實際上應該是 11 代 Rocket Lake 和更舊款的 CPUs 真的十分弱,幸好近期全靠 Lisa Su 博士的功勞才能夠令 Intel 自我反省:
https://audiophilestyle.com/forums/t...omment-1016777
引用:
Yes, I'm not aware of any hardware that can do EC modulators at DSD512. Number of cores don't help on that. Getting more clock speed does. Since DSD256 requires about 4 GHz clock speed and doubling rate doubles the load, you'd need about 8 GHz CPU clock to get DSD512...
https://audiophilestyle.com/forums/t...omment-1046172
引用:
With current CPU architectures you would need to have about 8 GHz clock speed to have DSD512 with EC modulators.
https://www.ptt.cc/bbs/Headphone/M.1...746.A.0BE.html
引用:
Intel 12 代 Alder Lake CPU 性能理想,可以不用 GPU 做 offload 就能用 ASDM7ECv2 調變到 DSD512。
而且那是 14nm 和 10nm 兩者之間的分別,實屬十分不公平的比較。
RAM 方面的分別:
https://www.sohu.com/a/507863448_126772
引用:
在時序和延遲上DDR5看上去要比DDR4大很多,但內存時序要換算成延遲時間,在同頻率下的內存條才能直接對比時序,例如3600MHz 的 c16延時是 16 / 2000MHz = 8.8ns,2400MHz 的 c14延時是 14 / 1200MHz = 11.6ns,而4800 cl40則是16.7ns,從指標上來說,DDR5延時並不會比DDR4大很多,在實際使用中對於系統性能的影響沒有帶寬那麼大。
每一代新內存推出時,都會遇到相同的問題,DDR3 1600時序一般為C12,DDR4 2666的時序大部分為 C19,而DDR5看上去時序指標不如DDR4內存,但是DDR5的架構改變、通道增加、頻率提升、電壓降低等特性使得DDR5在技術和提升空間上要比DDR4好很多。
對於消費者而言,每一代新的內存條推出因為供給側的原因,價格總是比上一代的內存條貴,但是隨著未來支持DDR5的主板越來越多,DDR5內存鋪貨及價格也會逐漸回到一個比較均衡的水平。
其實除了華碩 ROG MAXIMUS Z690 APEX 之外也沒有更理想的(同價位)選擇,官方(前所未有的第一次)僅僅標榜它超低的 latency,而且 24 相供電的每一相也高達 105A,合共 2520A:
https://audiophilestyle.com/forums/t...comment-941760
https://audiophilestyle.com/forums/t...comment-961068
https://audiophilestyle.com/forums/t...omment-1014096
引用:
As previously posted, the gaming motherboards I tested seem to have an edge over non-gaming motherboards presumably due to better VRMs and use of multi-layer PCBs with more copper in the trace paths resulting in better isolation, grounding, and power delivery. Because VRMs require real-estate, the ideal sized board would be a full sized ATX motherboard.
https://audiophilestyle.com/forums/t...comment-961583
引用:
I compared 2 Asrock boards side by side and the board with the better VRM sounded better to me. Whether this is specifically because of the VRM or some other variable is not entirely clear but it would make sense that the quality of the VRM "might" impact SQ since the VRM is responsible for the stability of the power that the CPU sees. As to whether more power phases also results in better power stability, this is not always true since the quality of the power phases matter just as the number of power phases.
Your Extreme4 board is actually a 10+2 design meaning 10 power phases are dedicated to the CPU while 2 power phases are dedicated to the integrated GPU and so only 10 of your power phases have significance. Your particular board uses SinoPower SM7431EH MOSFETS which are rated at 25A each. The Asrock Z390 Phantom 9 is a gaming ATX board that uses the same 10+2 design but uses the higher-end Texas Instruments NexFETs which are rated at 40A. The motherboard I have decided to focus on for now is the Asrock Z390 Phantom Gaming ITX/ac which is a mini-ITX board and because of its smaller size, it incorporates only a 5+2 design but this is where things get deceiving because this board utilizes higher quality power phases comprised of the Intersil Smart Power ISL99227 which many consider to be among the best and are rated at 60A each, more than 2X the current capacity of those used in your board. Regardless, each of these boards should be able to easily handle something like an i9-9900K that isn't being overclocked.
This illustrates the advantage of boards designed for gaming as they generally use better parts, especially with regards to power delivery. Another example, your board is only a 4-layer design while my board uses 8-layers which in theory, provides better isolation. Also, my board, even though it is smaller than your board uses a total of 8oz of copper in the traces resulting in better conductivity. If someone is building a server from scratch, it would be worthwhile to investigate the gaming boards. They don't cost that much more.
