Star Citizen 7700x, 5800x3d, Stock vs tuned 13900K, 10600K, featuring a Subscriber

Поделиться
HTML-код
  • Опубликовано: 30 июл 2024
  • The ultimate CPU analysis for star citizen 2023. Tuning is King
    ►Join the Game Room: www.game-day.xyz/
    Subscriber System Specs:
    CPU: 10600K
    RAM: Kingston Hyper X 3600Mhz CL 17
    GPU: RX 6800
    Motherboard: Asus Z490-p
    Storage: SSD m.2 Samsung 1T, Western Digital SN550
    CPU Cooler: Corsair H100x
    Power Supply: Coller Master 700W mpe-7001 acabw
    Monitor: LG OLED C2 48 Inch 120Hz
    System Specs:
    CPU: 13900KF All Cores at 5.8GHz
    RAM: Viper 32GB 4000MHZ CL16 @ 4100MHZ CL16 [Tuned]
    GPU: Nvidia RTX 4090
    Motherboard: MSI MAG Z690 TOMAHAWK
    CPU Cooler: EK Water Blocks EK-AIO 360 Basic
    Power supply: Be Quiet straight power 11 1200w
    OS M.2 NVMe: Crucial P2 1TB M.2
    Game M.2 NVMe: WD_BLACK SN770 2TB
    Case: Lian Li O11 Dynamic EVO
    Monitor: LG OLED C2 42 Inch 120Hz
    Gameday Graphics settings
    Resolution: 1080p
    Quality: Very High
    Scattered Object Distance: High
    Terrain Tessellation Distance: High
    Planet Volumetric Clouds: Very High
    Field of View: Max
    Motion Blur: Full
    Chromatic Aberration: Off
    Film: no
    ‪@tenpoundfortytwo‬
    CPU reference: • WOW! But Don't Buy One...
    Benchmarking Run: • Video
    GPU: GPU: RTX 3080/ 4080
    CPU: 5800X3D, 13700KF, 12700KF, 3600, 2600, 1600
    Graphics settings:
    Resolution: 1080p
    Quality: Very High
    Scattered Object Distance: Medium
    Terrain Tessellation Distance: Medium
    Planet Volumetric Clouds: Medium
    Field of View: Max
    Motion Blur: off
    Chromatic Aberration: Off
    Film: no
    CHAPTERS
    ---------------------------------------------------
    0:00 Into
    0:40 Subscriber Reaching Out
    1:05 Subscriber system specs
    2:00 Remote Tuning Methodology
    2:28 Windows Tuning and Optimisation
    3:01 CPU Overclock
    3:34 GPU Overclock
    3:54 RAM Tuning
    4:56 Lorville benchmarking run and graphic settings
    5:49 Benchmarking tuning results
    6:40 Benchmarking optimum results
    7:36 Tenpoundfortytwo graphic settings and Lorville benchmarking run
    8:49 Tenpoundfortytwo vs Gameday run and graphics comparison
    9:26 Tenpoundfortytwo CPU results vs Subscriber 10600K
    11:18 7700x, 5800x3d vs Tuned 13900K
    13:43 Conclusion and recommendation, wait for 7800X3D?
    Hope you enjoy the video. Comment with your thoughts about it. Don't forget to hit like, subscribe and share the video if you can.
    #13900k #7700 #5800x3d #4090 #starcitizen #overclock #rtx4090 #pcgaming
  • ИгрыИгры

Комментарии • 85

  • @mrch4nce
    @mrch4nce Год назад +3

    This is an amazing video and very well done, thank you

  • @demiadeyemi4434
    @demiadeyemi4434 Год назад

    Great video!! New to star citizen so really appreciating the comprehensive content and focus on performance benefit. Best in the game :)

  • @jimashby43
    @jimashby43 Год назад

    Thank you. Good helpful video. You have also been a nice addition and thought provoking person on 42s discord also. You've made a lot of this much easier to understand.
    SC isn't really for a person that bought a prebuilt and thinks they will never have to touch it. There are a lot of things to be done to have a good time with Star Citizen.

  • @ghenghisprawn6034
    @ghenghisprawn6034 Год назад +6

    Great job showing the impact of tuning. I always tell people there's a nice chunk of performance to be had in Star Citizen by tuning the ram, the subtimings are just as important as the main "advertised" timings when it comes to getting the most out of it.

