Intel Core i9-13900KS Review - The Empire Strikes Back

And this clearly shows us that Intel is far from Dead.

So, I don't know what squabbling about a couple percentage points here or there actually shows. AMD made some ground and depending on which article you cite they actually lost ground in the desktop space to intel...
AMD +2% say you and +10% says my source. That's more than a little. 5x difference. Keep pulling stats out of your ass.
 
AMD +2% say you and +10% says my source. That's more than a little. 5x difference. Keep pulling stats out of your ass.
Your stats and my stats don't mean shit. Because their market share is still, minimal. Intel is not dead. I don't have to pull anything out of my ass to prove that.

Keep on Shillin
 
Performance by watt, I feel that almost to the nose exactly what efficiency mean:
a) effective operation as measured by a comparison of production with cost (as in energy, time, and money)
b) the ratio of the useful energy delivered by a dynamic system to the energy supplied to it


Look at what it can do in eco-mode:

View attachment 566247

In 65 watt mode it does still beat a 170-195 watt 5950x
You just proved my point from earlier that I doubt it's more than 5-15% faster. That's only 10.77% maybe 11. And that's not my point, either and the above likely doesn't reflect constrained frequencies. It's not apples to apples. It had damn well better be more efficient at 5nm than 7.

It's a nothing burger uplift. The 7000 Series is an overclocked 5000 series chip on a better node.

And

No one runs the damn thing at 65 Watts... True efficiency gains would be a chip at like 30 Watts that is anything faster than the 5000 series. We won't see that, because we never have. Jesus, you could buy a Phenom II x6 that was 45 Watts back in the day. Not anymore...
 
You just proved my point from earlier that I doubt it's more than 5-15% faster. That's only 10.77% maybe 11.
The 34,300 (eco 105) is 34% higher than the 5950x while the 5950x is using nearly the double the watts than the 7950x.

No one runs the damn thing at 65 Watts...
People do run the latest zen in low wattage too, where it matters, laptop, when large cache on the 3d model make it hard to cool, etc...People do game running those around 65 watt, the 7800x3d average 50 watt while gaming or doing regular application
 
The 34,300 (eco 105) is 34% higher than the 5950x while the 5950x is using nearly the double the watts than the 7950x.


People do run the latest zen in low wattage too, where it matters, laptop, when large cache on the 3d model make it hard to cool, etc...People do game running those around 65 watt, the 7800x3d average 50 watt while gaming or doing regular application
I think your math is off a bit, 34% faster than 25,600 is 45,962...

843188_11_Eco_Cinebench_R23_nT.png

Now you're changing the conversation. First you were arguing 65 watts, then you just canged the narrative to suit your argument at 105 watts. At 105 Watts that's 25.4%

This is still not the comparison I was asking for from MY INITIAL POSTS. I stated in them I wonder how these chips would stack up at exact frequency constraints, not just TDP limitations. This is not an apples to apples comparison. The 5000 series maxes out (generally) around 4.65 Ghz, the 7000 series chips are boosting well beyond this even at constrained TDPs. Line em up side by side, clock for clock, the 7000 series is just an overclocked 5000 series chip on a better node with a bit more efficiency.

Nothing efficient about ANY modern processor. Because they all draw the same if not more power than the previous generation to get enhanced perfromance.
 
LukeTbk

"People do run the latest zen in low wattage too, where it matters, laptop, when large cache on the 3d model make it hard to cool, etc...People do game running those around 65 watt, the 7800x3d average 50 watt while gaming or doing regular application"

Wouldn't post the quote for some reason...

You're arguing around my point. The CPUs aren't marketed as efficient parts like their earlier predecessors. They ship with higher TDPs (in general).

We have already established from your examples above that the actual gains over the 5000 series are minimal and easily dismissed as a node advancement and perhaps some pipeline improvements.

Laptop Parts have to be in a constrained Wattage, yet we're seeing nothing innovative or wattage saving here either. And you can't compare laptop chips because they're cut down silicon with less cache memory and capabilities compared to their desktop counterparts. Because they run at a lower wattage does not make them efficient. They should be running at 1/2 their wattage, THAT would make them vastly more efficient than their predecessors.

The bold is so you know my text vs yours, I'm not yelling.
 
