Discussion in 'AMD x86 CPUs and chipsets' started by Skramit, Nov 7, 2019.
OJR went from a 3950x to a 5800x iirc
Awesome wonder if that was an 'upgrade' lol
Asus ROG Strix x570 Gaming e: Temp 2 and Temp 5 readings (under motherboard) - real? | HWiNFO Forum
Temp 2 and 5 Will be removed in the next update.
Heh, got bored and had a quick try at the beta AGESA 184.108.40.206 on my asus x570-e.
Still WHEAs above 1900 fabric even with excessive soc voltage etc. Keep hoping for a magic bullet to save my sorry ass chip and get that nice round 2000. bahaha
edit: oh hey, they put it on the actual asus download page now.
So we have two official betas at once, 220.127.116.11 and 18.104.22.168.....
File is unchanged from what's been floating around already.
Doesn't seem to give any more room on fabric clocks, and may have potential usb issues. Give em a miss unless you're super bored.
So with PBO2/Curve Optimizer I know the range is +/- 0 to 30 and each step is 3 to 5 mv. Obviously negative offset is what is needed to undervolt.
So is a higher number less voltage or is a lower number less voltage. I would think a higher number would be less voltage since it is set to be a negative. However I've seen some arguments about which it is.
My understanding is that you'er just moving up or down on AMD's boost curve, a negative offset moving you up the curve into a more aggressive position. I've found a positive voltage offset (say +50mv) can help quite a bit with squeesing a bit more out of a negative curve, to help fuel the more aggressive boost behavior.
I got an additional 10-15fps in CoD Warzone at 3440x1440 on a 2080 Super. It was definitely an uplift in IPC.
EDIT: I have since switched to a RTX3080 and have got an additional 40-50FPS depending on map and conditions.
EDIT2: If you turn on RTX it steals about 20-25fps whereas on the 2080 Super it would almost halve the performance.
Old mate did some testing on a 5800X and this gives some insight about how the PBO2/Curve Optimizer settings work.
Found this link over on OCUK: https://docs.google.com/spreadsheet...QvVSKyfD4DQghgUHb46-VIIzl3BQvX-hYt0D4/pubhtml
what kind of temps are people getting with a 5900x?
Corsair H100i V2, 1500rpm ML120 Pro (non RGB) fans, idles around 45 - 55c and hits about 80c load. My room is almost always 25 - 30c though and airflow around the case is horrible.
Gonna be replacing the Corsair cooler with a 60mm thick rad but not sure when yet.
I didn't realise you got that much from just the CPU. I can't remember if we compared mine with the 1080Ti and 3900X vs yours with the 5800X & 2080 Super.
We'll have to compare 3900X vs 5800X though since we both have same GPU now.
5900x under EK 360mm AIO in a Phanteks Evolve X (rad top mount, 3 fans exhausting out top, 2 fans front draw in and one fan rear to exhaust)
No OC per se, just doing it's own boosting. Also have an RTX3080 in the case, adding it's own spicy air
Room ambient usually 20 - 24C, idles 36 - 39C, loads up to 75 when gaming. Goes up to 80C under load with 30C+ ambient.
Yeah, as soon as get my fkn mobo.
I'll just enjoy my >1 K/D while i wait then xx
Still better than mine!
Honestly, i've seen you play and i'm not surprised
That Hardware Unboxed video is very interesting how some boards hold a lot higher clocks with the 5950X than other boards do. VRM on paper doesn't seem to give any indication if the clocks will hold higher or not.
I'm wondering if that is what I've experienced with my Strix X570-I Gaming. Stock 5950X performance is much lower than a lot of benches out there.
I don't think they have done an ITX X570 round up like that yet. I wish they had before I bought a board. I might have just gone with the Aorus B550 board that held the highest clocks.
It wouldn't surprise me if we're seeing the AMD boards starting to play funnies with PPT/TDP etc limits like most Z series intel boards do by default.
One thing I did find is anandtech found the PPT drops 17w for full core loading vs 9 cores loaded, perhaps some boards don't do this and some do. But that's still well short of the ~40w difference in power consumption (allowing for some psu efficiency loss) between the Asus and Gigabyte motherboards.
Wow.. 13% difference in clocks at stock, with the same bios generation.
This is a great motherboard review actually by HUB. Significant differences found, thermals, clocks, layout etc. Comparison to ATX boards.
I am kinda happy that I got a Gigabyte b550 ITX. IMO gigabyte did a great job with this board, and is a significantly better board than the B450 i have. It weighs as tonne, Has a solid backplate, heat sink, heat pipe to SSD, integrated back plate, 2 x NVME slots, flash back. My only gripe (and with nearly all b550 boards) is another 2 USB ports or at least a PS/2 port would have been nice. I've tried quite a bit of OC on this board, and it is a champ. I am limited by ITX and 240mm AIO.
The Asus board does seem to clock really low.. Hopefully now this issue is getting media light, Asus can patch the Bios and fix it. I am sure the Asus board is capable of supporting such clocks, they may have just not yet looked hard at their ITX board bios.
I got to say I am loving AMD on ITX. IMO this is the future. 16 cores, 64GB, on a ITX board with a 6800 and 16GB ram. It feels like I've got power for decades.