  • @sphaera3809
    @sphaera3809 Год назад +1

    It's great to see more SC optimization and benchmark content. Thank you for taking the time putting this video together. More performance can be achieved from Ryzen CPUs by tweaking PBO and tweaking thermal and power limits. I just moved to Ryzen 7000 with a 4080 OC GPU and things are better in SC for sure. I can sometimes get 3 digit FPS in 4K. My 7900x is now set to -20 in curve optimizer with a 170W PPT limit. And no gains in memory beyond 6000MHz, only timing can be optimized. But I'm still testing and will do a few Lorville runs on 3.17 to compare with your results. 3.18 seems to put less demand on CPU and GPU. Better FPS in SC for free should be just around the corner. 😃

  • @PatEagle787
    @PatEagle787 Год назад

    Very informative video. Thank you very much for the time you put into this. You've earnt a new subscribber. Hopefully you'll provide us with some tuned DDR5 numbers on the Intel platform and Zen 4 X3D numbers on the AMD side as well as following the evolution of this game performance with the upcoming implementation of the Vulkan API.

  • @TerraWare
    @TerraWare Год назад

    Very nice. I recently got to Star Citizen and am using a 7700X with 32Gb 6000mhz CL30 ram and RTX 4090. I don't know much about tuning my ram and seems quite intimidating because of the many different settings etc.

  • @mrm7058
    @mrm7058 Год назад +3

    @13:50 We have to wait until April for the 7800X3D. And since it seems that Star Citizen is very latency sensitive, I am not sure if a 7900X3D or 7950X3D would be a good idea, because you may get latencies between the chiplets there.

    • @gamedaytoday1
      @gamedaytoday1  Год назад +4

      100% agreed the reason why AMD is launching the 7800X3D later in april is because they don't want customers to buy it over the 7900X3D or 7950X3D as it is probably as fast or faster.
      And after 5 weeks the return window of 7900X3D or 7950X3D is done anyways a bit sneaky from AMD

  • @ahblooloo8639
    @ahblooloo8639 Год назад

    Thanks, would love to see a similar guide for "single player Star Citizen" aka X4: Foundations, where FPS can vary widely and very dependant on memory timings and CPU.

  • @meldepinda
    @meldepinda Год назад

    Oh also, is it possible to display error bars when you display the averages? Would be neat how much variation there is between the 3 runs

  • @123TheCloop
    @123TheCloop Год назад +1

    so just a heads up, why are you using W10 for both Intel and AMD CPUs? they dont have the updated schedular that W11 has for the 12/13th Gen and 5xxx/7xxx processors, Intel make this abundantly clear on the 12/13th gen stuff, and also are you using 3.18? this has the new Gen12 render which helped remove some of the thread limitations found in 3.17.5 so if anything ide expect performance to jump for pretty much all tests here assuming no other issues present.

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      Windows 11 has higher system overhead (probably due to the updated scheduler) which affects the ram latency by 1 to 2 ns

    • @123TheCloop
      @123TheCloop Год назад

      @@gamedaytoday1 that might be the case but W11 still offers higher performance on 5000/7000 by a significant margin, with 12/13th Gen CPUs being somewhat up for debate with the recent E and P cores approach, loss in performance from the higher latency increase might still be outweighed by increase in CPU performance. would be nice too see a comparison if at all possible and potentially a 3.17 vs 3.18 for the Gen12 improvements (if any) :)

  • @jesuschrist2284
    @jesuschrist2284 Год назад

    What do you suggest tor a oc 6600k & 1070 combo? ;)

  • @xxmeanyheadxx
    @xxmeanyheadxx Год назад

    I have a 7700x only undervolt (1.2v) to keep cool, 6950xt stock with 100% fan cuz my speakers cant be bothered by fan noise, 32gb ddr5-6000 flare expo 1/low latency/high bandwidth, i havent seen less than 70fps anywhere in game after the first hour of gameplay, it's always over 100 in my ship. i game on 60fps (big big ass oled) tv so it's pretty nice :) maybe if i get vr or high refresh monitor someday i will tinker but for now i'm just amazed at new technology. this is the first rig i've built since my fx-4350 LOL my mind is blown, seeing schematics of amd architecture reminds me a lot of why i quit being a mechanic: too complicated! (i miss the good old car days with only one computer: fuel injection, no traction or nothing lol) the whole infinity cache thing looks like a 2022 mercedes engine bay with no plastics can i get an amen

  • @meldepinda
    @meldepinda Год назад +1

    heya loving the video and subbed! One question, how much would it cost for someone to request tuning? Looking to upgrade to a 7000x3d series (after reviews are out) and a 4080 + some decent kit in like 1-2 months.