Because no games need the speed right now. Like all my previous CPUs I end up overclocking them down the line. I tend to keep my systems for a while. I already have the 6GHz all core settings written down, don't worry about that one bit lol it's just that I don't need to pump that amount of voltage into the chip at this very moment.
Also I love the binned CPU that does these super high clock speeds at a much lower voltage. Runs cooler in my PC room.

All that time and money spent for gaming performance and efficiency when the 7800x3d still beats it in gaming performance and efficiency right out the box.
https://www.techpowerup.com/review/intel-core-i9-13900ks/16.html
 
I think your math is off a bit, 34% faster than 25,600 is 45,962...
25,600*1.34 = 34.304, always a bit strange the faster than versus slower then phrasing.
Now you're changing the conversation. First you were arguing 65 watts, then you just canged the narrative to suit your argument at 105 watts. At 105 Watts that's 25.4%
No I am not all those examples are example of gain in efficacy, go at any watt the new Zen4 will have an impressive gain over the previous generation.

I was pointing out that even only at 65watt, using what 35% of the power of an 5950x it was still faster in something line cinebench.
We have already established from your examples above that the actual gains over the 5000 series are minimal and easily dismissed as a node advancement and perhaps some pipeline improvements.
We see a CPUs using around half the power and being 34% faster as minimal gain ? Having twice the fps in game with by watt with 8 core CPUs a minimal change ?

Can be i am sure dismissed as a node advancement, but that a giant node advancement, 3d cache seem to be both a TSMC achievement and an AMD design one.

You're arguing around my point. The CPUs aren't marketed as efficient parts like their earlier predecessors. They ship with higher TDPs (in general).
Except the 7950x3d-7800x3d where it was a big point, but yes desktop CPU will tend to go as high as they can where they can handle the temperature even if they are a lot in the diminishing return (one could say why not do it ?), which does not mean that the efficacy gain were not massive gen on gen and will not show up on mobile, server, gaming or if you power limit them to be able to see it.
 
Last edited:
25,600*1.34 = 34.304, always a bit strange the faster than versus slower then phrasing.

No I am not all those examples are example of gain in efficacy, go at any watt the new zen5 will have an impressive gain over the previous.

I was pointing out that even only at 65watt, using what 33% of the power of an 5950x it was still faster in something line cinebench.

We see a CPUs using around half the power and being 34% faster as minimal gain ? Having twice the fps in game by watt a minimal change ?

Can be i am sure dismissed as a node advancement, but that a giant node advancement.


Except the 7950x3d-7800x3d where it was a big emphased, but yes desktop CPU will go as high as they can where they can handled the temperature even if they are a lot in the diminishing return, which does not mean that the efficacy gain were not massive.
Man, either my calculator is bad or I just had a bout of retardation! LOL

This was my bad "I think your math is off a bit, 34% faster than 25,600 is 45,962..." I based this off of 34,300 .... Ugh

I totally fucked the math up... or I'm having an issue here.... 25,600 / 34,304 = 0.7462686567164179, which is 25.4% rounded up. I tried to do it backwards, I didn't use the right equation.

I should have done 34,304 - 25,600 = 8704 / 25600 = .34

Yup I'm retarded

Shit happens. At least I can admit it
 
Last edited:
25,600*1.34 = 34.304, always a bit strange the faster than versus slower then phrasing.

No I am not all those examples are example of gain in efficacy, go at any watt the new zen5 will have an impressive gain over the previous.

I was pointing out that even only at 65watt, using what 33% of the power of an 5950x it was still faster in something line cinebench.

We see a CPUs using around half the power and being 34% faster as minimal gain ? Having twice the fps in game by watt a minimal change ?

Can be i am sure dismissed as a node advancement, but that a giant node advancement.


Except the 7950x3d-7800x3d where it was a big emphased, but yes desktop CPU will go as high as they can where they can handled the temperature even if they are a lot in the diminishing return, which does not mean that the efficacy gain were not massive.
God I so love shitposting, I got carried away (y)
 
Last edited:
Mathhammer is hard.
Both CPUs are good.
Nothing is exceptionally efficient when pinned at 100%.
GPU will bottleneck long before either the 7800x3d or the 13900KS.

