These guys have no idea what they are talking about.

[humbug]

Yes my 3D gets run very hard with SC, seems to use whatever you throw at it. May improve with more optimisation? Have you noticed any difference with 3.17.2?

No difference with 3.17.2, performance is good tho....

I have my cheapo 3200MT/s ram solid at 3666MT/s now, i'm still experimenting with it. I had to manually set the IF clock, or North Bridge Frequency to 1833Mhz as anything over 1800Mhz (3600MT/s) it defaults to 1/2 speed, takes two clicks in the BIOS to override that

I'll shoot for 3733MT/s next.

 

[Troezar]

No difference with 3.17.2, performance is good tho....

I have my Cheapo 3200Mhz ram solid at 3666Mhz now, i'm still experimenting with it. I had to manually set the IF clock, or North Bridge Frequency to 1833Mhz as anything over 1800Mhz (3600MT/s) it defaults to 1/2 speed, takes two clicks to override that

I'll shoot for 3733MT/s next.

Good result. I just have XMP running 3600MHZ tight timings for XMP but sure I could get more out of them, not sure about 4 sticks though on an X570 Aorus Master. 3800MHz at CL14?

 

[humbug]

No difference with 3.17.2, performance is good tho....

I have my cheapo 3200MT/s ram solid at 3666MT/s now, i'm still experimenting with it. I had to manually set the IF clock, or North Bridge Frequency to 1833Mhz as anything over 1800Mhz (3600MT/s) it defaults to 1/2 speed, takes two clicks in the BIOS to override that

I'll shoot for 3733MT/s next.

Woop woop.... get a load of this £90 32GB kit

 

[Vulture]

I'm curious what Intel did to improve efficiency of RPL? ADL is terrible at it, and RPL is just little refresh of ADL so it isn't like Zen 2 to Zen 3 on the same node situation, they added more cache, 8 additional cores and increased clocks which increase power consumption by a lot, and if architecture is mostly the same then efficiency will be a lot worse.

 

[Troezar]

Woop woop.... get a load of this £90 32GB kit

You've got better latency than me on Memory 69.3ns. Read Write and Copy are a bit slower than my L2 and L3 but your L1 is faster. Not sure how much 4 sticks has an effect? I paid a lot more than £90 so you've done well there

 

[humbug]

2X 16GB sticks, is it yours that are 4 sticks? if so it can have a small effect on the bandwidth at slightly higher latency.

I would like to get it under 60ns but i think i'm kidding myself with that...

 

[Troezar]

2X 16GB sticks, is it yours that are 4 sticks? if so it can have a small effect on the bandwidth at slightly higher latency.

I would like to get it under 60ns but i think i'm kidding myself with that...

Yes 4x4GB 8Pack Ripped Team Group.

 

[Cyber-Mav]

What about efficiency measured over time?
If cpu A pulls 200w and renders in 1 minute and cpu B pulls 50w and renders same scene in 5 minutes which would be more efficient?
People see 200w and think straight away that its inefficient.

 

[Cyber-Mav]

Tested it again at 65 watts, 142 average and 105 minimum. That's 2.23 fps per watt, exactly the same as the 3d gets


Tested one last time with e cores on at 65w, surprisingly it does way better. Peak wattage was 61 and fps was 154 average and 116 minimum. I don't know why but with e cores on it doesn't hit the power limit

test again with just e cores getting 5 or 10w. would be interesting

 

[Bencher]

test again with just e cores getting 5 or 10w. would be interesting

Cant have only e cores on, tested with 1P + 8E cores at 10w. 67 fps average 47 minimum

 

[Nexus18]

 

[Disco_P]

Listened to and agree with what he's saying. Lotta users with niche requirements completely oblivious to how niche they are.