    • @gamedaytoday1
      @gamedaytoday1  Год назад +1

      Send me a email if interested in Tuning

    • @meldepinda
      @meldepinda Год назад

      Gotcha, will do when I got all the kit!

  • @chiron03
    @chiron03 Год назад +3

    I have a 7700x running at 5.5ghz all core with an RTX 4090 OC'd +200core and +1000mem. With DDR5 6000mhz CL30 ram. That I could add into this mix of rigs.

    • @gamedaytoday1
      @gamedaytoday1  Год назад +1

      Awesome what results do you get when doing the Tenpund run?

  • @SuperEjizz
    @SuperEjizz 2 месяца назад

    i have a 7700x and was wondering if the x3d cpus were worth it for star citizen or if the 7700x has what it takes

    • @gamedaytoday1
      @gamedaytoday1  2 месяца назад

      If you tune the 7700x it’s better than a tuned 7800X3D in star citizen ruclips.net/video/N159yBADy1I/видео.htmlsi=tIVEbH5eaTElyBZj

  • @fullergaming1354
    @fullergaming1354 Год назад

    I have a 5800x3d and a 4090 and i dont get the numbers most times. Like in loriville i top out at 40-60 tops when doing the run that you mentioned in the video. Can you help me out?

    • @gamedaytoday1
      @gamedaytoday1  Год назад +1

      I will get the 5800x3D soon and find the tuned settings for it, then after that I can help.

    • @fullergaming1354
      @fullergaming1354 Год назад

      @@gamedaytoday1 awesome thanks

  • @Hyperion1722
    @Hyperion1722 Год назад +2

    13900K OC with 4090 is not apples to apples comparison with 7700x with 4080 that's why their is a big difference. Could possibly match 13900K if 4090 is also used. Performance uplift could be significant when switching gpu from 4080 to 4090.

    • @gamedaytoday1
      @gamedaytoday1  Год назад +3

      Incorrect I have the 7900 XTX which is like a 4080 and it get the same FPS as the 4090 in 1080p so there is no performance uplift by using the 4090.
      mentioned
      The 13900K OC with 4100mhz ram can push Higher FPS to the GPU
      While the Stock 7700x can push less to the GPU.
      If I am wrong please show me some benchmarks proving a 4090 will improve performance over a 4080 when the system is CPU/RAM bottleneck

    • @DKTronics70
      @DKTronics70 Год назад +1

      @@gamedaytoday1 This is nonsense. The fact that you had to use a TUNED, far more expensive 13900k, pulling nearly double power than a stock 7700x, all the while the 7700x has a slower GPU. Tune the 7700x, and add the 4090, then we'll see different results. You come across as a typical Intel fanboy.

    • @gamedaytoday1
      @gamedaytoday1  Год назад +1

      @@DKTronics70 Here is a video on why a more powerful GPU does not improve performance of the CPU/RAM when your limited by the performance of those components. ruclips.net/video/pZHTMMpjJLQ/видео.html If I had the 7700x I'd tune it too. Will get the 5800X3D next week and tune it to see how far it will go. I am hardware agnostic. The purpose of the video is to show that tuned systems go further than plug and play, if your conclusion is that I am a fan boy then it seems like you missed the whole point.

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      ruclips.net/video/pZHTMMpjJLQ/видео.html , No performance uplift when swapping GPUs in CPU/RAM bound scenarios

    • @younglingslayer2896
      @younglingslayer2896 Год назад

      @@DKTronics70 jesus christ

  • @MrGatz
    @MrGatz Год назад

    I suppose that if you game SC at 4K you can run a 7700x and get the same performance as the top chips?

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      nope there are alot of areas in space even in 4K where the GPU (4090) is not 100% utilized, in the 50% range. Just look at the video here at Area 18 4k same performance as 1080p ruclips.net/video/ESMyu4wuT4Q/видео.html

    • @MrGatz
      @MrGatz Год назад

      @@gamedaytoday1 So what do you recommend? i’m gaming SC in 4k (4090) i can buy the 7700x for 300$ usd or wait and (maybe get the 7800x3d) money’s no object but I don’t want to be wasteful. What do u recommend? 80% gaming and 20% productivity)

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      @@MrGatz 7800x3d

  • @rust2156
    @rust2156 Год назад

    ok just fyi never oc AMD GPUs with MSI afterburner. I have a feeling that GPU is clock stretching, almost 3Ghz on a rx 6800 is sketchy

  • @ItsJustVV
    @ItsJustVV Год назад

    Hey GameDay, can you please tell me the exact model name/code of the Viper 32GB 4000MHZ CL16 memory kit?

