Shhh, we're not allowed to talk about that here

i mean if you oc the 7700k to around 4.7-4.8 the physics scores will be the same

You think .2-.3GHz more will close a physics difference of 20%? lol OK.

This post was edited by Incendiary on Dec 7 2017 07:10am

it's 18% and i was still assuming 4.0ghz
4.5? oc to 5.3 for 166.6% more single core perf and same physics with 66.667% less cores

This post was edited by DCSS on Dec 7 2017 07:13am

Except it won't be the same physics because the physics score in FireStrike scales higher with more cores and threads than core clock if the IPC is the same. Very few 7700k go that high. Are you just spamming and trolling or..?

"As of 12/01/17, the top 8% of tested 7700Ks were able to hit 5.2GHz or greater." -SiliconLottery

and you can close the core gap with clock speed if ipc is the same (which it isn't in this case but that's another story), 8 cores at 1ghz aren't going to get higher physics than 2 cores at 6ghz. Forget about the Ryzen for a moment, your physics score should be roughly 18% higher with 18% higher clock speed, barring other factors like gpu/ram speed which are present for the ryzen as well, unless, for some reason, on 3dmark (my favorite game btw) a 10% OC on 16 cores is going to yield a huge relative increase in that physics score when compared to the relative increase a 4 core chip would get from a 10% OC, in which case this synthetic is quite stupid and has a nonsensical way of calculating score.

OS and driver overhead should be slightly higher with fewer cores which could be a factor but I don't think it would result in a very different score if the overall throughput of the chips is nearly evened out like it would be between an r5 1600 and sufficiently overclocked 7700k.

This post was edited by DCSS on Dec 7 2017 08:16am

https://www.3dmark.com/compare/fs/11951867/fs/14336701

Here's a 7700k at 5GHz vs my 1600 at 3.8GHz. Still behind in physics by 9%.

Here's another 7700k, this time at 5.3GHz and still behind my 1600 by 1%. https://www.3dmark.com/compare/fs/12655276/fs/14336701

This post was edited by Incendiary on Dec 7 2017 08:57am

where's the evidence that it's even 5ghz? max turbo just means 1 maybe 2 cores, physics uses all



within margin for error, this is basically exactly what i expected. Wasn't exactly 5306 mhz that i calculated the scores evening out at, it's slightly higher but still under 5.4

This post was edited by DCSS on Dec 7 2017 09:00am

cores are big deal on 3dmark,

check out intel 6 core physics scores 7800x and back and see how they are, I'd look but need to focus on 2 files at work lol

So to make up the difference in 2 cores/4 threads you need a near 40% increase in clock speed. Very, very, very few 7700k will reach this 5.3GHz mark, and it requires a de-lid. Any 1600 will hit 3.8GHz on a $50 mobo with a box cooler. I'll take the Ryzen and its cores.



As they should be when measuring physics!

This post was edited by Incendiary on Dec 7 2017 09:02am