I look forward to the OMGWTFBBQ Ninja release of either processor where they do away with all power limits get an extra 200MHZ out of them but somehow double power consumption in the process.
 
All that time and money spent for gaming performance and efficiency when the 7800x3d still beats it in gaming performance and efficiency right out the box.
https://www.techpowerup.com/review/intel-core-i9-13900ks/16.html
Yea It's a nice chip. Although I am not interested in going AMD. I am comfortable with the Intel chips and prefer them over AMD CPUs and GPUs. Besides they trade blows in performance even if the new 3d chip is more efficient I'm not going to jump ship 😂 Also maybe the raptor lake refresh is going to be faster and/or more efficient who knows let's see. There is always something better on the horizon. So this argument of what's better or faster or more efficient is pointless. I'm happy with my gaming system 😃
 
You folk really like to argue :grabs popcorn:

My opinion is simple:
The 13900ks is a halo product targeted at people who want the best at any cost, much like the 4090.

Arguing power consumption about such a part is a bit of a farce.

Doing the intel vs AMD thing.. well the answer is “it depends”. All midrange processors from both brands are plenty enough for gaming and no benchmark has proved otherwise to me to date.

Regarding power, yeah intel uses more when it is running full out, but it isn’t running full out most of the time and a lot of benchmarks disable things like core parking and c states, which is counterproductive. AMD chips idle at a higher wattage than intel, when c states etc are enabled, because of the I/O die (seek to prove me wrong in that one)

Essentially it’s a wash, and that is a good thing for the consumer
 
Last edited:
The 13900ks is a halo product targeted at people who want the best at any cost, much like the 4090.
Yes for sure, but the execution is quite different the 4090 which achieved to be clearly the fastest "consumer" gaming GPU

This ? Not obviously the fastest gaming CPU (A common audience for that type of halo product), maybe it is with some really fast ram and overclocking, but by a very small amount:
average-fps-1920-1080.png
minimum-fps-1920-1080.png
minimum-fps-3840-2160.png


It seems to beat a 7950x in general in applications by a very small amount, but people that want the best in that realm have Threadripper pros, Epyc, Apple, new Xeon with 8 channel of DDR-5 ram to compete with, the 13900KS could be the best in some case, but I would imagine not obvious how often it would be the fastest system money can buy for your application.

Would it be clearly the fastest at something, even by a smaller amount, that would make it way more interesting.
 
Yes for sure, but the execution is quite different the 4090 which achieved to be clearly the fastest "consumer" gaming GPU

This ? Not obviously the fastest gaming CPU (A common audience for that type of halo product), maybe it is with some really fast ram and overclocking, but by a very small amount:
View attachment 566304View attachment 566305View attachment 566306

It seems to beat a 7950x in general in applications by a very small amount, but people that want the best in that realm have Threadripper pros, Epyc, Apple, new Xeon with 8 channel of DDR-5 ram to compete with, the 13900KS could be the best in some case, but I would imagine not obvious how often it would be the fastest system money can buy for your application.

Would it be clearly the fastest at something, even by a smaller amount, that would make it way more interesting.
Thanks for the charts. I have been on 4k for years so looking at the 4k results (I ignore the 1080p results even the 1440p because I'm never going below 4k) the results are within margin of error because games are more GPU bound/dependant at 2160p 144hz.
 
Thanks for the charts. I have been on 4k for years so looking at the 4k results (I ignore the 1080p results even the 1440p because I'm never going below 4k) the results are within margin of error because games are more GPU bound/dependant at 2160p 144hz.

And you will note that the midrange is (in general) within 5% of the high end.
 
Ya that's great for consumers. Don't we wish it was the same for GPUs lol
If some of the rumors are true of what Intel has on the go it very well could be, Cache quantities in the multiple GB, supposedly between Intels 2.5 and 3D stacking tech they can get something stupid like 32GB of L4 cache on a package.
 
Honestly, your numbers didn't prove anything. You entire point was to prove that Intel was dead. You failed in that regard. So, you can cite all the numbers you want to. But none of them indicate the demise of Intel.

Cheers
My point is to prove your numbers are bullshit. Mission accomplished. You lack basic reading comprehension skills.
 
