"It's only good for workstations" le meme

> "It's only good for workstations" le meme
Stop repeating that stupid meme, it's not good for Workstations either in most cases. Only for 1% of the time you ever use a computer you leave it the fuck alone to render something. For 99% of the time, you will be actually using the computer in an interactive manner and interactivity can not be parallelized well on CPU space.
Let alone if you knew anything about computers you'd know most of the parallelism nowadays is done by GPUs, NOT CPUs, and that notion might become even MORE prevalent, NOT less with the newest GPU APIs that offload even more and more stuff from the CPU space to the GPU space even if they claim to help more at multithreading *but that is not their main advantage, the batch commands are.
inb4 niche-usage guy that runs some kind of mesh software with his professor.
you are still a tiny minority

Other urls found in this thread:

anandtech.com/show/11170/the-amd-zen-and-ryzen-7-review-a-deep-dive-on-1800x-1700x-and-1700/18
anandtech.com/show/11170/the-amd-zen-and-ryzen-7-review-a-deep-dive-on-1800x-1700x-and-1700/20
archive.rebeccablacktech.com/g/?task=search&ghost=yes&search_text= "It's only good for workstations" le meme
twitter.com/SFWRedditImages

Rizen is shit, my 3930k which I bought in 2010 for less than what rizen costs now is just as powerful. Mind you it now goes for about $140 on eBay, 1/3 of the top of the line rizen CPU.

Oh, it's that thread again. Stop shilling, mate.

>newest AMD chip raped by any and every intel part.


noice

>Literally post the same thread again
Intels marketing team isn't even trying anymore

>single threaded benchmark
Oh no! Applying my image filters will take .2 seconds slower!

>real world cases don't conform to my narrative
>oh no let's find a niche usage
post gentoo compiling again

I want the Intel shills to leave

off fucking topic but why do people choose gentoo over other DIY distros

the fucking compiling is stealing all my CPU cycles

>video encoding, 3d rendering and physics simulation aren't real world

whatever company provides the cheapest 8 core multithreaded cpus is what gets put in my computer farm

more cores at lower cost = more virtual servers at higher profit margin

a few percentage disadvantage in performance doesnt matter when the only affordable option on intel's side for years has been cheapo salvage xeons

i host minecraft servers for a living. i have to keep my costs low enough so that kids can get mommy to buy them a server

once the costs on the 1700 model come down a bit i can stop scouring ebay for old overworked 8 core xeons that i try to push a few extra months of life out of

Read the OP
> Only for 1% of the time you ever use a computer you leave it the fuck alone to render something. For 99% of the time, you will be actually using the computer in an interactive manner and interactivity can not be parallelized well on CPU space.
>Let alone if you knew anything about computers you'd know most of the parallelism nowadays is done by GPUs, NOT CPUs, and that notion might become even MORE prevalent, NOT less with the newest GPU APIs that offload even more and more stuff from the CPU space to the GPU space even if they claim to help more at multithreading *but that is not their main advantage, the batch commands are.
>inb4 niche-usage guy that runs some kind of mesh software with his professor.

>computer farm
>Workstations
you are off topic

...

Oh that paranoid guy that thinks everyone that isn't him is 1 guy. He is in need of medical attention.

Also, unless you want your game physics to be locked to a single GPU vendor, you have no choice but to run it on a CPU.

>here is another niche usage
get over it paranoid fuck
the real world does work on computers that the "moar cores" meme fails, and 2-3 very strong cores win.

>AYYYMDPOOJEETS ON SUICIDE WATCH

>niche usage
Physics is not a niche usage, you stupid fuck. It's literally used in every single 3d game out there.

Why aren't we using Prescott Pentiums again? One single core, more than anyone needs.

and you have tesla for that

>posting worst girl
At least shitpost with Velvet.

yet your narrative does not seem to be supported by the overwhelming majority of benchmarks in gaming or workstation use of the real world, on software that people actually use.
get over it. the niche usage you googled does not cut it.
most people can use 2-3 strong cores > 99 weak.

Yes, Tesla for serious scientific calculations where you're not concerned with being locked to a single vendor. Otherwise, you would want to run your physics on CPU if you want to reach the widest audience possible.

anandtech.com/show/11170/the-amd-zen-and-ryzen-7-review-a-deep-dive-on-1800x-1700x-and-1700/18

anandtech.com/show/11170/the-amd-zen-and-ryzen-7-review-a-deep-dive-on-1800x-1700x-and-1700/20

You can stop posting now

I'm sorry you have shit taste.

>he proceeds to post rendering and encoding results
read the OP and fuck off:
>Only for 1% of the time you ever use a computer you leave it the fuck alone to render something. For 99% of the time, you will be actually using the computer in an interactive manner and interactivity can not be parallelized well on CPU space.
>Let alone if you knew anything about computers you'd know most of the parallelism nowadays is done by GPUs, NOT CPUs, and that notion might become even MORE prevalent, NOT less with the newest GPU APIs that offload even more and more stuff from the CPU space to the GPU space even if they claim to help more at multithreading *but that is not their main advantage, the batch commands are.
>inb4 niche-usage guy that runs some kind of mesh software with his professor.

