Then just cap the game to 117fps, set monitor refresh rate to 120hz and call it a day?
141fps @ 144hz is really not a whole lot different to 117fps @ 120hz, despite what marketing and armchair believers will tell you.

Sure you're theoretically "wasting" your 3080Ti, but if it makes you feel any better - I really only play tekken 7, a 6 year old game that could run on a toaster and it's locked to 60fps at 1440p. There, I'm wasting my 3080 more than you're wasting your 3080Ti.

Otherwise just run at a higher resolution using DSR if you feel you need to put your GPU to use.

This post has been edited by SSJBen: Aug 25 2021, 03:33 PM

Not for me.

But this issue was somewhat common with the 3800x and also the 5800x. This is down to faulty chiplet, probably the bottom of the barrel silicon that barely pass for 8 cores but still gets binned for production.

Oof Win 11 perf with Zen 2/3 is very noticeably bad. Most issues stem from workload not being properly spread to multiple cores, there are instances where even if a single core is already running at above 90% process, the OS will still send more work to that core despite others being idle. MS said they were going to make W11 a true multi-core focused OS, but so far it's doing a poorer job than Win10 and even Win8 lol.

Lucky I'm not migrating to Win 11 full time.

No. The feature I like most about W11 on launch is auto HDR and the proper usage of WCG. That works as expected and it's a good feature.

I think W11 has the potential of being a good OS though.


I'm talking about performance regression, not driver issues. So far everything hardware wise in terms of driver has worked on W11 like it did on W10.

Experimented a bit more over the weekend, the OS is not being allocating processes to free cores even on something as simple as Excel crunching and code compiling on Python. This is happening on both my 3900x and 5900x machines, very repeatable issue also.

Found another big bug too - HAGS (hardware accelerated GPU scheduling) will peg the GPU at 100% load if using gamestream with Nvidia devices like the Shield. If using Plex/Kodi to watch movies, same issue could happen to 4k movies with HDR tonemapping enabled. Very stupid bug.

Conclusion = Win11 can eat a dick for now.

Actually don't have high hopes for their 3D v-cache either. Zen 3 is complicated enough as is, just tacking on massive amounts of L3 isn't going to alleviate the problems it already has. This is just a stop gap for AMD so they can have something to compete with ADL (eventhough it's a futile attempt) so that they have a buffer before launching Zen 4.

It's made for people who are on Zen 2 who has a good X570 board, good b-die DDR4s and not wanting to wait until Zen 4 releases in Q4 2022 because FOMO.

Honestly, it's a pointless release just like Coffee Lake Refresh was.

It's practically still Skylake lol. There was no need for Coffee Lake and Coffee Lake refresh to even exist, they could have just gone straight to Comet Lake but Intel was LCLY at the time.

Yes Zen 2 XT was also a pointless release, even more so when it went AWOL just 2 months after launch.

Previous CEO was a clown.

Current CEO is actually an engineer first, business man second. He was one of the key pillars that led Intel to dominance with Conroe after the joke that was Pentium D.

Intel hobbled down the stairs after Haswell actually. Remember Broadwell? Oh wait, it barely existed. They were lucky AMD was a shit show at the time, otherwise Devil's Canyon would have been Intel's downfall instead of having leeway until Skylake.

Yup, glad I kept my 3900x and 5900x. Mining with them to fund for Raptor Lake and Lovelace.

5900x and 5950x are still being sold at scalper prices anyway, no difference whether people buy it up or not for mining.

RTM mining is very nice on Zen 2 and 3 I must say though. ~70c no AC room with 3.5 - 4k h/s average using AVX workload.

CPUMiner is trash. Use XMRig instead.

My 5900x and 3900x only hitting 75c max mining RTM, both only equipped with a NH-D15s. Don't even need 360mm AIO.

For a stock cooler those temps are normal.

Yes you can always spend for a bigger air cooler, go for AIO or go custom loop. The lower the temps for Zen 2/3, the faster it performs.

Yes, absolutely change the thermal paste if you're going to change the cooler. It's conventional wisdom not to re-use thermal paste.

EDIT-
I suggest to use thermal paste that are not susceptible to "pump-out" effect, so higher viscosity paste like Gelid GC-Extreme instead of the typical low viscosity Thermal Grizzly Kryonaut.

This post has been edited by SSJBen: Mar 1 2022, 12:23 AM

Generally it is wise to change thermal paste after 1 or 2 years of usage. Thermal paste may have been "pumped out" towards the edges of the CPU IHS, leaving not very much in the middle which is where the chiplet and IOD is.

USB disconnectivity pun not yet fix despite lying about it for nearly a year, now want to launch 7 new CPUs with new BIOS.

Lmao AMD.

USB disconnects and AMD is synonymous with one another. I can guarantee disconnects also happen on the USB A ports, just that you haven't run into it yet.

Nothing much you can do other than updating your BIOS and hoping it fixes itself. AGESA 1.2.0.1 Patch A was the alleged USB fix, which AMD lied about IMO.

I had no USB drop outs with AGESA 1.0.0.1 (which was when the USB issues were reported, eventhough it existed long before this) until I tested an iFi Zen DAC v2. Had to update to AGESA 1.2.0.3 Patch C (BIOS 3801) in order to eliminate the issue (hopefully, so far it's been good for a week now).

Motherboard I'm running is Crosshair Hero VIII, which runs on the same BIOS as the Dark Hero. So try out 3801, it's working good so far for me.

I have fTPM enabled because I'm also using W11, no stutters here for me in any apps or games.

AGESA 1.2.0.7 is supposedly to be the last major "clean up" BIOS release. I'm willing to bet all my 10 BTC it won't be.

This post has been edited by SSJBen: Mar 24 2022, 08:03 PM

I do also agree that USB disconnection can be related with cable issue.

All the DACs I'm currently testing at the moment (Topping DX3, FiiO K9 Pro, JDS Labs Element 3 and iFi Zen Dac v2) all stopped having USB disconnections after I switched my old ugreen USB-B 3.0 cable to a new Vention USB-B 3.0 cable.

So while AGESA does have something to do with USB disconnecting, the USB cable is also another culprit to look into.

I don't know if it's exactly to do with ugreen cables being specifically the issue or not.
But personally from my own anecdotal experience, ugreen cables has always failed me in one way or another.

Had several USB A to micro USB 3m cables, all of them failed within a year. Yes I know micro USB is a bitch to begin with due to the shitty head design, so I'll dismiss that as a connector flaw.

Had a ugreen active HDMI fiber optic cable, 15m in length supposedly rated for 2.0. Had audio drop outs within 3 months, video drop outs after 8 months and failed after that with a short out. Thank goodness it didn't take out the HDMI port on my 3080 or my receiver, would be pissed AF.

Also still have several USB C and B ugreen cables of varying lengths between 1m and 2m, they've all started to exhibit disconnectivity issues.

Only cables that so far hasn't have any issues are the CAT6, CAT7 and passive HDMI 2.0 cables from ugreen.

This post has been edited by SSJBen: Mar 30 2022, 04:55 PM