 

[Cyber-Mav]

Cant have only e cores on, tested with 1P + 8E cores at 10w. 67 fps average 47 minimum

Test cinebench with just ecores

 

[Cyber-Mav]

Holy moly:

Intel Core i9-12900K E-Cores Only Performance Review

With Alder Lake, Intel is betting big on hybrid CPU core configurations. The Core i9-12900K has eight P(erformance) cores and eight E(fficient) cores. We were curious and tested the processor running the E-Cores only to see how well they perform against architectures like Zen 2, Zen 3, Skylake...

www.techpowerup.com

Those e cores alone are almost a match for a 5600x !!!!!!!

 

[Bencher]

Holy moly:

Intel Core i9-12900K E-Cores Only Performance Review

With Alder Lake, Intel is betting big on hybrid CPU core configurations. The Core i9-12900K has eight P(erformance) cores and eight E(fficient) cores. We were curious and tested the processor running the E-Cores only to see how well they perform against architectures like Zen 2, Zen 3, Skylake...

www.techpowerup.com

Those e cores alone are almost a match for a 5600x !!!!!!!

No no, they are "e waste cores".. Amd users trolling around various forums, at least you are being honest. Yeah, e cores are insane, especially for the die size. Nothing gets close in terms of performance per die space

Since everyone else doesnt dare test it, do you mind running the in game farcry 6 benchmark at 25w? Set the resolution to lowest possible, 768p, and ultra settings. I wanna see that zen 3 "efficiency" in action

 

[eddyr]

Holy moly:

Intel Core i9-12900K E-Cores Only Performance Review

With Alder Lake, Intel is betting big on hybrid CPU core configurations. The Core i9-12900K has eight P(erformance) cores and eight E(fficient) cores. We were curious and tested the processor running the E-Cores only to see how well they perform against architectures like Zen 2, Zen 3, Skylake...

www.techpowerup.com

Those e cores alone are almost a match for a 5600x !!!!!!!

Are we looking at the same review?
The 8 e cores are no where near the 5600X. The e cores even have a 200mhz clock advantage.

The E cores are just about matching the Zen 2 3300X 4c/8t chip.

 

[Bencher]

Are we looking at the same review?
The 8 e cores are no where near the 5600X. The e cores even have a 200mhz clock advantage.

The E cores are just about matching the Zen 2 3300X 4c/8t chip.

He is talking about MT performance, which is the reason they exist.

https://tpucdn.com/review/intel-core-i9-12900k-e-cores-only-performance/images/cinebench-multi.png

 

[Cyber-Mav]

No no, they are "e waste cores".. Amd users trolling around various forums, at least you are being honest. Yeah, e cores are insane, especially for the die size. Nothing gets close in terms of performance per die space

Since everyone else doesnt dare test it, do you mind running the in game farcry 6 benchmark at 25w? Set the resolution to lowest possible, 768p, and ultra settings. I wanna see that zen 3 "efficiency" in action

I dont have farcry 6 unfortunately. Pc mostly used for video encoding and some games like forza 4

 

[Cyber-Mav]

Are we looking at the same review?
The 8 e cores are no where near the 5600X. The e cores even have a 200mhz clock advantage.

The E cores are just about matching the Zen 2 3300X 4c/8t chip.

I was looking at cinebench test

 

[eddyr]

I was looking at cinebench test

Fair enough.

He is talking about MT performance, which is the reason they exist.

https://tpucdn.com/review/intel-core-i9-12900k-e-cores-only-performance/images/cinebench-multi.png

There is more to assessing a core's multithread throughput than just cinebench.

On the criteria of MT performance, they are clearly not equivalent to Zen 3, as is evidenced by all the multithreaded tests in that same TPU review, including several other rendering benchmarks, not to mention other highly parallel workloads also shown.
You are very focused on a benchmark where the e-cores are particularly good at maximising instruction throughput for that workload to prove a conclusive point, but which is not indicative of real world multi thread performance or efficiency.

Speaking of cinebench, what would be far more interesting than a loaded comparison using a preferred favourable benchmark, is why exactly does cinebench favour the e-core.