No you. Velvet a best.

t. shit taste

>Only my benchmarks in applications that doesn't properly make use of 8 cores and 16 threads are valid and my use case is the only use case that exist on the planet
You're mentally ill

t. dumb Magilou poster

>Implying heavily multithreaded workloads aren't the most appropriate programs to test on a multi-core CPU
You have no point unless my GPU has the ability to compile code.

Ayymds on suicide watch

amd on lifesupport

>The AutoCAD results remind us of what we reported on the previous page when we looked at output devices. AMD’s Ryzen 7 1800X ends up exactly where you’d think, based on those findings. There’s no real parallelization, but there’s plenty of system memory and cache usage, which turns out to be a combination that puts its performance squarely in the middling category.
>There's no real parallelization

If you are that guy that thinks everyone is the same "intel guy", you are the one that is in need of medical attention I'm sorry to say.
The reality is, the fact you found a niche usage with your professor is not the mainstream workstation.
The real world does not work with whatever niche usage you googled, it works with OP pic and related.

>interactivity can not be parallelized well on CPU space.

ur full of shit m8

You know the easiest way to have parallel interactivity processing in your PC? Open more than one program at once.

>appropriate programs to test on a multi-core CPU
Do you even what this topic is? It's specifically about Workstations.
Workstations, do human WORK, they do not do server farms.

>I have no argument to the greentext and the other post
>I'm just going to assume they're the same person and call them mentally ill

Ryzen is fucking horrible

Literally the worst CPU launch I have ever seen

You mean like playing a game and streaming at the same time?

Oh wait...

>Preset is 3.5mbits of laggy, artifacting video with a whole load of drop frames
You literally can't make this shit up

Exactly, real world cases of real world workstations often just not need parallelism on the CPU level for 99% of the time.
In fact, Autodesk DOES have software that can render offline, like CAD software that does even ray tracing.
But, most people either never use that, or when they do, it often can or should better be done with GPUs.

Nah, I'm pretty sure it's the guy who keeps copypasting the same OP who's mentally ill, unless he's on the Intel payroll which I doubt since he seems to have a double digit IQ

archive.rebeccablacktech.com/g/?task=search&ghost=yes&search_text= "It's only good for workstations" le meme

And no, rendering and encoding isn't niche tasks no matter what you need to tell yourself to be able to sleep at night.

>Open more than one program at once.
And 99% of users will have them idle. They stay in the background doing nothing. Stop reaching.

get a load of this faggot

???

>rendering and encoding isn't niche tasks

Jesus AMDtards truly do live in an alternate universe.

>Ray Tracing
>GPU
Try making sure you actually know shit before you post.

AMDtards have no idea whatsoever how computers work, they think the cores are running full blaze at all times.

If using Adobe Premiere is a niche task then so is Photoshop, you retard

>rendering and encoding isn't niche tasks
It's literally you and the other 4 criminals of the "scene" that use encoding so much.

I'm encoding right now.

And heavy photoshop usage, somehow, isn't when it conforms to your opinions

Again you are completely out of touch with reality

Way more people use Photoshop than Premiere

Then why are you using that benchmark to make a point?

Which is still a niche use case. The only thing that isn't a niche use case for Intel kiddies is video games.

>ray tracing
>GPU
Try making sure you actually know shit before you post

Actually you can run physics on any GPU in parallel with normal rendering using Vulkan.

really?

7700K
With OBS
-11.8% Average FPS
-20.7% Min FPS

Ryzen
With OBS
-2.7% Average FPS
-3.4% Min FPS

AMDtards truly are braindead, look at the actual FPS and tell me why on earth you would buy a $500 Ryzen CPU that performs worse in streaming than a $340 Intel CPU

True games are hugely popular, way moreso than photoshop or basically any software other than like Microsoft Office

Guess what: Ryzen fails at both!

You can look at the difference between the average and the minimums and then gauge how much frames are being dropped by the 7700k.

>shaving .2 nanoseconds off of that =SUM function
Might as well get a pentium for all it's worth

Ars used the shittiest possible settings in their test, settings that would yield a stream full of stuttering and artifacts. Any high bitrate recording will kill the frame rate on the i7. Everyone knows this. Even streamers playing games like Fallout 4 at 1280x720 have frame rate issues on their mainstream i7s.

D A Y 0
A
Y
0

multitasking can not be parallelized well on cpu

THE FUCK DID I JUST SAW

A retard that keeps on posting the same shit over and over again ever since the Ryzen release.

Really. IMO it is the major feature of Vulkan, but all reviews I've read when Vulkan came out did not pay enough attention to it. But you can learn about it from Khronos conferences videos on YouTube. The problem is that Nvidia is not going rewrite their libraries for Vulkan for obvious reasons, and Havok who were a large provider of physics libraries is now acquired by M$. And as you know M$ has their on API. Most likely M$ bought Havok to prevent them from building APIs for Vulkan.

>games
>gaymes
>gumes
>games

I hate this board. I can make up a shit program for atrocious coding and have BOTH Intel and AMD do shit.

