On Intel Raptor Lake, any truth to the rumors that disabling all e-cores hurts single threaded performance of the p-cores??

[Cyber-Mav]

My daddy as a twin turbo v12 zimmer.

Dihatsu 1.3 here

 

[Cyber-Mav]

So is it just a die shrink that will bring intels power consumption down or some architectural changes needed too?

 

[MartinPrince]

You seem to be taking your time with the figures. Guess the CPU is slower than i thought

Had to take the son out to see and photograph the blue supermoon and put him to bed.

I'll do more than @Wing-Man asked and go through some other software and games to give him a better picture, so might be tomorrow before I get through them all and put them up. I'll avoid fabricating information, unlike some...

 

[Robert896r1]

Had to take the son out to see and photograph the blue supermoon and put him to bed.

I'll do more than @Wing-Man asked and go through some other software and games to give him a better picture, so might be tomorrow before I get through them all and put them up. I'll avoid fabricating information, unlike some...

Be sure to tune the mesh on your rpl chip. Really pays off.

 

[jigger]

Be sure to tune the mesh on your rpl chip. Really pays off.

But what about the ring bus half?

 

[MartinPrince]

The first post i read when disabling the E cores he overclocked the ring. This is not a fair comparison. I’d like to see just on and off everything to stock.

In bench marks the scores give a false reading as you get a full 24t score. But under full heavy load they should be off.. so should be 16t

Sorry for the delay, what with fitting a new kitchen and keeping up with the way more important pre-release hype and reviews of Starfield it was hard to find the time! I also just read the "stock" bit in your request but I've tested it on my system as is, though the E-cores have not been overclocked just the P-Cores (with HT disabled for both)
The OS:
Edition Windows 11 Pro
Version 22H2
Installed on ‎07/‎11/‎2022
OS build 22621.675
Experience Windows Feature Experience Pack 1000.22634.1000.0


With E-Cores

No E-cores.

With E-cores

No E-cores

With E-cores

No E-cores

With E-cores

No E-cores

With E-cores

No E-cores

With E-cores

No E-cores

With E-Cores

No E-cores

 

[MartinPrince]

@Wolverine2349

As you can see, for me anyway, when it comes to single threaded performance E-cores do not seem to harm performance as all the results are well within run to run variance.

 

[Wing-Man]

@Wolverine2349

As you can see, for me anyway, when it comes to single threaded performance E-cores do not seem to harm performance as all the results are well within run to run variance.

So literally no difference, on a single core and game basis.
Multi threaded bench will always show a difference because there is less threads.

Thank you for talking the time to clear that up for me