Info Statuscore core stresser

[igor_kavinski]

Statuscore Home

Cool utility.



Notice how stressing a selected core causes the next core to rise in temperature too, despite nothing happening on it.

Recommend not to stress all cores unless you have good cooling.

Also, the driver in the folder needs to be installed for it to work.

 

[dullard]

Notice how stressing a selected core causes the next core to rise in temperature too, despite nothing happening on it.

That is the purpose of a heat spreader. Spread the heat around so it can be cooled more readily.

 

[Det0x]

Cant read Zen5 temps and dont understand ECLK overclock it seems
But pretty much zero heating up of the neighbouring cores with DD + LM

 

[igor_kavinski]

Yeah, the author last added support for Zen 4. Maybe something changed with Zen 5.

 

[igor_kavinski]

Notice how stressing a selected core causes the next core to rise in temperature too, despite nothing happening on it.

That was dumb of me. Of course the next core will show the same temp because it's the virtual HT core!

 

[igor_kavinski]

 

[igor_kavinski]

@511 Could you record a similar video for your 185H? Want to see the IPC of the LP E-cores.

 

[511]

@511 Could you record a similar video for your 185H? Want to see the IPC of the LP E-cores.

Yeah sure only LP-E ?

 

[igor_kavinski]

Yeah sure only LP-E ?

No, all of them please. If you can run rudibench alongside, that would be great too.

Power profile should be High or Ultimate and plugged in please.

 

[511]

No, all of them please. If you can run rudibench alongside, that would be great too.

Power profile should be High or Ultimate and plugged in please.

Ok i have limited my U9 185H TO 45w using Throttlestop though.

 

[igor_kavinski]

Ok i have limited my U9 185H TO 45w using Throttlestop though.

I guess that's ok.

 

[511]

@igor_kavinski Here are the results with power figures

LP-E core roughly 8.5-9.5W @2.5 Ghz


E core 8.5-9.5W @ 3.8 Ghz


P core 14.5-15.5W @ 4.8Ghz


P core 19-20W @ 5.1 Ghz

 

[511]

@igor_kavinski if you look carefully at frequency Intel 4 is able to clock the frequency nearly 1.5X higher vs N6 Crestmont which is ridiculous lol

 

[igor_kavinski]

Kinda confused by the LP E-core frequency. It is boosting pretty high (base is just 1 GHz). Didn't expect that.

 

[511]

Kinda confused by the LP E-core frequency. It is boosting pretty high (base is just 1 GHz). Didn't expect that.

 

[igor_kavinski]

So the LP E-core is just the exact same E-core with same architecture just clocked lower? Because the IPC and frequency tracks perfectly between them. 50% higher frequency for 50% higher MIPS.

 

[511]

So the LP E-core is just the exact same E-core with same architecture just clocked lower? Because the IPC and frequency tracks perfectly between them. 50% higher frequency for 50% higher MIPS.

Yes exactly same core also I think this workload is very small so it fits inside L1 or spill to L2 at max otherwise we would have seen worse results in IPC.

 

[MS_AT]

So the LP E-core is just the exact same E-core with same architecture just clocked lower? Because the IPC and frequency tracks perfectly between them. 50% higher frequency for 50% higher MIPS.

Comparing Crestmonts: No L3 Hurts

Meteor Lake’s standard E-Cores share a 24 MB L3 cache with the P-Cores, while the LPE-Cores have to make do with just their private 2 MB L2.

chipsandcheese.com