trident Z5 RGB DDR5 7200 CL34-45-45-115
I'll admit it is not stable with a memory torture test (it is at 6800), but I want to compare these results to running at a slower speed. When the 3060ti runs out of VRAM, the simulator should use system RAM... and I want to see if there's a measurable difference. Subscribe and stay tuned!
Thank you. I am since your first video. I happen to have exactly same setup but with 4090 and run a 6800 c32 Kingston ram using the buildzoid settings shared with hardware unboxed. I asked because I was wondering whether I should try a 7200 kit but it should run at c32 in order to make a difference I think
@@BenchmarkOdysseys I follow you also in the iracing forum
Oh that's great! The little 'subscribed' icon wasn't showing in my chat window, perhaps that's a bug with UA-cam Studio. Thanks for telling me! I don't want to bug people who have already signed up heh
I'm really interested in AMS2, thanks for the benchmark, It would be great if you did a CPU benchmark like you did for ACC, I'm trying to check if my 10400f is holding me back.
I'm currently using a 6750 Xt no complaints really...it gives me stable frame rate no matter what sim I play
Really interesting video for sim racers. Guessing that the results using the more current 6750xt would be around 7% - 10% higher fps than the 6700xt used in the video.
Thank you :) Yes I think you're right, and the faster memory speed of 6750xt should help at higher resolutions when it uses +10GB of VRAM.
Thanks for the tip about higher resolutions and vram @@BenchmarkOdysseys 👍
Good to see that 1440p triples for iracing on my 3060ti may suffice
Awesome racing set up! 3:00 cool to see 🏁 🏎
You say you're going to bench with AMS 2 & iRacing, but later mention Asseto Corsa gives a distinct advantage to vRam. So what are we trying to evaluate here? Also it's important to note that the 6700 series runs at 110c as "normal" based on feedback from Radeon dev team. No idea why, but it's very temp dependant.
I'm sorry for the confusion. I keep learning with each run of benchmarks. If you have a question for a specific bench, I'll address it :)
I have a 5800X3D and an 6800XT and 32” triples. With I racing, big fields at night I have a lot of trouble, way below 60FPS. Even some day time races are tough, I thought this system would be fine but it’s not.
With triples and/ or VR better go with NVidia in iRacing
I am looking to build a brand new PC to power my sim rig, what GPU would you recommend? (Running triple 1080s)
You have lots of options for Triple 1080p! Something from the 4070 lineup is more than enough, even a used 3080 would be plenty fast. The 4060Ti is also a fine card, it's just overpriced by $50 or $100. I haven't done any testing with the Radeon 7000 series (yet), so I cannot recommend until I do.
bruh....thank you for reaffirming my 6700XT buy last month, its the best bang for buck card when it comes to vram. its a 1440p card, but I game at 1080 locked at 60fps so its an absolute beast per my requirements. the 12 gb vram should keep this card relevant for at least the next 2 or so years. testing it at 4k is a bit unfair because its a 1440p card. it was never advertised at 4k.
VRAM is not as important as you think, you can find a good explanation as to why that is on the gamersnexus channel, the 6700 xt video to be precise, but it doesn't only apply to this card, it applies to all medium budget gpus that have more than 8gb
I'd be very interested to see if the results changed at all when using an AMD CPU rather than the Intel.
Me too! I'm also curious about what it looks like when a CPU bottleneck creeps in.
Fantastic work! Very well done video
Thank you :)
I'm upgrading to a 4080 super coming from a 2080 it's time. Installing with a older the week. My older CPU is running at 55% with the 2080 at 95% to benchmark( triple 32's @ 2k) , then pairing with a new build with a 78003dx. I just want to see the jumps. Thanks for all your work.
Congrats on the upgrade :) It will be a massive jump. I will be doing more testing with 7800X3D, so please subscribe and check it out.
You probably don't even know what 2K is.
An other person looking for attention. I'll appease you. LOL @@GrainGrown
@@logo8009 So you don't know. Figures.
Very informative. Subbed!
Thank you!!
Right so the question will Radeon run triples in iRacing seems to be answered. They will.
So that brings Team Red back on to the table as an upgrade path.
Where the Radeon dips inexplicably and all the performance is not being utilised throws the thought into my mind would that have a different result if the CPU and GPU were both AMD? Perhaps the Radeon might play better with a Ryzen? Or z670 or b550 motherboard, or might it be the drivers on the Radeon?
If it’s a +/-5% up or down from circuit to circuit, I think I could live with that.
Great work again.
Thanks! I doubt AMD will explicitly help the Radeon dips, however, 3D V-Cache might have an influence. Unfortunately I do not have another Radeon to test and see if the same dips repeat. It could even be the case that the dips experienced by the 6700XT represent a 'performance floor' of 24.1.1 drivers and faster Radeons will also dip down to. I still have so many questions! lol
Fair game pitting them to eachother based on the similar price you payed. I would expect the 3060Ti to blow the 6700XT away. I know Tflops are not everything, but that Nvidia card has 30% raw power and a raytracing advantage. It's the better deal. Normally the 6700XT is pitted against the 3060 non-Ti.
It was a far closer battle than I expected as well. Cheers :)
dude you are clearly sitting too close at the wheel, at least at 3:10
The camera POV is probably misleading. I drive with a 90 degree bend between forearm and bicep, that's the most comfortable for me and the cars I drive at moderate FFB. I appreciate you lookin' out for me tho ;)
The benchmarks that no one else has done. Really nice!
For the iRacing results, I think you are seeing the higher AMD CPU overhead that only seems to affect iRacing. Could be that is fixed now for single screen, but for triples that definitely seems to be the case, with the relatively powerful CPU used for testing masking it in some of the benchmarks.
Cheers! Yeah it's a strange when Radeon fumbles. I'm tempted to throw more CPUs at it and see how 6700XT scales as single core performance changes...