• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Intel Core i9-13900KS

They are in the process of doing exactly that by working on Meteor Lake and Arrow Lake.
Time will tell if that's really what's going on and if we're just going to have another refresh of the same garbage that we have now. The current architecture is at the end of its lifespan, it's got no more gas in the tank.
 
I feel the architecture is OK, but this is a clear sign that Intel is lagging behind on their fab. While it is true that the Xnm metric is no longer relevant because it is not how the actual transistors measure, but Intel 10nm is clearly behind TSMC 5nm in both density and power efficiency. You can tell Intel is struggling ever since they got stuck on 14nm multiple years and they have never been in this state before.
 
They are in the process of doing exactly that by working on Meteor Lake and Arrow Lake.
Yes, just like rocket lake just improved the power sucking while removed 2 cores from 10900k, just like 12900ks completely new architecture on their new shiny node keep the power consumption and just like their new improvement just improved the power sucking again. Do you seriously believe in what you said?
 
I think the 13900KS needs some power justice here from what I've been reading in the comments.

My testing here is 8P cores only. If I need any more cores than that ill switch them on.

5.5Ghz - 66.7w.

Cyberpunk2077_2023_04_22_11_34_35_947.jpg


4.5Ghz - 44.5w

Cyberpunk2077_2023_04_22_11_41_47_429.jpg


4.5Ghz - 37.2w

HorizonZeroDawn_2023_04_22_11_55_29_770.jpg


As you can see with a bit of tuning these things can be very efficient.

I've been running mine on Air @ 4.5Ghz which is super cool, quiet and fast.

The reason for 4.5Ghz is because I'm not noticing any more FPS advantage over 5.5Ghz that's noticeable and temps, power and noise has been excellent.

If I need to do any more threading I've got 16 E-core's at my disposal plus HT if need be.
 
Last edited:
By the way, why when Zen4 on air hit 95C everyone second coment is for the crazy temps, but when 13900ks hit 115C on air with stock settings, there are zero complains?
I think the 13900KS needs some power justice here from what I've been reading in the comments.

My testing here is 8P cores only. If I need any more cores than that ill switch them on.

5.5Ghz - 66.7w.



4.5Ghz - 44.5w



4.5Ghz - 37.2w

As you can see with a bit of tuning these things can be very efficient.

I've been running mine on Air @ 4.5Ghz which is super cool, quiet and fast.

The reason for 4.5Ghz is because I'm not noticing any more FPS advantage over 5.5Ghz that's noticeable and temps, power and noise has been excellent.

If I need to do any more threading I've got 16 E-core's and my disposal plus HT if need be.
Its called GPU bottleneck
 
I think the 13900KS needs some power justice here from what I've been reading in the comments.

Ofcourse, every chip can be tuned to extremely efficient levels. But thats not the point. This is the "best" you can get for the most money, and the tradeoff is a awefull high power requirement. 355W at peak consumption, and when going OC'ed even beyond 400W. Mind you thats 3x the power requirement AMD is doing for the same performance. oh how times have changed.

I feel the architecture is OK, but this is a clear sign that Intel is lagging behind on their fab. While it is true that the Xnm metric is no longer relevant because it is not how the actual transistors measure, but Intel 10nm is clearly behind TSMC 5nm in both density and power efficiency. You can tell Intel is struggling ever since they got stuck on 14nm multiple years and they have never been in this state before.

TSMC 5nm cannot be compared to Intel's 14nm or even 10nm. It's completely different nodes.
 
Are you sure you got the CPU right? Because the 5950X is an AMD chip.
He had a 5950X, now a 13900KS and tested it with the same AIO.
 
What a pleasant surprise to have a new W1zzard review for a relaxing Friday evening read. :D

Page 24 - Clock Frequencies still has "The weird sawtooth pattern..." blurb from your 7800X3D review.

I think the 13900KS needs some power justice here from what I've been reading in the comments.

My testing here is 8P cores only. If I need any more cores than that ill switch them on.
[...]
I've been running mine on Air @ 4.5Ghz which is super cool, quiet and fast.

The reason for 4.5Ghz is because I'm not noticing any more FPS advantage over 5.5Ghz that's noticeable and temps, power and noise has been excellent.

If I need to do any more threading I've got 16 E-core's and my disposal plus HT if need be.

Question: in your current configuration, is there any difference between the 13900KS and a 13700K?
 
355W, utter madness. You shouldn't need to tweak a $700 CPU to get acceptable power consumption. Intel regressed back to Netburst levels?
 
Its called GPU bottleneck

I'm not sure this is the case. I'm running a 4080.

RealBench stress testing under full load CPU is 58w @ 4.5Ghz 8P cores only. Games will use less power.

Question: in your current configuration, is there any difference between the 13900KS and a 13700K?

Do you mean like P cores only? Then no.

A 13900KS should achieve lower power and thermals at the same frequency though as the silicon on the 13900KS will be better so less volts will be needed.
 
I'm not sure this is the case. I'm running a 4080.
Update your system specs page :)
You have disabled the E cores and the HT as far as I understood, also undervolted and underclocked it so its normal to consume fraction of the power, but that is to far and is not something that most users will make with their high end CPU, if not 13600k, 13700k will make same job. Do you know how much of the power consumption drop when you disable only the E cores vs stock?
The fact that you get same fps at 4.5 and 5.5ghz make me think that you still have bottleneck.
 
