News Core i9-13900KS 6 GHz CPU Beats Core i9-13900K By 10 Percent In New Benchmarks

DougMcC

Prominent
Sep 16, 2021
55
26
560
0
Would it have been so hard to completely redesign the K series with 10 p-cores instead of 8 for a marginal performance advantage in certain benchmarks? Stupid lazy Intel.
 
Dec 8, 2022
1
0
10
0
Would it have been so hard to completely redesign the K series with 10 p-cores instead of 8 for a marginal performance advantage in certain benchmarks? Stupid lazy Intel.
Sure bro you're smarter than those stupid lazy Intel employees with their so-called "degrees" in "computer engineering"
 

spongiemaster

Admirable
Dec 12, 2019
2,162
1,181
7,560
0
RIP, 7950X3D will be much faster in many games where cache is important. Just look at the 5800X3D and this time the 3D cache is even faster and more optimized.
12900KS was on average about 1% faster than the 5800X3D in gaming. And just like the previous gen, the 13900KS will crush the 7800X3D in pretty much everything else. Most people don't use their computer for gaming.
 
12900KS was on average about 1% faster than the 5800X3D in gaming. And just like the previous gen, the 13900KS will crush the 7800X3D in pretty much everything else. Most people don't use their computer for gaming.
I said in cache optimized games, 5800x3d is significantly faster than 12900KS there and now its a different situation as the 7950X is already too close to the performance of the 13900K, on top of that this time the 3D cache is faster and more optimized vs 5800X3D.

https://www.tomshardware.com/reviews/amd-ryzen-7-5800x3d-review
 
Last edited:

spongiemaster

Admirable
Dec 12, 2019
2,162
1,181
7,560
0
I said in cache optimized games, 5800x3d is significantly faster than 12900KS there and now its a different situation as the 7950X is already too close to the performance of the 13900K, on top of that this time the 3D cache is faster and more optimized vs 5800X3D.

https://www.tomshardware.com/reviews/amd-ryzen-7-5800x3d-review
What games does a 12900k struggle to run that magically become playable on a 5800X3D with its "significantly faster" performance?
 
I said in cache optimized games, 5800x3d is significantly faster than 12900KS there and now its a different situation as the 7950X is already too close to the performance of the 13900K, on top of that this time the 3D cache is faster and more optimized vs 5800X3D.

https://www.tomshardware.com/reviews/amd-ryzen-7-5800x3d-review
Yeah but it could be close because that's the limit of the games so no matter how much better the cache is it won't improve the games.
We can only guess and as far as guessing goes you can go both ways with the same amount of sureness.
 

zecoeco

Proper
Sep 24, 2022
46
54
110
0
I saw the Ryzen 7900x boost to 6.2GHz via PBO out of the box. So I don't think it's the world's first 6GHz as Intel claim.

Instantaneous 1ns pulses of 6GHz in one core isn't going to help much in performance, but its Intel's special chip. +200MHz for $100 more is not a good idea.
 
Last edited:
How does a 3-4% clock speed increase deliver a 10% performance gain?
They increased the PL1 to accommodate higher base clocks. It'll be 150W instead of 125W for the 13900KS. This means all-core workloads will have higher clocks.

der8auer has a really good video where he pushes a 13900K to 6.1Ghz, I think? It's basically a good preview of what the performance uplift should look like.

Regards.
 

magbarn

Commendable
Dec 9, 2020
84
47
1,570
1
Well if AMD releases a 7950x3d then they’ll fix my biggest beef with the 5800x3d, poor productivity performance. And it will beat the 13900KS in gaming if the uplift is just as good as it was for Zen 3.
 

JamesJones44

Respectable
Jan 22, 2021
287
207
2,060
0
RIP, 7950X3D will be much faster in many games where cache is important. Just look at the 5800X3D and this time the 3D cache is even faster and more optimized.
Why would you buy 7950X3D over a 7800X3D? Games are not going to come anywhere close to using 16 cores. Just a waste of compute power if all you do is play games.
 
Sep 21, 2022
38
6
35
0
Sure bro you're smarter than those stupid lazy Intel employees with their so-called "degrees" in "computer engineering"
it is all about the cost . not the possibilities. and yes a smart guy does not need a " degree" .. actually many never had one , like Thomas Edison who never even attended high school nor any university. and Steve Jobs who never completed his degree.
 

HideOut

Distinguished
Dec 24, 2005
481
36
18,820
1
Would it have been so hard to completely redesign the K series with 10 p-cores instead of 8 for a marginal performance advantage in certain benchmarks? Stupid lazy Intel.
yes it would have. Thats an entirely new chip. Not using parts from todays design. Probably a billion dollar design to sell a few thousand chips.
 

HideOut

Distinguished
Dec 24, 2005
481
36
18,820
1
It does not have the GPU power draw and heat creation, meaning it can hold higher clocks for longer periods of time and thus get more CPU work completed is the most likely reason.
No, he's right. The math does not add up. The 13900K does not have any difference at all except for no more than 200mhz of speed. By math its a MAX increase of what, 4%? They likely did not do same drivers on same rig thing. They cheated.
 

cyrusfox

Distinguished
Sep 24, 2009
285
129
18,990
8
The math does not add up. The 13900K does not have any difference at all except for no more than 200mhz of speed. By math its a MAX increase of what, 4%?
I think the answer lies in efficiency. the KS is a better bin chip, able to both sustain boost longer at lower energy use. Higher performance at same voltage, so pulls 10% better numbers even if max boost is only 4% higher, it is able to sustain a higher clock overal yielding a 10% improvement on this metric.

Too much cash for me for a KS, I am planning on getting at least a 13700k and maybe a 13900k(Replacing a 12900k and 12600k). I don't need the 6Ghz boost but will take all those threads for software encoding.
 
No, he's right. The math does not add up. The 13900K does not have any difference at all except for no more than 200mhz of speed. By math its a MAX increase of what, 4%? They likely did not do same drivers on same rig thing. They cheated.
That 200Mhz change is for single clock where the 13900k already boost extremely high and in single they do report a ~4% difference, the 10% difference is in MULTI core where a better binned core can reach higher clocks with the same amount of power, so you have 10% better bin between KS and normal K, or more power, or a combination of the two.
 

ASK THE COMMUNITY