Userbenchmark updated their cpu performance index calculation to reflect more on single and quad core results...

userbenchmark updated their cpu performance index calculation to reflect more on single and quad core results, and a lot less on multi core results
if you check the top 10 and top 20 cpus, you'll notice it's mostly intel now, even though in actual WORK-related tasks they're vastly inferior to AMD's offerings

previous weighing:
>10% multi core
>60% quad core
>30% single core
new weighing:
>2% multi core (KEK)
>58% quad core
>40% single core
the resulting move displays AMD cpus at a major disadvantage, making them look as if they're a lot slower than in reality
now a 8350 has a higher CPU performance index than a 2990WX
post yfw userbenchmark pulls a
>cores don't matter!
>multithreading doesn't matter!

Attached: 1562344082422.jpg (386x455, 32K)

Other urls found in this thread:

cpu.userbenchmark.com/Faq/What-is-the-effective-CPU-speed-index/55
cpu.userbenchmark.com/Compare/Intel-Xeon-W-3175X-vs-Intel-Core-i3-9350KF/m797172vsm775825
cpu.userbenchmark.com/Compare/AMD-Ryzen-5-2600X-vs-AMD-Ryzen-TR-2990WX/3956vsm560423
cpubenchmark.net/singleThread.html
cpubenchmark.net/high_end_cpus.html
guru3d.com/news-story/new-cpu-z-upgrade-lowers-ryzen-performance.html
twitter.com/NSFWRedditImage

cores don't matter
multithreading doesn't matter

lolwut
Who actually uses userbenchmark? Gaymers? Anyone with a defined use case already should know what they're going to buy

forgot link: cpu.userbenchmark.com/Faq/What-is-the-effective-CPU-speed-index/55
>Gaming CPU performance does not typically scale well with core count. Extra cores work well for server and workstation orientated workloads where there are often several CPU intensive tasks running simultaneously. Beware the army of shills who would happily sell ice to Eskimos.

And why is there a unified score in the first place? Why not split it into single/quad (why quad and not dual?)/multi numbers from the get go?

classic shenanigans. oh the CPU we expect to win is losing? then let's redefine what's important.

then why is qc 58% of the score?

yeah dude, cores don't matter, that's why the game I've made and released is aware of as many threads as your system can provide and the more cores you throw at it, the higher your fps is going to be

>higher your fps

>semantics
you know exactly what I mean

Yeah fuck you Jow Forums. Fucking shilling your ice and shit. Intel is future!

I think you meant "single core is the future"

damn, that's some spicy damage-control

>video encoding speed doesn't matter
whatever helps you sleep at night

>Why not split it into single/quad (why quad and not dual?)/multi numbers from the get go?
What are you talking about? There's always been a split score in there if you scroll down just a little, I always ignored the unified score because it never made any sense and compared by the split partial scores.

>cpu.userbenchmark.com/Compare/Intel-Xeon-W-3175X-vs-Intel-Core-i3-9350KF/m797172vsm775825

A I3 9350K now beats a 28-core Intel Xeon 3175-X.
AHAHAHAHA

a 8350 and 2990WX is 60% apart
60
fucking
percent

S I X T Y
F U C K I N G
P E R C E N T

Attached: intel wojak.png (633x900, 162K)

oh jeeze

Attached: bite.jpg (674x538, 50K)

intel btfo themselfs

bravo paying your shills to btfo yourself intcel

>now a 8350 has a higher CPU performance index than a 2990WX
NO GOY YOU DON'T UNDERSTAND

jesus fucking christ this site is a fucking joke now

can someone please post the current and up-to-date "doesn't matter" list?

>current and up-to-date
it's never current nor up-to-date, it is constantly expanding.

aay version will suffice

lists don't matter

POOZEN PATCHED INTO IRRELEVANCE INTEL WINS AGAIN BABY

Attached: gkek.jpg (5000x5000, 1.76M)

>Intel bribes ubm to shift the goal post
Oh wow, as if they weren't showing us how desperate they are already.

What the fuck is wrong with these retards

cpu.userbenchmark.com/Compare/AMD-Ryzen-5-2600X-vs-AMD-Ryzen-TR-2990WX/3956vsm560423

This will not age well.

AMD BTFO once again.

top fucking kek

Jow Forumsbros we can do something

on the very bottom of the userbenchmark page, covered by a popup you can close, is a 'feedback' button. spam that with 'Our effective CPU index is now sponsored by Intel:TM:.' or something similar

>lowering the multicore weight this hard
>when more and more games are starting to actually use those cores
Seriously, just ask the retards who fell for the 7700K how it performs on modern games, it's fucking retarded. At least they should have replaced the quad core test with six or eight cores.

What a crock of shit, GTA V and bf1 ran like shit on my quad thread 4690k @4.5ghz. nothing but CPU bottlenecking.

4 CORES 4 EVA LOL

Quad cores are a fucking joke for gaming these days.

