lopri
Elite Member
- Jul 27, 2002
- 13,173
- 557
- 126
It takes about a minute to cold boot for me with 4x16GB memory. I might time it with a stop watch some time. 1 minute is long for a boot, for sure. Thank goodness I am done tuning.
How much effort, to poweroff, remove 2 stick and try it again ! I would love to see that with a real timed set.From powered-off state to the Windows desktop, I just timed 1 minute and 10 seconds. This is with 4 DIMMs, though, and it might be shorter with 2 sticks.
I am unaware with a USB problem but then again I do not use a headset, so can't comment on it.
so, 15 seconds faster with 2 dimms.... They must be integral to the boot.44 seconds to Windows logo and 54 seconds to the desktop from I switched the system on. With 2 sticks.
You owe me a dinner.
Nah. That's a problem I've encountered with newer boards. You are not alone.For whatever reason this board's DIMM sockets are very stiff. It makes me nervous every time I take the sticks out.
It does not bother me, one minute to boot every few months ? Just put it to sleep when you are not using it.I wish I could fast forward 12 months just to see what can be fixed about boot times and memory with AGESA revisions. Right now it feels like we're just making guesses about limitations of the platform.
To be fair boot time of OSes is still lightning fast (mainly thanks to NVMe etc.). What regressed is the time POST requires for initializing the hardware, especially RAM training AMD introduced on AM4 and appears to have aggravated on AM5.I recall when Windows 8 used to advertise fast boot times as a key feature. It feels like we have regressed a lot from that.
If you use EXPO settings, you lose the Sleep function (at least I do). Is that the same for everyone else?It does not bother me, one minute to boot every few months ? Just put it to sleep when you are not using it.
Fixed for me as of Agesa 1.0.0.4 on my Asrock board.If you use EXPO settings, you lose the Sleep function (at least I do). Is that the same for everyone else?
I'm looking at the POST time + time spent on the Windows loading screen. Basically until the time when the mouse cursor and desktop show up (but before startup apps load). Both of these things went up a lot over the years.To be fair boot time of OSes is still lightning fast (mainly thanks to NVMe etc.). What regressed is the time POST requires for initializing the hardware, especially RAM training AMD introduced on AM4 and appears to have aggravated on AM5.
IMO something's clearly wrong there and it's odd that not more is being talked about that.
Something must not be getting alone with memory timings. Or maybe the board is changing something that it is not supposed to. Have you tried entering EXPO values manually? My understanding is that they are not that special anyway. It's mostly primary timings or primary+secondary timings that memory vendors program, and the motherboard does the rest.If you use EXPO settings, you lose the Sleep function (at least I do). Is that the same for everyone else?
it changes all sorts of memory voltages, all sorts of timings, memory speed and a couple other things. Its WAY easier to use it on EXPO memory than trying to set it yourself. I have an EXP 6600 kit that I paid $400 for (way before the chip was available) and could not get it to run better than 5800 or 5600. The EXPO kit of 6000 CL30, I set it (one click) on 3 different motherboards, and all 3 have worked perfectly ever since.Something must not be getting alone with memory timings. Or maybe the board is changing something that it is not supposed to. Have you tried entering EXPO values manually? My understanding is that they are not that special anyway. It's mostly primary timings or primary+secondary timings that memory vendors program, and the motherboard does the rest.
I have not seen that setting. THAT is the issue, all boards need that in the bios.I'm not certain why boot times are an issue now. My Asus board had a setting to stop training after the initial one. Expo settings and sleep work no problems either.
I see. I'm not a Windows user so I didn't know Microsoft was slacking in that aspect as well.I'm looking at the POST time + time spent on the Windows loading screen. Basically until the time when the mouse cursor and desktop show up (but before startup apps load). Both of these things went up a lot over the years.