My point is to prove your numbers are bullshit. Mission accomplished. You lack basic reading comprehension skills.
All you managed to prove is that you are a linear thinker and the "my numbers are better than yours" argument is retarded.
 
All you managed to prove is that you are a linear thinker and the "my numbers are better than yours" argument is retarded.
You've proven you are full of made up numbers you pull out of your ass to try and prove your point. Basic shill definition.
 
You've proven you are full of made up numbers you pull out of your ass to try and prove your point. Basic shill definition.
https://mezha.media/en/2022/11/12/x...and-laptops-and-improves-position-in-servers/
That one is from December of last year. That's even more bleak than your numbers.

x86_CPU_Market_Share_Graph.png

https://hothardware.com/news/amd-steals-intels-lunch-money-increasing-cpu-market-share

https://wccftech.com/amd-grabs-over...re-with-epyc-cpus-intel-continues-to-decline/

AMD Q4 2022 x86 CPU Market Share (via Mercury Research):​

MERCURY RESEARCHQ4 2022Q3 2022Q2 2022Q1 2022Q4 2021Q3 2021Q2 2021Q1 2021Q4 2020Q3 2020Q2 2020Q1 2020Q4 2019Q3 2019Q2 2019Q1 2019Q4 2018Q3 2018Q2 2018Q1 2018
AMD Desktop CPU Market Share18.6%13.9%20.6%18.3%16.2%17.0%17.1%19.3%19.3%20.1%19.2%18.6%18.3%18.0%17.1%17.1%15.8%13.0%12.3%12.2%
AMD Mobility CPU Market Share16.4%15.7%24.8%22.5%21.6%22.0%20.0%18.0%19.0%20.2%19.9%17.1%16.2%14.7%14.1%13.1%12.2%10.9%8.8%N/A
AMD Server CPU Market Share17.6%17.5%13.9%11.6%10.7%10.2%9.50%8.9%7.1%6.6%5.8%5.1%4.5%4.3%3.4%2.9%4.2%1.6%1.4%N/A
AMD Overall x86 CPU Market Share31.3%28.5%29.2%27.7%25.6%24.6%22.5%20.7%21.7%22.4%18.3%14.8%15.5%14

21 to 22 AMD's Desktop Market Share only went up by 2% (which was accurate-ish), technically 2.4%

Your numbers are wrong as well

have a nice day
 
Last edited:
Y’all getting’ caught up on some pretty minor stuff, AMD is doing good, they are putting serious pressure on Intel and forced them to restructure and find new ways to compete.

The AMD & TSMC tag team are putting on some hurt and it’s about time. Intel put bean counters in charge of an engineering based tech company and they made a decade of shit decisions based on a perceived position of absolute authority and now AMD has forced Intel to walk it back, and figure their shit.
This is good for us, Intel needed to slim down and actually focus on something.
 
https://mezha.media/en/2022/11/12/x...and-laptops-and-improves-position-in-servers/
That one is from December of last year. That's even more bleak than your numbers.

View attachment 566559
https://hothardware.com/news/amd-steals-intels-lunch-money-increasing-cpu-market-share

https://wccftech.com/amd-grabs-over...re-with-epyc-cpus-intel-continues-to-decline/

AMD Q4 2022 x86 CPU Market Share (via Mercury Research):​

MERCURY RESEARCHQ4 2022Q3 2022Q2 2022Q1 2022Q4 2021Q3 2021Q2 2021Q1 2021Q4 2020Q3 2020Q2 2020Q1 2020Q4 2019Q3 2019Q2 2019Q1 2019Q4 2018Q3 2018Q2 2018Q1 2018
AMD Desktop CPU Market Share18.6%13.9%20.6%18.3%16.2%17.0%17.1%19.3%19.3%20.1%19.2%18.6%18.3%18.0%17.1%17.1%15.8%13.0%12.3%12.2%
AMD Mobility CPU Market Share16.4%15.7%24.8%22.5%21.6%22.0%20.0%18.0%19.0%20.2%19.9%17.1%16.2%14.7%14.1%13.1%12.2%10.9%8.8%N/A
AMD Server CPU Market Share17.6%17.5%13.9%11.6%10.7%10.2%9.50%8.9%7.1%6.6%5.8%5.1%4.5%4.3%3.4%2.9%4.2%1.6%1.4%N/A
AMD Overall x86 CPU Market Share31.3%28.5%29.2%27.7%25.6%24.6%22.5%20.7%21.7%22.4%18.3%14.8%15.5%14

21 to 22 AMD's Desktop Market Share only went up by 2% (which was accurate-ish), technically 2.4%

Your numbers are wrong as well

have a nice day
Nice, move the goalposts to desktop only. Pathetic. How is limiting market share analysis to Desktop only a valid representation? It's misleading.
 