The only difference is the Intel cpu's are actually still pretty decent..... Bulldozer was not only inefficient but performed like shite......
It is true that the Intel chip still performed well. But we also need to remember that AMD was using GF 28nm vs Intel's 14nm when AMD introduced Bulldozer. So just the fab itself gave Intel a significant advantage.

I think the 13900KS needs some power justice here from what I've been reading in the comments.

My testing here is 8P cores only. If I need any more cores than that ill switch them on.

5.5Ghz - 66.7w.

View attachment 292582

4.5Ghz - 44.5w

View attachment 292583

4.5Ghz - 37.2w

View attachment 292585

As you can see with a bit of tuning these things can be very efficient.

I've been running mine on Air @ 4.5Ghz which is super cool, quiet and fast.

The reason for 4.5Ghz is because I'm not noticing any more FPS advantage over 5.5Ghz that's noticeable and temps, power and noise has been excellent.

If I need to do any more threading I've got 16 E-core's at my disposal plus HT if need be.
Reviews generally take out of the box experience since this is the way they are designed to be used. So while I understand your point, but I believe we can still squeeze more efficiency or performance out of the Zen 4 chips with 3D cache.
 
The fact that you get same fps at 4.5 and 5.5ghz make me think that you still have bottleneck.

Cyberpunk is a game where even low end CPU's wont make a diff in FPS especially at high res. I'm just showing the power consumption @ 5.5Ghz in game which I think is quite impressive with my settings.

All I'm doing is trying to show case a CPU that is typical of being a power hog doesn't have to be when manipulated the way you want to run it.

Yeah, Yeah, I get it people can write me off about "why didn't you just get this cpu etc" but I wanted the best silicone I can get for my money and am happy to run it how I like.

Do you know how much of the power consumption drop when you disable only the E cores vs stock?

Nah havent done this testing yet. Maybe down the track when Ill need games to run more cores.
 
Are you sure you got the CPU right? Because the 5950X is an AMD chip.

Yes, I sold my 5950X and bought one of these. CPU coolers come with mounts for both Intel and AMD CPUs, yes?
 
Why a review of this product since everyone and there dog knows this is just a binned 13900k and it's been out a good while now.
Even more pointless considering most reviewers don't include it in benchmarks because well the 13900k.
Hopefully anyone considering this CPU at least got the point not to use an air cooler.
 
Why a review of this product since everyone and there dog knows this is just a binned 13900k and it's been out a good while now.
Even more pointless considering most reviewers don't include it in benchmarks because well the 13900k.
Hopefully anyone considering this CPU at least got the point not to use an air cooler.

A competent (NH-D15 or similar heavyweight) air cooling solution will handle this CPU.

You're also given the option to configure its power limit manually.
 
Might I say the 13900KS does as well? Might not suck at performance, but sucks juice outta the mains with a pump-powered straw.
Youd be wrong then. The FX was inefficient, not just out of the box, but at whatever wattage. The 13900k and the 13900ks are not. Not at all. It's just that the stock settings intel has chosen are way outside its efficiency range. If you care about efficiency, the 13900 nonk and the 13900t are the most efficient cpus out of the box in existence right now. Grab em
 
A competent (NH-D15 or similar heavyweight) air cooling solution will handle this CPU.

You're also given the option to configure its power limit manually.
Was running my 13900k on a u12a. No problems whatsovever.
 
Intel ~ the latest entrant in the lucrative Toaster/oven business :peace:

cpu-temperature-blender.png

I'd like to know how this was achieved, my motherboard at least throttles the CPU down when it reaches 100 and it only touched such a temperature during torture testing with OCCT and XTU, where the CPU has exceeded 300W. I'm using an id cooling Frostflow X 360 with Kryonaut Extreme (the giga expensive pink one). Worth mentioning I suppose that I've also installed a thermalright contact frame, so it's an aftermarket socket ILM.

I didn't run Prime95 small FFT because I did not overclock the processor (nor do I have any real need to - it'll live a comfy life playing my Todd Howard classics), but the large FFT for RAM testing didn't get it much above 70.

In any case, on the MSI Z690 ACE, you're offered three presets for power the first time you enter BIOS, a "boxed cooler" preset with a 125W PL1, a high end tower cooler at around 250 watts limit (which imo realistically serves this processor well) and a water cooler preset which is fully unrestricted.

I only wish I could figure out why Windows 11 is so uncooperative with my system, as my thread detailed.
 
I'd like to know how this was achieved, my motherboard at least throttles the CPU down when it reaches 100 and it only touched such a temperature during torture testing with OCCT and XTU, where the CPU has exceeded 300W. I'm using an id cooling Frostflow X 360 with Kryonaut Extreme (the giga expensive pink one). Worth mentioning I suppose that I've also installed a thermalright contact frame, so it's an aftermarket socket ILM.

I didn't run Prime95 small FFT because I did not overclock the processor (nor do I have any real need to - it'll live a comfy life playing my Todd Howard classics), but the large FFT for RAM testing didn't get it much above 70.

In any case, on the MSI Z690 ACE, you're offered three presets for power the first time you enter BIOS, a "boxed cooler" preset with a 125W PL1, a high end tower cooler at around 250 watts limit (which imo realistically serves this processor well) and a water cooler preset which is fully unrestricted.

I only wish I could figure out why Windows 11 is so uncooperative with my system, as my thread detailed.
You have the option in bios to remove the thermal limit or raise it to 115c from the normal 100c. I assume that's what w1z did. The only other time I managed to exceed the 100c limit was when I was testing 5.6ghz all core on my 12900k @ 1.64 volts, the moment I launched a game it raced to 107c degrees.
 
Back
Top