Main benefit of 8000/2000 is that you can undervolt soc heavly so cpu is more power efficent+ generates less heat so core clock run little higher if using pbo (ryzen are throttling past 60 temp). The 8000 should be only faster if you have poor cpu cooler, have 7800x3d with run much hotter or cant run stable 2133fclk.
Someone said its because the higher u go in frequency, the less ur able to push ur cpu to compensate the ram oc. Best is still 6000-6400 with fclk match and tigher timings
I'm curious on difference at 6400 of having FCLK at 2133 (3:2 ratio) or 2200. I can get both stable but went with 2133 because of the latency decrease.. curious on your thoughts.
When UCLK and FCLK are set at a 3:2 ratio, the SyncFIFO buffer operates, compensating for latency in the process. Refer to the following link for more details: www.reddit.com/r/overclocking/comments/18z4rm9/some_fresh_zen4_ramif_overclock_scaling_data/.
@코코33 I've seen the wonderful data here already, thank you! I was just curious about gaming performance as I've not seen a direct comparison between 2133 and 2200 apart from latency and synthetic benchmarks.
Hey man nice video, but quick question if you dont mind. I assume you manually tweak both configs right? If thats the case, did you only bother touching the CAS latency or did you tweak the subtimings as well? like can we get a full description of both configs from zen timings? I would appreciate a lot if you could provide this much context because I find your test quite useful.
Nice to meet you! I also referred to another tester's configuration. You can check the details at this link: zod.kr/cpu/1456874#comment_15... I’ve already obtained permission to share this resource, so feel free to use it as you like. However, if you plan to share it further, please share the link to give credit to the original creator. Have a great day!
This is pure gold, almost nobody does this... its a close call but basically staying in sync with your fclk is the way to go. 6200-6600 1:1 depending on your IMC quality.
Thank you for your kind words! Recently, there’s been a lot of information being shared, but no comprehensive comparisons or tests between different settings. I was losing sleep at night, wondering which setup would be better. I’m so glad to hear that my data was helpful!
@@코코33 Yes a lot of people pulling shit out of their ass claiming 8000Mhz is better when you really understand computer logic and how they work beyond the surface level most people tend to argue on, 8000Mhz doesnt make sense to be faster in gaming than 6400Mhz CL28. Using logic and theory, you need to prioritize as low as possible latency, with the most bandwidth you can pack with, if you sacrifise the latency for bandwidth this should be the result you expect logically, 8000Mhz being slower. Its kind of the same thing of saying "Oh if I upgrade my internet speed from 100Mbs to 1Gigabit I will get better ping" but the reality is that ping doesnt care about raw speed, it transfers very little data in actual practise what makes ping slow or fast is the total length it has to reach between the server and the computer + other variables of course but that's the biggest factor imo, thats why "LAN" parties are between 0-1ms because the server and the client are practically in the same network/next to each other. Cool video and thanks for uploading!
@@sadasd-n2f That's an excellent analogy. I, too, had only a vague understanding, but your explanation has really clarified my thoughts. Thank you for sharing your insights.
Main benefit of 8000/2000 is that you can undervolt soc heavly so cpu is more power efficent+ generates less heat so core clock run little higher if using pbo (ryzen are throttling past 60 temp). The 8000 should be only faster if you have poor cpu cooler, have 7800x3d with run much hotter or cant run stable 2133fclk.
Someone said its because the higher u go in frequency, the less ur able to push ur cpu to compensate the ram oc. Best is still 6000-6400 with fclk match and tigher timings
그래서 뭐가 더 좋나요 ? 낮은 cl 값 ? 높은 클럭 값?
스윗 스팟이 6400 이라던데
와우 대박이네요 ㅋㅋ 저도 드래곤볼 도착하면 램오버 할건데 ddr5는 처음이라 기대되네요. 저거 램 뱅크24x2 6800로 판매되는 제품 맞나요? 마침 똑같은램같아서요.
6400에 cl28이면 상당한 성능이고 전압도 꽤 들어갔을텐데 전압은 얼마주셨는지 스팟쿨링은 하시는지 여쭤봐도될까요?
@@달총-j1c 스팟쿨링은 안하고, 전압 1.54까지 줬습니다. 실내 온도 올렸더니 백그라운드 프로그램 돌아갈 때 게임이 간헐적으로 오류나는 문제 있어서 현재는 추가 테스트 중입니다.
@@코코33 감사합니다 ^^ 저도 배송 다오면 얼른해봐야겠어요 ㅎㅎ
🫡👍👍👍
I'm curious on difference at 6400 of having FCLK at 2133 (3:2 ratio) or 2200. I can get both stable but went with 2133 because of the latency decrease.. curious on your thoughts.
When UCLK and FCLK are set at a 3:2 ratio, the SyncFIFO buffer operates, compensating for latency in the process. Refer to the following link for more details: www.reddit.com/r/overclocking/comments/18z4rm9/some_fresh_zen4_ramif_overclock_scaling_data/.
@코코33 I've seen the wonderful data here already, thank you! I was just curious about gaming performance as I've not seen a direct comparison between 2133 and 2200 apart from latency and synthetic benchmarks.
Hey man nice video, but quick question if you dont mind.
I assume you manually tweak both configs right? If thats the case, did you only bother touching the CAS latency or did you tweak the subtimings as well? like can we get a full description of both configs from zen timings? I would appreciate a lot if you could provide this much context because I find your test quite useful.
Nice to meet you! I also referred to another tester's configuration. You can check the details at this link: zod.kr/cpu/1456874#comment_15...
I’ve already obtained permission to share this resource, so feel free to use it as you like. However, if you plan to share it further, please share the link to give credit to the original creator. Have a great day!
@@코코33 Thank you
모니터 해상도 어떻게 되시나요?
QHD 입니다
This is pure gold, almost nobody does this... its a close call but basically staying in sync with your fclk is the way to go. 6200-6600 1:1 depending on your IMC quality.
Thank you for your kind words! Recently, there’s been a lot of information being shared, but no comprehensive comparisons or tests between different settings. I was losing sleep at night, wondering which setup would be better. I’m so glad to hear that my data was helpful!
@@코코33 Yes a lot of people pulling shit out of their ass claiming 8000Mhz is better when you really understand computer logic and how they work beyond the surface level most people tend to argue on, 8000Mhz doesnt make sense to be faster in gaming than 6400Mhz CL28.
Using logic and theory, you need to prioritize as low as possible latency, with the most bandwidth you can pack with, if you sacrifise the latency for bandwidth this should be the result you expect logically, 8000Mhz being slower.
Its kind of the same thing of saying "Oh if I upgrade my internet speed from 100Mbs to 1Gigabit I will get better ping" but the reality is that ping doesnt care about raw speed, it transfers very little data in actual practise what makes ping slow or fast is the total length it has to reach between the server and the computer + other variables of course but that's the biggest factor imo, thats why "LAN" parties are between 0-1ms because the server and the client are practically in the same network/next to each other.
Cool video and thanks for uploading!
@@sadasd-n2f That's an excellent analogy. I, too, had only a vague understanding, but your explanation has really clarified my thoughts. Thank you for sharing your insights.
램오버 진짜 ㅈ또 의미없구ㅏ