Last edited:
Yeah, intel should focus !

This 13900ks tries to do too many things a once.

Like it wants to be an an egg laying woolmilk pig.

Personally i´ve deactived the e-cores and hyperthreading.

I am running it as an all-core 6 Ghz rig on a 4k 120Hz 50 inch TV with a 4090.

My Windows 10 latency and ssd transferrate do appreciate this (+5% IPC due to hyperthreading deactivation)

Due to my 4k/120Hz limitation I cannot even be bothered to check if the games run slower or faster, soo...

I simply run 6 Ghz on an AIO without ever hitting the thermal limit.

And THAT puts a smile on my face every time i turn the rig on. :)


Rant on e-cores in a gaming rig:

Why would i let my background tasks ever run slower than they could except in a hotel with a laptop whilst having forgotten my powerbrick?

I´m too old to let ANYTHING run on slow waste cores.
 
Last edited:
Yeah, intel should focus !

This 13900ks tries to do too many things a once.

Like it wants to be an an egg laying woolmilk pig.

Personally i deactived the e-cores and hyperthrading and i am running it as an allcore 6Ghz rig on a 4k 120Hz 50 inch TV.

My Windows 10 latency and ssd transferrate do appreciate this.

I cannot be bothered to check if the games run slower or faster, since i have a 4090, so...

I simply run 6Ghz on an AIO without ever hitting the thermal limit.

And THAT puts a smile on my face every time i turn the rig on.

For me it was a good investment.:)
Hyper-V and thread director have been really good at keeping all my little VM's over on the E cores and leaving them off the P cores so that has made doing stuff a fair bit easier for me personally so I see it as a win.
I am looking forward to the Sierra Forest Xeons, an HCI stack of all E cores for hosting all the tiny VM's that gobble ram but have very small workloads, delicious.
 
Yeah, intel should focus !

This 13900ks tries to do too many things a once.

Like it wants to be an an egg laying woolmilk pig.

Personally i´ve deactived the e-cores and hyperthreading.

I am running it as an all-core 6 Ghz rig on a 4k 120Hz 50 inch TV with a 4090.

My Windows 10 latency and ssd transferrate do appreciate this (+5% IPC due to hyperthreading deactivation)

Due to my 4k/120Hz limitation I cannot even be bothered to check if the games run slower or faster, soo...

I simply run 6 Ghz on an AIO without ever hitting the thermal limit.

And THAT puts a smile on my face every time i turn the rig on. :)


Rant on e-cores in a gaming rig:

Why would i let my background tasks ever run slower than they could except in a hotel with a laptop whilst having forgotten my powerbrick?

I´m too old to let ANYTHING run on slow waste cores.
Well you have effectively turned your CPU into an 8C/8T one. Why not save $300 and just get the 13700k?
 
The way I see them right now is AMD makes boutique silicon, they are using the best processes and making small batch releases of what is in demand.
Intel is the bulk provider, they make a little of everything and it might not reach the best performance metrics at the best efficiency, they can produce a shitload of them for relatively cheap.
It's a good old-fashioned Ford/Fararri thing.
 
Nice, move the goalposts to desktop only. Pathetic. How is limiting market share analysis to Desktop only a valid representation? It's misleading.
You bore me. You've backed up your argument with " your word". I have provided multiple sources and you in all of them you are incorrect. I was originally citing Desktop Market Share. Unfortunately, dealing with you is like feeding a troll. I'm done with you.
 
Last edited:
The way I see them right now is AMD makes boutique silicon, they are using the best processes and making small batch releases of what is in demand.
Intel is the bulk provider, they make a little of everything and it might not reach the best performance metrics at the best efficiency, they can produce a shitload of them for relatively cheap.
It's a good old-fashioned Ford/Fararri thing.