Maybe use something that's not verified garbage to test it.

man i cant even begin to question some shit i saw here...

as a former autocad user i need to remind those idiots here that autocad doesnt render anything on 2d anymore since 2014 is highly suspicious on how and why they choose to do that..

then did anyone did a live rendering or live editing test?

did anyone saw any virtualization test?

So? It still has higher average (absolute)

Ryzen 7 1800X
>With OBS average FPS 108
>With OBS 99th Percentile FPS 54
7700k
With OBS average FPS 135
With OBS 99th Percentile FPS 69

It's amazing that AMD shills will take the fact that 7700k runs HIGHER frames by default and because of that has higher drop with OBS and try to use that as proof that AMD is better.
EVEN THOUGH with the drops of OBS the 7700k is STILL FASTER.
It's literally faster in every possible way while streaming and even more so when not streaming!

It's not proof that AMD is better; it's proof that the 4/8 7700K is tapped out and no sane person would spend $300+ on it.

Why do you retards try to benchmark CPUs with fucking video games anyway, at a gimped resolution that nobody since mid 2000s has used

and is that drop not testaement to something faulty in INtel?

let me give a hint

ryzen is currently facing
a) an imporper balancing mapping
b) a virtual l3 size of 138mb on which the cpu has only 16mb...
c) noumerous bios problems resulting on lower imc and having no oc capabilities on xfr currently on
d)the minor core parking which is currently being fixed to release on the 17.3.1 drivers

and still despite everything that is currently crippling them THEY ARE SO DAMN CLOSE to 6900k and 7700k and you idiots dont realise that by shilling so much against them you are literally creating the perfect enviroment

This. It's incredible that AMD manage to make Ryzen in the position they were, and it's outstanding that the Ryzen CPUs manage to be so close to the 6900k and 7700k even with all the problems they are facing.

Which MoBo manufacturer has the best support for 1800X for now?

I think ASRock and Gigabyte are fairing better than ASUS and MSI right now.

AMD kids are desperate. They constantly try to google niche uses nobody in the real world cares about. They constantly invent conspiracy theories.

The latest conspiracy theory is that they need more time to get better software support on Level 2/3 caching.

They are so mentally challenged they don't realize Level 2/3 caching is not manipulated by software.

software reading the cache wrong can result in actual performance hits, though

It's obvious it's a conspiracy theory that was invented after the fact of their failures. AMD engineers might be stupider than Intel but they are not that stupid. They would tell you before-hand that they failed at that, they would not tell you after the fiasco that "oh wait, look at this failure".
In any case it's a failure.

Windows load balancing haphazardly moving a single thread between the CCXs incurs a large latency hit, and that absolutely does impact performance.

Don't respond to shitposting shills. They're not even people.

$20 says once this
and this
Get fixed my MS... when ever that happens, gaymen performance will be a lot closer.

Plus add in game dev patches and optimizations for future titles.

They never learn. Most people neither render 100% of the time, or do single-threaded 100% of the time. They do both a little.

Mentally sound people seek balance. They know how to find the sweet spot. And the sweet spot is obviously 4c/8t.

The mentally unstable go to extremes. This isn't about your personal niche use. It's about the average Work PC.

I like how 6900k and 6950x become unwanted nuisances for Sup Forums only after Ryzen was released.

> Level 2/3 caching is not manipulated by software.
Ryzen has 2 core complexes and each of them have separate caches. Windows scheduler moving threads between cores in different complexes result in bad performance

That's complete bullshit, their settings are fine

>all games are programmed badly

I love how AMDtards blame the software, when 99% of software runs like shit on Ryzen.

AMD's job is to make a fucking CPU that works with the software that is already out and they failed miserably.

>3.5mbits/sec bitrate
>no other details
Considering that the broadwell-e is being beaten, that benchmark is completely bullshit

AMD shills are truly insane. They live in an alternate universe.

>t-t-t-t-t-he benchmark is fake!

The last bastion of AMDelusion always comes out when they have no other retarded way of spinning reality.

No. No they're not.
How can you dipshits just say things and not expect to be called out for it?

So the 6900k is a shit CPU too since 99% of benchmarks that are being posted shows it being beaten by the 7700k?

>No details other than that low bitrate
>broadwell-e being beaten by a 7700k in realtime video encoding
>literally no argument

The mentally unstable of the AyyMD meme don't realize that the real subject isn't what they do in their basement. The real question is what most people do at their work. They do not encode all day or they don't compile gentoo all day, and they don't do single threaded gaming all day either.

>THE NUMBERS ARENT REAL
>REALITY IS FALSE

AMDelusion off the charts

Didn't you know?
A Kaby Lake i3 is the best processor on the planet. Overclock it to 5ghz and it beats everything.
i5 and i7fags eternally BTFO. Workstation users BTFO. Enterprise market BTFO. A 20k node HPC cluster can't even run CS:GO at 400fps but an i3 can. BTFO.

It's even worse when they try to shill on Sup Forums
Literally denying reality tier.

You have to prove that reality first, m8. Show me the fucking settings used