[energie80]

You can make a decent os by yourself

 

[J7SC]

...full-fat Windows benching...

 

[eighty20]

Anyone who still play PUBG, do you guys have issue where CPU (Tctl/Tdie) suddenly spike to 85C. I only launched the game, enter the lobby and it's reported 85C.

 

[HappySheep]

bios 1415 vs 1709 asus gene

1440p 7800x3d RTX 4090

1709 4% worse in watchdogs legion benchmark 1080p very high preset, same performance in MW3 benchmark, 70 points better in CB23 and .4ns better in Aida64
TimeSpy Extreme Custom Test 1 is 1% or more worse

I'm confused why in watchdog legion benchmark is much worse

 

[Word_Up71]

Testing weeks now with ECLK.

Very difficult to get the right values in the co without crashing windows.

Push one core up, one core goes down and so on.
CCD1 is very sensitive for everything above 5750. CPU lottery.

But in my opinion, it can change CPU-Performance (CCD0) in games really well.

 

[Kelutrel]

The file in attachment contains version 1.82 of the RopBench tool.
The SHA-256 of the original archive is: 1653ab67100527b78609938b4d29880f398d3a8e0da5a8451cd4ee34551f36d6

In this version:

• Added support for Zen 5 (thanks to @RedF for testing)
• Added custom icon (the artist is @RedF)
• Updated UI library to latest
• Fixed detection of some older Zen 4 models

 

[J7SC]

Testing weeks now with ECLK.

Very difficult to get the right values in the co without crashing windows.

Push one core up, one core goes down and so on.
CCD1 is very sensitive for everything above 5750. CPU lottery.

But in my opinion, it can change CPU-Performance (CCD0) in games really well.

View attachment 2644604

ECLK is really enjoyable, keeping in mind that once beyond a certain MHz off base 100, it will start to change CCD0 vsCCD1 auto-preferences in games and other apps.

I found my ECLK happy place by only using the 90L5 PBO enhancement on the Aorus mobo (which Skatterbencher suggests is an all-core CO), but no per-core CO at all, unlike on some of the other bios profiles. An important step is to test for max effective per-core speed - again, without any per-core CO, in stock BCLK.... move (negative) boost offset up / down until it can pass all tests in repertoire (single- and multi-core stress tests), including AIDA such as SHA3 and FPU Julia. So there is no per-core on my fav ECLK setting which is 105.4 and one of my 'daily profiles' re-posted below; it beats a similar ECLK profile with tested per-core CO in basically everything.

To some extent, it can depend on the 'spread' of cores in terms of max MHz - if dramatic MHz differences exist within a given CCD (0,1) then you can use per-core CO to 'flatten' the max effective MHz differences of cores within a CCD. But overall, IMO, it is better to run higher-clock ECLK with no per-core CO and use (negative) boost offset as the main adjustment tool...

 

[notwalkin]

Anyone who still play PUBG, do you guys have issue where CPU (Tctl/Tdie) suddenly spike to 85C. I only launched the game, enter the lobby and it's reported 85C.

I don't play pubg but isn't this the standard for x3d cpus?

At least my 7800x3d spikes to 80 or 85 at times when launching games/programs.
The game itself will run much cooler though, on average.

 

[eighty20]

I don't play pubg but isn't this the standard for x3d cpus?

At least my 7800x3d spikes to 80 or 85 at times when launching games/programs.
The game itself will run much cooler though, on average.

I run various games and programs but only pubg spike like that. I wonder what the cause of it so maybe using it to test each core separately since it can increase that much temperature.

I also forgot to say that I haven't seen any of stress test (both MT & ST) reach above 70 except cb23( barely). I'm not using chiller but the temperature at the spot where I placed my rads is ~14c, it's beeb like that for awhile so I'm supprised how pubg can reach that high temp while stress testing using corecycler can't.

 

[sendap]

I will play PubG tonight for a bit and let you know.

 

[sendap]

I will play PubG tonight for a bit and let you know.

it does spike up to 86 degrees and runs overall warmer than my 5800X3D. I mainly play Tarkov, no spikes at all with that game.

 

[Melan]

I was messing around with soc voltages and noticed if I play music while running linpack (or any other stress test), on unstable fclk music will start skipping. Wish I knew that before testing for fclk instability for hours lol.