2nd place performance metrics with complete dogshit efficiency. FTFY.
 
You bore me. You've backed up your argument with " your word". I have provided multiple sources and you in all of them you are incorrect. I was originally citing Desktop Market Share. Unfortunately, dealing with you is like feeding a troll. I'm done with you.
Intel '21 fourth Q DEC total net revenue - $20,528M
Client Computing (desktop) - $9,408

Desktop is less than half of intel's business. Not nearly representative of the company. You are just taking the "recovering" part of the business to paint it in the best light, after intel desktop took a beating the last couple years.
 
https://mezha.media/en/2022/11/12/x...and-laptops-and-improves-position-in-servers/
That one is from December of last year. That's even more bleak than your numbers.

View attachment 566559
https://hothardware.com/news/amd-steals-intels-lunch-money-increasing-cpu-market-share

https://wccftech.com/amd-grabs-over...re-with-epyc-cpus-intel-continues-to-decline/

AMD Q4 2022 x86 CPU Market Share (via Mercury Research):​

MERCURY RESEARCHQ4 2022Q3 2022Q2 2022Q1 2022Q4 2021Q3 2021Q2 2021Q1 2021Q4 2020Q3 2020Q2 2020Q1 2020Q4 2019Q3 2019Q2 2019Q1 2019Q4 2018Q3 2018Q2 2018Q1 2018
AMD Desktop CPU Market Share18.6%13.9%20.6%18.3%16.2%17.0%17.1%19.3%19.3%20.1%19.2%18.6%18.3%18.0%17.1%17.1%15.8%13.0%12.3%12.2%
AMD Mobility CPU Market Share16.4%15.7%24.8%22.5%21.6%22.0%20.0%18.0%19.0%20.2%19.9%17.1%16.2%14.7%14.1%13.1%12.2%10.9%8.8%N/A
AMD Server CPU Market Share17.6%17.5%13.9%11.6%10.7%10.2%9.50%8.9%7.1%6.6%5.8%5.1%4.5%4.3%3.4%2.9%4.2%1.6%1.4%N/A
AMD Overall x86 CPU Market Share31.3%28.5%29.2%27.7%25.6%24.6%22.5%20.7%21.7%22.4%18.3%14.8%15.5%14

21 to 22 AMD's Desktop Market Share only went up by 2% (which was accurate-ish), technically 2.4%

Your numbers are wrong as well

have a nice day
AMD has been growing market share in the most lucrative, if I remember right, area, which is datacenter. That's paying a lot of bills. They need to fix their mobile trend though.
 
AMD has been growing market share in the most lucrative, if I remember right, area, which is datacenter. That's paying a lot of bills. They need to fix their mobile trend though.
I have no argument there. I had been in this running fight with someone on here that said Intel was dead (or strongly inferred that they were dying). That's not the case, even with AMD's growing market share. Lets hope for AMD that it not only pays the bills but keeps them competitive with Intel in the years ahead.

If we are lucky, next year will fix the mobile trend if Leaks about Strix Halo are true. With AMD possibly bringing a 40CU Graphics Card to their APUs with 32 Megs of Infinity Cache on the high end. Even the mid point of the Strix GPU stack will be impressive. We will finally be able to game on Battery. Take that with some salt, Adored and Moores Law is Dead are both spouting off about it. Mobile graphics that kick ass are exciting, but they need market segment penetration here terribly. I have a 5700u laptop and I have been very happy with it. It's basically cut down 5000 series everything but I was surprised I could play Battletech and Sins Of a Solar Empire II on it. A bit choppy on the latter title but cool nonetheless for a bullshit APU.
 
Last edited:
2nd place performance metrics with complete dogshit efficiency. FTFY.
Not everyone cares about efficiency these days so long as the processors do what people need em to do. Intel does have TDP limited processors for OEMs, they can be run more efficiently at performance loss. However, Lakados has a point. Intel is churning out these processors on a mature 10nm node. The cost to produce these has to be low. Plus, it's all done in house. A huge cost savings.

Intel has widespread adoption. It's everywhere in every company. I have never seen someone go exclusively Team Red. AMD is gaining but for how long, no idea. They do seem to be kicking ass with their Epyc processors though. Only time will tell.
 
Back
Top