digitalbath
Mr. Oldschool-Bios
- Mitglied seit
- 04.05.2008
- Beiträge
- 4.121
- Ort
- Moers
- Prozessor
- AMD R7 5800X
- Mainboard
- ASRock B550 Steel Legend
- Kühler
- Scythe Fuma 2 + Noctua
- Speicher
- Crucial Ballistix DDR4 3600 32GB
- Grafikprozessor
- AMD XFX RX6600
- Display
- BenQ 24" XL2410t
- SSD
- 2x500GB; 1x1000GB
- HDD
- Ja, für Datengrab
- Soundkarte
- O2 + AKG K702 / K340
- Gehäuse
- Deepcool Matrexx 55 V3 RGB
- Netzteil
- BeQuiet! SystemPower11 Platinum 550W
- Keyboard
- Corsair K90
- Mouse
- Kingston Plastikmüll-X
- Betriebssystem
- Win11
- Webbrowser
- der rote Fuchs
Thanks, I see we're interested in researching the same things so I've registered to keep the info together.
pifast is a good idea! I will use it for the next tests too.Yes, latency alone might not always help. I use a combination of AIDA memory subsystem test (read/write/copy/latency) and Pifast. For example, CPC has small boost for AIDA but bigger boost for Pifast. Same for Tref.
My max FSB in 1:1 is ~273MHz. My other CPUs limit somewhere around 250-260MHz FSB.By the way, do you have info about max benchable memory frequency? I've been playing a several days ago and managed to complete Pifast on Winbond 1x256MB at 2-2-2-x at 289MHz (1:1).
With TCCD I couldn't get such clocks even at 3-4-4-8.
The max benchable memory, that I was able to get is 300Mhz in dual channel and with a 4:5 FSB : dimm ratio.
The max frequency, that I was able to validate ist 308MHz.
Pifast with 2-2-2-x / 289MHz is awesome! I wonder if you did this with a ASUS board? I mean, you probably needed a lot of voltage for this result and as far as I remember the P4x800 boards are tied to the 3,3V rail?