 

[Imprezzion]

Anyone who still play PUBG, do you guys have issue where CPU (Tctl/Tdie) suddenly spike to 85C. I only launched the game, enter the lobby and it's reported 85C.

Mine does the same with the division (1). Spikes Tctl to 82-83 but cores never go over 67c. During gaming (solo legendary runs for the fun of completing all of them solo, only warrengate and general assembly left, got all the others so far.) it never really goes much over 55c. It does run mid 60's when I run black desert minimized afk horse training as well but yeah. Makes sense running 2 games lol.

 

[Chodiddy]

Hi, I would like some advice on why my CCD1 on 7950X3D is no longer boosting to 5.75. I've had this problem before but it would go back to 5.75 when I reset the bios. I'm on a MSI X670E Carbon.

Right now, I'm using bios 7D70v1C, the last version before the beta. I tried the beta but that's when I noticed that CCD1 hasn't been jumping to 5.75. I've been monitoring on HWINFO. When I noticed that the top core speed was only going to around 5.4, I went back to the previous bios, but the pattern continued.

I'm using FCLK 2100 and all core CO of negative 20. This has worked for more than a month for me without any crashes. All cores speed is still about the same as always, around 5.0 on CCD0 and 5.2-5.3 on CCD1. Performance is about the same but it's just annoying to see that my top core speed is no longer hitting what it used to.

 

[eighty20]

it does spike up to 86 degrees and runs overall warmer than my 5800X3D. I mainly play Tarkov, no spikes at all with that game.

Mine does the same with the division (1). Spikes Tctl to 82-83 but cores never go over 67c. During gaming (solo legendary runs for the fun of completing all of them solo, only warrengate and general assembly left, got all the others so far.) it never really goes much over 55c. It does run mid 60's when I run black desert minimized afk horse training as well but yeah. Makes sense running 2 games lol.

Yeah it's strange that each core temp is a lot lower than tctl. I ddu, uninstalled the game, cleared cache and then used Lasso to assign core. Any core I assigned to pubg will have temperature spike 85C real fast when launch then chill around 50-ish, and it's random for only two cores. I found a number of ppl report having the same issue. Some suggest that dx11e/dx12 of the game itself is not optimized so switch back to dx11 may help, but that's not working on AM5&W11 combination.
One of the the reason why I still play pubg is bc I saw @der8auer use it for some testing on Intel plaform so I think it's good for testing cpu. However, when the temp spike like that in a short amount of time, I wonder if it would cause any damage to vrm.

Anyway, I'm glad you guys actually tested. Because most response I found on other sites are just bland critism of how poor the contact of heatsink, unstable OC and some even suggest set tjmax at 75C instead of actually playing the game and test it. I also found that modded driver for AMD gpu like nimez ( thank to those entitled on that discord group) causing unstability for FCLK. No idea how but when I installed the offcial one on AMD site, I can manage to run FFT and Karhu back to back without slowing down or crash ( not thoroughly tested yet) at 2200FCLK.

@BlueShark you mentioned the game on PS3 emulator that act like a virus, do you remember the name?

 

[sendap]

Pubg has always been a little strange. Dont worry about it. And those spikes are not harmful at all. It still is within limits.

 

[eighty20]

I may have to pass PUBG until there's a improvement, i usually solo anyway 😆 I plan to play w BCLK so if it's 85C then wonder how high it could spike at 1.15vcore.

 

[HappySheep]

@LtMatt dis you happen to lose fps with recent bios's compared to 1415 specially on WDL SOTTR and MW3 benchmark?

tried 1709 and 1801 few times but they were super weird, one time even my bios kept freezing for few secs when sticking USB in

 

[LtMatt]

@LtMatt dis you happen to lose fps with recent bios's compared to 1415 specially on WDL SOTTR and MW3 benchmark?

tried 1709 and 1801 few times but they were super weird, one time even my bios kept freezing for few secs when sticking USB in

I don't think so, but it's been so long since I used a BIOS like 1415 or the betas that were around at the time I've not really kept track.

 

[energie80]

@LtMatt dis you happen to lose fps with recent bios's compared to 1415 specially on WDL SOTTR and MW3 benchmark?

tried 1709 and 1801 few times but they were super weird, one time even my bios kept freezing for few secs when sticking USB in

Not really