Yeah intelbux have been flowing like mad lately, they know they have nothing

i told them they were retarded niggers

BUY INTEEEL AAAAGHHHHHHHHHHHHHHHHHH

Attached: 1527629778452.jpg (679x758, 54K)

First CPU-Z, now userbenchmark.

Next up is Passmark and Geekbench, right?

So when are THEY gonna patch passmark?

cpubenchmark.net/singleThread.html
cpubenchmark.net/high_end_cpus.html

Attached: 1436734391281.jpg (202x249, 30K)

CPU-Z is intel biased now?
didn't use it for a few years but alwaysed liked the program

CPU-Z got "patched" after they found out AMD's branch predictor owned the benchmark. Happened like a year ago.
The excuse was "it's not a realistic performance indicator" or something

guru3d.com/news-story/new-cpu-z-upgrade-lowers-ryzen-performance.html

speccy is intel biased too.
they won't patch the temp reading
thanks

so this IS damage control

OY VEY PATCH IT GOYIM

Attached: subversion.png (614x424, 22K)

You fucking bet your reptilian ass they're going to, Schlomo.

haven't been to Jow Forums in ages
when did we get so anti-semitic
i love it

> Beware the army of shills who would happily sell ice to Eskimos.

Attached: palpatine ironic.png (500x333, 104K)

Single core performance is all that matters. 99% of video games and 90% all software are single core.

So what is Intel going to do if Ryzen 4000 beats them in single core across the board?

Attached: question mark 6.jpg (1920x1080, 120K)

userbenchmark would just ban amd cpus since they will be better and being better than intel cpus is antisemitic so it is hateful therefore it is hatespeech.

PATCH IT

Attached: 1506977929561.png (200x200, 45K)

if (AMD != AVX512) {
use AVX512;
else
use some obscure instruction set;
}

12 threads? What do you need 8 threads for? Here's your 4 thread CPU LMOA

Attached: SotTR_1080p.png (1328x1222, 53K)

>AVX512
nice housefire

Intel compiler does this exact same shit, it refuses to use SSE and AVX if it's not an Intel CPU.

It says the Xeon is much more recent while showing Q2 '19 for both. Intels own website says Xeon is Q4 '18 and i3 Q1 '19.

>now

Did Intel buy them?

any potential buyer that types something like "i3 8350K vs 2700X" on google will get pic related.

Attached: oh no no non o.png (1664x1208, 479K)

Since forever?

Attached: 1492139297054.jpg (752x548, 282K)

Attached: 1550107261226.png (600x387, 408K)

how new

Attached: 1546433289733.jpg (921x960, 212K)

It largely doesn't.
I just put together a Ryzen 2700 system in the other room, ran a couple encodes on it, yet am happily crunching away on a 50 episode BD series here on my i7 4790. (and I'm thinking of releasing another, smaller encode, which would put it at 100 episodes encoded)
That 2700 is going to need a new cooler to quiet it down (the stock isn't loud but it's still too noisy for my main) and the mobo is severely lacking in SATA ports so I'm going to need to find a RAID card before I could even consider it for my main, and I'm really in no hurry to get it finished. Oh yeah, I'd need another 8GB DDR4 too.
In all likelihood I will never do the swap since moving 8 drives over is a lot of work and I'm really happy with quietness of my main, so I'll probably just end up remoting into the Ryzen to run the occasional large encode while doing most everything else on my i7.

Having more than twice the encoding performance just isn't that big a deal since proofing subs takes longer than the encode.

Saying that multicore is the be all and end all of CPU performance is hilariously overstating the case. If I'm in no hurry to do the swap with PCs I literally have in my house as an encoder, guess what the value is to the vast majority of PC users?

>That 2700 is going to need a new cooler to quiet it down
>I'd need another 8GB DDR4 too.
>the mobo is severely lacking in SATA ports
>I'm going to need to find a RAID card
so you built a ryzen PC, presumably for its superior video encoding performance, but you refuse to actually put in any of the hardware to make it useful for encoding your 8 jiggabytes of anime
>Having more than twice the encoding performance just isn't that big a deal since proofing subs takes longer than the encode.
you would still be saving 200% more time on encoding
>Saying that multicore is the be all and end all of CPU performance is hilariously overstating the case.
>If I'm in no hurry to do the swap with PCs I literally have in my house as an encoder, guess what the value is to the vast majority of PC users
i'm not even sure how your encoder PC swap anecdote is even relevant to multicore performance, it sounds to me that you built a ryzen PC to replace your i7 but gave up when you realized you can't plan a build for shit

>he spends $5k on all his builds

My 4790/GTX 1060 build was $280.
The Ryzen is at $375, but that will come down after I sell the 1050 Ti and 2200G it came with.

The cheapest AM4 motherboard with 8 SATA ports looks to be $154 on its own. See why I don't have one?
I have a spare 212 EVO but it's too old for AM4 compatibility, as is the one attached to my 4790. I'm not keen on spending $30 on a brand new one.

I put together what I find. With the i7 I managed a build with zero compromises. With the Ryzen there are a few.