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      PVS416G400C6K: amzn.to/3YI62rn

    • @ItsJustVV
      @ItsJustVV Год назад

      @@gamedaytoday1 That one you linked is the 16GB kit. I was referring to the 32 GB one. Also is it Dual Rank too?
      I already have a 16GB kit myself, but I need a good 32GB B-die kit to upgrade to.

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      @@ItsJustVV amzn.to/3XMeKDm any 32GB pair sticks with 3600Mhz with CL 14-14-14 are B-die

    • @ItsJustVV
      @ItsJustVV Год назад

      @@gamedaytoday1 Thank you, that's what I was looking for. 👍

    • @ItsJustVV
      @ItsJustVV Год назад

      @@gamedaytoday1 Are these hard to OC at 4100 C16 with a RPL CPU?
      I have a 13600K with the kit you mentioned 1st, the 16GB 4000 one, but I want to upgrade to 32GB.
      But this one being at 3600, I suppose it can get to 4000+ right?

  • @Pabula
    @Pabula Год назад

    You think the 13900k with 8000mhz ddr5 Gear2 will beat a 13900k 4100mhz ddr4 gear 1 on Star Citizen?

    • @gamedaytoday1
      @gamedaytoday1  Год назад +2

      Yes the question is how much will find out soon

    • @jjlw2378
      @jjlw2378 Год назад +2

      13900k/8000mhz+ DDR5 will certainly beat 13900k/4133c15 but by how much and for how much extra cost. 8000+ DDR5 requires a 13900k with a good IMC, an expensive 2 dimm mobo like a Z790 Apex, and very expensive 7600+ DDR5. Meanwhile, pretty much every 13900k can hit 4133c15 and even the cheapest ddr4 mobo can achieve it. I'm definitely excited to see how much faster the DDR5 actually is though!!!

    • @jameyt1
      @jameyt1 Год назад +1

      ofc but I can't even hit 8000mhz on a z790i edge, most I can do is 7800mhz. It might possibly be the ram kit I bought tho, I didn't buy the 1.4v 7600mhz kit, I bought the 7200mhz kit which people say is a worse bin. you def need a top tier mb+cooling and a binned CPU to get there

  • @lostnwalmart9110
    @lostnwalmart9110 Год назад

    I have a 7900x, DDr5 6000 with a 4090 if you want to help me max this thing out we could do a better comparison I'm sure.

    • @gamedaytoday1
      @gamedaytoday1  Год назад +2

      I will get the 7950x3d soon and DDR5 once I get my hands on this will be able to tune other Rysen systems remotely. Currently working on the 7900 XTX vs 4090 video

  • @thiagodelgado3128
    @thiagodelgado3128 Год назад

    Do you use klipsch monitor?

  • @TheEVEInspiration
    @TheEVEInspiration Год назад

    Clearly some games like SC are very latency sensitive.
    Meaning they see a lot of cache misses due to the game code/data needs are not being predicted well by the CPU. This means either the cache is really to small for the problem (unlikely these days) or the game code/data-structures are not well tuned for modern architectures.
    CPUs love and need predictable code/memory access as their prefetchers are then able to hide most memory latency, resulting in better performance. But programs need to be written with this in mind, sequential memory access for example is blazingly fast, but random access is not.
    There is a lot programmers can do in this area with respect to the algorithms used, the compactness of their data-structures and how they are accessed. Categorized and/or sorted pools of game elements are much easier to work with as they eliminate slow code branching as well as being easier to pre-fetch and they are more compact on top of this all (need less cache to begin with).

    • @freelancerthe2561
      @freelancerthe2561 Год назад +1

      But are the cache misses on code, or caused by how assets are handled? I've seen a few other games that have similar quirks in performance, and all of them involve huge asset pools and (practically) monolithic main threads. Thinking about it..... the size of the main thread probably is the reason for the variable branching it has to do, based on how many entities and assets its juggling. I think I might had answered my own question.

  • @IM0001
    @IM0001 Год назад

    I did pretty much the same process (Cell Phone Camera Bios Fun and all) for another SC RUclipsr a while back on his modest AM4 build. It went from slow and crashing pretty consistently, to quite playable even though the specs are rather dated compared to current gen hardware. It is amazing how with just some proper tuning, SC can be made to run much, much better on all sorts of hardware.
    It even runs semi playable on the Steam Deck, and amazingly, each Gen12 bit that gets updated improves on that little by little as well.

  • @LekkaBier
    @LekkaBier 4 месяца назад

    Why is the most important part is missing here?
    He has only 16GB RAM. Everyone knows that SC is running much better with 32GB RAM.

