Is this accurate? a processor from 2011 clocked at 4.5 GHz is beating the per-core performance of a Threadripper 1950x?

Is this accurate? a processor from 2011 clocked at 4.5 GHz is beating the per-core performance of a Threadripper 1950x?

Attached: oc-i7-2600k-45ghz.png (417x413, 42K)

Other urls found in this thread:

is2.Jow
youtube.com/watch?v=XJNQ44GYChc
twitter.com/AnonBabble

yes. if you buy a 32 thread processor to use 1 thread you are fucking retarded

this

Yeah. The AMD not clocking as high should account for most of the difference.

Threadrippers are about multicore performance. Compare to a Xeon and you'll see the same.

if you ignore the fact that tr has 4x the amount of cores and threads you have, yes.

this is some pretty weak bait. Apply yourself.

Yes. Old tech is better.

>amd shill to the rescue
we are talking about single core: op already said that

Attached: thinking emoji.png (469x495, 24K)

Can't match a stock 2700X though, despite being overclocked.

Attached: Sad!.png (403x402, 18K)

a $1500 i7-6950x which launched in 2016 also gets beat in single performance to that same $300 2011 i7
>b-but amd have poor single performance goy

Attached: Untitled.png (419x431, 71K)

if we are talking about single core then why are we comparing it with threadripper?

Seems so. My Ryzen 1700 does 439 singlethread at 3.75 GHz, so you need 750 MHz more to archieve the same performance.

The 1700 is a rendering beast. It gets 80-90% of the performance at 100 W. Also it runs at these temps on full load in a fucking HTPC case with only one exhaust.

Attached: 1700.jpg (2351x1730, 657K)

That Threadripper score seems wrong. Pretty much all Ryzen's have the same single core score at the same clockspeed. Here's my Ryzen 5 at 4GHz. Theoretically any Ryzen CPU with a single core that hits 4GHz should get around 460, and the TR1950X has a single core boost up to 4-4.2GHz.

Attached: cpuz.png (396x396, 17K)

Clock? This one seems pretty high for 1st gen Ryzen.

All cores at 4GHz.

Not bad, how much voltage? The 2700X does 490 at 4.2 GHz

>per core
single core
>is this accurate
I don't know but probably, the 1950x has four times the cores and of course if you leave it at 3.5ghz and overclock the other processor to 4.5 it will beat it.

>(overclocked) 95 watt quad core beats 180 watt 16 core

>impressive

Attached: 1529146993817.png (1210x1190, 116K)

Slightly more than 1.4V, I got unlucky. I bought the 1600X on release when the rumor was that the X models were much better binned, unfortunately it's impossible to keep my 1600X at 4GHz at 1.4V stable. Maybe with a high end mobo but mine isn't great.

Not him but I got a pretty damn good chip. 4.2ghz at 1.31v on my 2700x. Single core is a little off sometimes though as I need to turn some stuff off in the background to properly benchmark.

Attached: unnamed.jpg (593x592, 197K)

I only tried 4 GHz once but it crashed before I could even run a benchmark. 3.75 work very well. Its my efficiency build, while the 2700X rig is my performance build.
Yes, you need to set it to realtime in task manager, or freshly boot and close all backup programs.

I ran 4.25 at 1.425 V but the chip can run less, I first figure out the memory OC.

I had 3600 CL14 working, but CR2, now I get almost the same score with 3466 CR1. Time to improve all timings and see how it affects latency.

Attached: ram.jpg (829x797, 104K)

Fortunately even with the high voltage it's not too bad, my CPU hits about 75 max in stress tests with my decent air cooler. VRMs were shit so I had to buy some 140mm fans to blow air across the mobo.

Yeah a fresh boot and setting to real time gets me around 488-491 though it varies a lot due to wangblows having a bunch of shit running

I assume I could get higher scores if I had faster memory, I only have 3200 CL16 which I should mess with the timings a little. it's not B-die though as it was hard to find 32gb of high frequency B-die last July when I bought it.

The Crosshair is a dream when it comes to this, 50-60° VRMs under load, I guess they prepared for the 12core part I get next year. CPU 60-70 liquid metal, 3x ML120 H150i.
~490 is the average exactly what a 5820K at 4.4 GHz gets
Fast memory doesn't affect CPUz, but it affects Cinebench.

Why do you need 32 GB? I want to buy another kit too, but only for enjoyment. You have to sell your soul nowadays for good RAM. I got my 4266 kit for 215 € but cant find another.

Luckily I bought my ram right before everything completely went to shit but there was still shortages at the time, really nobody had above 3200 B-die 32gb under 500 dollars. I need 32gb because I have tons of Ram heavy programs constantly running like photoshop, blender, autodesk suite, Substance ___, Quixel, etc. I haven't completely used it all yet but I've gotten pretty close at 26gb.

I did remember just now that I was able to get a 498 at one point when I closed nearly everything running in the background. I wonder if I can do that again.

Single core performance is still relevant today. Just think of all those classic single core games you play (lots of people play them still you know). In a perfect world you would play them @ butter smooth fps, no jerky/studdering shit. In these modern times you'd think it would be childs play for modem hardware to render say UT 99 w/everything maxed out @ high res. Sadly this ain't so. FX Proved this to be true. Poor chip design? Yeah. Alas the good old Phenom II can and does deliver the performance single core games need while also handling multi core tasks with ease. Granted I don't know how Ryzen handles old games but if it can't deliver on what I need then what good is it?