[Official] Intel Z690 Motherboard Roundup (OCN Edition)
https://www.overclock.net/threads/of...ition.1794800/
[2520A MADNESS] Z690 VRM Discussion Thread 27/10/21
https://www.overclock.net/threads/25...10-21.1794798/
Z690 VRM
https://docs.google.com/spreadsheets..._jXnMf2jY/edit
https://www.matrox.com/en/video/apps...wnload/?id=620
引用:
Compatible with newer systems that use UEFI system BIOS
https://www.matrox.com/en/video/apps.../previous/bios
引用:
G550 LP PCIe, G550 PCIe
https://ftp.matrox.com/pub/mga/archi...300_02_001.exe
沒有任何 VGA/DP/HDMI 輸出的主機板就需要亮機卡,而它的 Video BIOS 必須支援 UEFI:
https://item.taobao.com/item.htm?id=41390893636
https://item.taobao.com/item.htm?id=42387653003
https://item.taobao.com/item.htm?id=594936164866
https://item.taobao.com/item.htm?id=612811953880
https://item.taobao.com/item.htm?id=624734638423
https://item.taobao.com/item.htm?id=660962162920
節錄一些比較有趣的討論,基本上都是關於 Intel Optane 的:
https://www.ptt.cc/bbs/Headphone/M.1...210.A.286.html
https://www.ptt.cc/bbs/Headphone/M.1...500.A.8AB.html
https://www.ptt.cc/bbs/Headphone/M.1...644.A.2E0.html
https://www.ptt.cc/bbs/Headphone/M.1...758.A.4A1.html
https://www.ptt.cc/bbs/Headphone/M.1...466.A.965.html
https://www.ptt.cc/bbs/Headphone/M.1...683.A.D67.html
https://www.ptt.cc/bbs/Headphone/M.1...757.A.ABB.html
https://www.ptt.cc/bbs/Headphone/M.1...030.A.8D3.html
https://www.ptt.cc/bbs/Headphone/M.1...434.A.5E7.html
https://www.ptt.cc/bbs/Headphone/M.1...558.A.FC1.html
https://www.ptt.cc/bbs/Headphone/M.1...844.A.1FF.html
https://www.ptt.cc/bbs/Headphone/M.1...047.A.7E7.html
https://www.ptt.cc/bbs/Audiophile/M....695.A.ABC.html
引用:
900P就算用轉接線進M.2,聲音也比普通M.2 SSD好ww
引用:
900p的半高pcie和U.2差異,U2會稍微活躍、分離度好些,反義就稍微沒這麼厚、沒這麼穩
就類似於545走sata這樣
引用:
然後講ramos這個屁話,事實上主機板只要PCIE插槽上面插東西就會影響聲音,即使你用ramos,我只要pcie槽上面插900P一樣也會讓出來的聲音不同
而未經優化的 Windows 或 Linux 只會變成 Optane 的瓶頸:
引用:
之前就是900p雖然細節多音響性好但會薄聲不耐聽,現在就沒這個問題,換句話說900p目前沒有弱於970的點了
實際上官方公佈的 latency 數字好像全都是相當接近:
https://www.intel.com/content/dam/ww...900p-brief.pdf
https://hexus.net/tech/news/storage/...est-ssds-ever/
引用:
For example “exceptional random storage performance of up to 575K/550K IOPs (4K random reads/writes), and is complemented with low latency of less than 10μs,” says Intel’s official product page.
https://www.intel.com/content/www/us...ies-brief.html
引用:
Read/Write Latency (average sequential): < 6.75 μs / < 18 μs
https://www.storagereview.com/review...-memory-reviewhttps://lenovopress.com/lp1484.pdf#page=2
引用:
Ultra-low I/O latency, with a read/write latency of as low as 5 µs.
但是 RAID 的話就或許擁有無限的可能性?
https://thetechrevolutionist.com/201...nder-aero.html
引用:
Now the question, what about speed & latency? Did it improve? The answer is a whole hearted yes. Both latency & speed though not everything improved. High que depth tests did not improve but neither did they lose much either. Latency & que depth one transfers improved dramatically as did sequential writes, things that are important to client computer usage.
Optane Raid 0 Array終於成功了
https://pctantao.com/297264039.html
Optane Raid0 作磁盤緩盤
https://www.chiphell.com/thread-2299063-1-1.html
Intel Quad RAID-0 Optane Memory 32GB Bootable Without VROC Key!
https://pcper.com/news/intel-quad-ra...hout-vroc-key/
Intel VROC Tested! - X299 VROC vs. Z270 RST, Quad Optane vs. Quad 960 PRO
https://pcper.com/2017/10/intel-vroc...-quad-960-pro/
問題是經過精簡的 kernel 也不一定能夠支援 RAID 了,而且 Z690 的主機板只可以一拖二:
https://www.asus.com/support/FAQ/1037507/
最後就是機殼方面的選擇,空間足夠的話就找那些重量十足的全鋁外殼:
https://item.jd.com/10034104850693.html
https://item.taobao.com/item.htm?id=582302895064
日後 12900K/12900KF 跌價的時候就可以升級到這種無風扇的機殼,聲稱單 CPU 持續功耗最大 150W:
https://item.taobao.com/item.htm?id=626163494407
http://www.zerorpm.cn