Windows 11 vs. Ubuntu Linux Performance On The Intel Core Ultra 7 Meteor Lake

[duby229]

Linux has way better IO performance, a lot of what you're seeing is a rsult of Linux pushing bits around the system, ie between SSD and RAM, at a lower latency than Windows. Windows tries to do too much it has like a dozen processes that each can individually thrash a drive and then it makes them each try to thrash your drive simultaneously. Search, superfetch, defrag, defender, antimalware, windows update, etc, etc, and on and on and on.


This is why Linux outperforms windows^^^

EDIT: And has done so for -decades-, Windows 2000 was the last time MS had an OS that was actually capable of -IDLING-...

[smitty3268]

...

In all seriousness, I hope you are receiving proper treatment and live a good life.

[qarium]

Here's what you claimed:
You took exception with my statements that AMD's Threadrippers are a scam and a waste of money and responded with some garbage about Intel's "fake" E-cores.

I have 2 Threadripper systems and your claim that they are a scam are really nonsense.

if i had bought a ryzen system. the result would had be;
Maximum 8cores for the ryzen 1000/2000 generation and 16 cores for the ryzen3000/5000 generation
and maximum 64GB ram with 16gb modules on 4 DIMM slots.

my system is a Threatripper 1920X with 12cores and 128GB ram with 16gb modules in 8 dimm slots.
i can upgrade it to a 32core 2990WX and also up to 256GB ram with 32gb modules.

https://cpu.userbenchmark.com/Compare/AMD-Ryzen-9-5950X-vs-AMD-Ryzen-TR-2990WX/4086vsm560423

according to userbenchmark.com the 2990WX is 58% faster in massive multicore benchmarks than the fastest AM4 socked CPU 5950X

this alone shows that your statement "that AMD's Threadrippers are a scam" is absolute nonsense.
sure my example is historical but its not different with the newest Threadrippers

if you are happy with 16cores you can buy a AM5 system and use a 7950X or 7950X3D (or whatever intel system you like)

but if you do something what needs massive multithreat performance then a 64core Threadripper is the way to go.

what is a little sad in my point of view is that these TR5 mainboards can only handle 256GB of DDR5 ram. (with not yet released 64GB modules)

I responded by posting links to articles by Puget Systems where they tested Intel's i7 CPUs against AMD CPUs, including 64C/128T TR and in some benchmarks an i7 matched the 64C/128T TR and beat a 32C/64T TR.
This really pissed you off and you claimed that the links I posted somehow "hacked" your computer, but that both U.S. and German military cyber command units had alerted you to the "attack", stopped it and launched a "hack back" campaign. I think a judge also started an investigation, special forces were on the way, aliens from the Andromeda galaxy decided to have Bigfoot ride Nessie, The Loch Ness Monster on his way to see me.
It's possible that I took creative liberties with the last portion of the above, but the rest is accurate.

the benchmarks of course did not piss me off. i had not even time to read the benchmarks because the browser tab of this website crashed.

"aliens from the Andromeda galaxy decided to have Bigfoot ride Nessie, The Loch Ness Monster on his way to see me."

yes very funny. sadly for you the other entities are not so funny or friendly to deal with.

You claimed you had "proof" and posted a link to that Apache ActiveMQ CVE and claimed that you were hit by LogoFAIL and provided a link to that.
I have pointed out that Puget Systems is a reputable company with years of expertise that uses CloudFlare as their hosting provider.
I have pointed out that CloudFlare uses NGINX, not Apache and i have pointed out that one normally pairs KubeMQ with NGINX and have explained the reasons why.
I have also explained that the virus you claim you were hit by is only a proof of concept and does not exist in the wild and I have pointed out that the ActiveMQ venerability does not deliver LogoFAIL, it delivers a cryptominer.

there was never any virus it was a spy-trojan. and also cryptominers are very simple to detect because the CPU fan and the GPU fan spin high and loud but of course it was quiet. thats why your babbling about a cryptominer is nonsense.

The scenario you have concocted is that I am a member of Israeli Intelligence, I somehow know who you are, have targeted you because of your political views, I somehow managed to penetrated Cloudflares infrastructure, and infect one of their client's VM servers, without Cloudflare, or the client, Puget Systems noticing, with a virus that is capable of gaining root access on a fully up to date Fedora 39 install running an fully up to date Firefox build, the virus was capable of surviving a full format and reinstall the first time, but not the second time and you were the only one that was hit with this, no one else's computer was infected.
And all of this makes perfect sense to you?
To everyone else, how do I ignore a poster?
I think it's best if we give this user the silent treatment and not feed into his disillusion any more.

"fully up to date Firefox build"

this is plain and simple not true the exploid was for firefox 119 and the newest update packages in fedora 39 where version 120.

"you were the only one that was hit"

who knows this ? i do not know this. there is a possibility that other people where it to.