If it's "classic" any 2018 CPU should handle it perfectly.

>25 year old boomer who has no idea what he's talking about
Piledriver (FX 8350) should have the same or more single core performance as a Phenom II and Ryzen is drastically better than both. Why on Earth would you think that Ryzen would do worse in single core games?

are you retarded? even the really shitty bulldozer cpus were an improvement over phenom

just fuck off with your inane ramblings
this is why we cant have nice things

>Granted I don't know how Ryzen handles old games
Great, I play mostly older games with a 1200 and I haven't encountered any issues even in emulation

You'd need to have some major form of mental retardation to believe that even the lowest-clocked Ryzen on the market doesn't have significantly better single-threaded performance than a Phenom II.

>are you retarded? even the really shitty bulldozer cpus were an improvement over phenom
Bulldozer was a flat IPC regression from the Phenom II chips. Piledriver brought them up about par, though there are still workloads where a hexa-core Phenom II can outperform an octa-core Piledriver, clock for clock.

>Bulldozer was a flat IPC regression from the Phenom II chips. Piledriver brought them up about par, though there are still workloads where a hexa-core Phenom II can outperform an octa-core Piledriver, clock for clock.
yikes

A processor from 2011 that achieves its high level of performance cutting corners where security is concerned. It's only just been coming to light just how much of a price was paid for Intel's extra performance - meltdown as a big example, and now the context switching thing that's just been in the news. Once all the flaws are found and patched, the performance will probably end up being less.

based intel skipping all the nerd BS to give us gamers what really matters. #respect

Attached: Intel anal beads.jpg (1212x909, 218K)

>In these modern times you'd think it would be childs play for modem hardware to render say UT 99 w/everything maxed out @ high res. Sadly this ain't so.

because the software can't handle clock gating you fuckwit. go ahead and try to play it on coffee lake with zero workarounds. it's got nothing to do with single core performance you absolute cretin.

Yes, because guess what, threads/cores/hyper threads, whatever you wanna call them, only matter if you actually use them.

>FX
>modern hardware

Ryzen is on par with a Haswell i5, depending on the clock. Unless the Haswell runs on 5 GHz, but most couldn't do that.

I think its a myth, it only accounts to shittier parts like the 1400 and 1700, both only do 3.2-3.4 on multiple cores. Ryzen needs manual tweaking, but it isn't as hard. You can run any Ryzen at least at 3.8 GHz at a moderate voltage.

>mfw I'm going to use my i7-2700K for more than a decade and it's barely going to be outdated

Attached: F990C4DECC266E8AF5CA3A73C1E913A7.jpg (800x600, 102K)

Depends on the use. Even a C2D is still pretty usable for browsing. Of course you can't game or run editing software on it.

truth, I use 1700, can't wait for 4700, if my motherboard survives.

It matches most modern Intel processors in single core performance too.

This thread is SO fucking retarded

Good luck getting a phenom to 5Ghz retard

is2.Jow Forums.org/g/1529174529749.png
are you retarded? Ut99 is easily done 500+ fps in a FX chip, almost any of them...csgo
youtube.com/watch?v=XJNQ44GYChc

You overclocked it by 32% and it only beats threadripper by 8% in single core.
Is this supposed to make us praise Intel?

Attached: 1523452147562.jpg (958x958, 62K)

>Pretty much all Ryzen's have the same single core score at the same clockspeed
First gen yeah

Letting XFR and PB do it's thing. CPU boosts to 4.15Ghz on all cores with Vcore locked at 1.275v, on the stock cooler.

Still waiting for my AM4 mounting kit from coolermaster so I can put my 212 on it and see how well XFR can handle it with increased thermal headroom.

Attached: cpuz_2018-06-17_08-24-34.png (403x402, 17K)

i never got why the 6950x was so pricey

Intel

because it was released in a time when AMD didn't have anything on the HEDT market and everything from them performed like dogshit
skylake-x would actually top out at 10 cores too if threadripper hadn't happened

seems like it has the same IPC as that 7700k

/thread
Intel shills are getting desperate

>truth, I use 1700, can't wait for 4700, if my motherboard survives.

4700? Don't you mean 3700?

ok

Attached: ejfg8gmmsn111.png (961x952, 417K)

Cpuz seems to have the default ryzen scores quite low these days. My stock 1700 beats their listed 1700x, probably recorded with 2400 ram and original bios.

Attached: download.jpg (250x202, 6K)

Security doesn't matter.

shut up goy

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

YAS QUEEN

Attached: Untitled.png (428x425, 33K)

because op is retarded
>check this picture that has anything to add to the discussion
my last intelaviv cpu was a 2380P

>single-threaded performance
nobody gives a fuck you faggot
and if you do, there's a place for you

A FUCKING CHILLER

>is this accurate
>no, cpu-z is shit
>discussion devolves into even more shit
There you go.

>The Crosshair is a dream when it comes to this

Given the level fo support the C6H has I anticipate my C6H to not only support next gen ryzen it will also most likely support the SoC features like whatever the next iteration of precision boost is. My plan is to go from first gen ryzen to last AM4 ryzen on the same motherboard which is a factor into why I went all out and bought the crosshair in the first place as asus doesn't fuck about with their high end AMD cpu stuff.

Remember, a socket change a year keeps the goyim in fear!