    • @gamedaytoday1
      @gamedaytoday1  4 месяца назад

      Send him 16 GB plz thanks 🙏🏽

  • @Syron21
    @Syron21 Год назад

    I highly doubt that 12700k and 13700k have the same performance.

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      The data is from the other RUclipsr, arguably his results are low due to his system not being tuned

  • @EvilStromboli
    @EvilStromboli Год назад +1

    Or CIG could just optimize Star Citizen better.....if its the only game in my library needing RAM tweaks......

    • @gamedaytoday1
      @gamedaytoday1  Год назад +3

      Good point they can just remove all the game logic that makes the gameplay features, so you can get better performance.

    • @EvilStromboli
      @EvilStromboli Год назад +2

      @@gamedaytoday1 Clearly all my fault. BRB OCing my mouse

    • @gamedaytoday1
      @gamedaytoday1  Год назад +3

      @@EvilStromboli Naa you need to OC your headphones

    • @gamedaytoday1
      @gamedaytoday1  Год назад +3

      Most open world games benefit from RAM tuning like the Witcher 3, GTA, BF series, COD, etc.

  • @sjoervanderploeg4340
    @sjoervanderploeg4340 Год назад

    Star Citizen doesn't make a great benchmark, this is because it uses a CPU bound render thread that takes as much time if not more than the main thread...

  • @heezflash
    @heezflash 2 месяца назад

    And dont it i have 5800x3d VERY tuned memories 32 gigas 3600 and dlss on the lowest quality im getting 40 fps in lorvill AND TONS AND TONS OF stuyters allllll the time its unplayable

  • @mud8723
    @mud8723 Год назад

    well ofc an i9-13900k with a 4090 will do better than a ryzen 7 7700x with a 4080

    • @gamedaytoday1
      @gamedaytoday1  Год назад +1

      The 4090 only has 50% GPU utilization and the 4080 is also CPU BOTTLENCKED so it is comparing which CPU can push the highest frame

    • @mud8723
      @mud8723 Год назад

      @@gamedaytoday1 ok so wouldnt it make more sense to stuff the 3080 into each pc if its ab which cpu pushes the highest frame

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      @@mud8723 It does not mater I just got the 7900 XTX, and it gets the same FPS as the 4090 in 1080p due to the CPU/ram Bottleneck but the 4090 is 75% faster in 4k

    • @mud8723
      @mud8723 Год назад

      @@gamedaytoday1 that makes even less sense
      if ur testing to see bottleneck, even if 7900xtx and 4090 have same fps bc of the bottleneck, the data has nothing to stand on because too many things are being changed. if everything was tested with a 4090 finding which one has the biggest bottleneck makes sense, but if one is tested with a 4090 and the other is tested with a 4080 (basically this graph) it wouldnt work bc for all we know the 7700x could have more bottleneck with a 4090. or it could even be strong enough to handle the 4090, probably, but i dont know that one for sure
      also whats with the 4k thing? if thats going into the graph thats not even related to ur cpu bottleneck experiment

  • @jasonhenninger8220
    @jasonhenninger8220 Год назад

    Too many variables. Keep pretending that swapping around different GPUs isn't important. It is. That's factual. Redit told you the same thing.

    • @gamedaytoday1
      @gamedaytoday1  Год назад +3

      Well I just got the 7900 XTX, which is slower than the 4090 by a lot, its 40% slower in 4k. The 7900 XTX gets the same performance in 1080p as the 4090 in 1080p on my system. So unless there is any evidence that show swapping GPUs increases CPU/RAM performance or removes CPU/RAM bottleneck then please let me know

  • @scarumanga
    @scarumanga Год назад

    Star Citizen is a hot mess.
    RTX 4080, 13600k with no power limit, 64gb 5600mhz ddr5.
    What this game really needs is DLSS 3 and Frame Generation.

    • @jameyt1
      @jameyt1 Год назад

      if you have the money to burn buy an apex board and 7600mhz ram

    • @PatEagle787
      @PatEagle787 Год назад

      DLSS 3 is a high latency mess. Only good for single player games and Star Citizen isn't one of them. I don't want high latency garbage added to my mouse, keyboard head tracker and flight stick inputs.

  • @rooster1012
    @rooster1012 Год назад

    One of the worst system comparisons I have ever seen in a video, 4080 0n one system and a 4090 on another obviously amauter hour at its best.

    • @gamedaytoday1
      @gamedaytoday1  Год назад

      I made this video for you, if you had commented earlier I would have included you in minute 3:34: ruclips.net/video/pZHTMMpjJLQ/видео.html