"but not the second time"

LogoFail fail's to execute if you disable the UEFI logo in the bios settings then you have a ugly text based bios post but the logo is not loaded at startup.

i am sure there are many users in this forum who would be happy if you give us all the "silent treatment"

[qarium]

In all seriousness, I hope you are receiving proper treatment and live a good life.

right. good wishes are always good.

[sophisticles]

Linux has way better IO performance, a lot of what you're seeing is a rsult of Linux pushing bits around the system, ie between SSD and RAM, at a lower latency than Windows. Windows tries to do too much it has like a dozen processes that each can individually thrash a drive and then it makes them each try to thrash your drive simultaneously. Search, superfetch, defrag, defender, antimalware, windows update, etc, etc, and on and on and on.


This is why Linux outperforms windows^^^

EDIT: And has done so for -decades-, Windows 2000 was the last time MS had an OS that was actually capable of -IDLING-...

All the things you describe can easily be disabled.

I think the main reason that Linux outperforms Windows is many benchmarks comes down to the following:

1) Poorly optimized open source software. I think many open source developers tend to write their code with Linux in mind and are not familiar with Windows optimization techniques and/or too lazy to include code paths targeting Windows.

2) I think this is another big reason, the Windows HAL. With Linux, all drivers are part of the kernel and the kernel runs bare metal, with Windows everything except the graphics drivers goes through the Windows HAL. This is a big reason why 3d graphics tests are faster on Windows than Linux.

The above notwithstanding, if MS wanted to they could easily restructure Windows to have an architecture similar to Linux or just do what Apple did years ago and base their next Windows version on BSD.

Maybe MS will do this, they already have WSL and a Linux distro for the cloud, so who knows?

[duby229]

All the things you describe can easily be disabled.

I think the main reason that Linux outperforms Windows is many benchmarks comes down to the following:

1) Poorly optimized open source software. I think many open source developers tend to write their code with Linux in mind and are not familiar with Windows optimization techniques and/or too lazy to include code paths targeting Windows.

2) I think this is another big reason, the Windows HAL. With Linux, all drivers are part of the kernel and the kernel runs bare metal, with Windows everything except the graphics drivers goes through the Windows HAL. This is a big reason why 3d graphics tests are faster on Windows than Linux.

The above notwithstanding, if MS wanted to they could easily restructure Windows to have an architecture similar to Linux or just do what Apple did years ago and base their next Windows version on BSD.

Maybe MS will do this, they already have WSL and a Linux distro for the cloud, so who knows?

Your number 1 up there -isn't- the problem at all... Open source software is open source for a reason and your number 1 up there -IS- that reason. Basically no open source software is poorly optimized.The term "Poorly optimized open source software" is an oxymoron of epic proportions. It is itself proof you have no idea what the hell your talking about.

and for your number 2, Microkernel dude... Hashed and rehashed a million times. MS proved them in commercial use decades ago, they aren't necessarily slower, just more complex. But that isn't what's slowing windows down.

As you say many processes that thrash windows so bad can be disabled by people like me an you.... But most people aren't like me an you.....

[sophisticles]

Your number 1 up there -isn't- the problem at all... Open source software is open source for a reason and your number 1 up there -IS- that reason. Basically no open source software is poorly optimized.The term "Poorly optimized open source software" is an oxymoron of epic proportions. It is itself proof you have no idea what the hell your talking about.

I think the fact that you think open source software is optimized shows that you have no idea what you are talking about.

If the software were optimized then simply compiling with more aggressive flags, or a different compiler, would not result in faster binaries but as has been shown many times when Michael ran tests of various compile flags or Clang vs GCC, a simple change can result in up to a 50% performance increase.

If the code is optimized by someone that knows what they are doing then you should not be able to find that much performance increase.

[SophTherapy]

Hello everybody,

I would like to apologize for my posts under nickname "sophisticles" and "hel88".

the thing is, I am very sick person. Schizophrenia with bipolar disorder.
When I'm on my medication like now, I feel ashamed for the things that I do when not on medication.

For example, when I'm not using my therapy properly I get this crazy tendency to troll on linux forums. For that devious purpose I am using nicknames "sophisticles" and "hel88". under those nicknames I write crazy, insane things. when I am on regular therapy like now, I cannot believe the crap that I wrote under those 2 nicknames.

overall, I would like all of you to know that I don't really mean what I write under those 2 nicknames and also, I love linux, open source and gpl. and yes, microsoft sucks.

[duby229]

I think the fact that you think open source software is optimized shows that you have no idea what you are talking about.

If the software were optimized then simply compiling with more aggressive flags, or a different compiler, would not result in faster binaries but as has been shown many times when Michael ran tests of various compile flags or Clang vs GCC, a simple change can result in up to a 50% performance increase.

If the code is optimized by someone that knows what they are doing then you should not be able to find that much performance increase.

More proof you have no idea what the hell you're saying. Different compiler flags do different things..... Obviously it depends on what your code is actually doing vs what the compiler is trying to do.... It makes perfect sense....