Intel is the king of 8 cores/16 threads

You can buy a Xeon X7560 8 core/16 threads for as low as 3.15 USD.

AMD BTFO!!!!!!!!!!!!!! How can AMD even compete?????

Attached: AMD BTFO.jpg (1163x536, 147K)

Other urls found in this thread:

progforperf.github.io/nehalem.pdf
twitter.com/SFWRedditVideos

>2.26GHz
I can just get Core2 Duo shit out of the dumpster for free and use that $3 on a hamburger.

lol amerifats

>single thread rating: 961
i7 2700k is twice that. Most apps nowadays still use 1-2 cores. More cores won't always equate the best performance for daily browsing.

Attached: doubt.jpg (609x250, 37K)

>2.26GHz
>hyperthreading disabled due to zombieload
>40% IPC performance reduction due to spectre/meltdown mitigations
>intel ME spyware
>expensive socket motherboard, dead platform

Basically, current and past gen Intel CPUs are going to go to zero and no one will be picking them up off the streets because that's what they are--complete garbage.

ahhhh nehalem.... now THAT was a CPU... *sip*

LMAO @ all these people falling for this obvious bait.

"Hey guys did you know ancient enterprise hardware is cheap af?"

>3 usd for cpu
>600 usd for motherboard
That's a good deal

Attached: Screen_Shot_20190508_at_10.37.07.jpg (1648x930, 167K)

*crack* I'm telling ya, those suckers'd overclock 50% or more on a good cooler. *siiiiiiip*

from 2 to 3 Niggerhurtz!

This ancient $3 boomer chip running on DDR3-1066 in a single socket configuration unironically has better RAM latency than Poozen 3.

progforperf.github.io/nehalem.pdf
>The latency to access the local memory is, approximately, 65 nano-seconds.

AyyMD is finished.

Attached: 1560280779458.jpg (2000x1125, 207K)

e5-2687w would be 8 core 16 thread with ipc between sandy and ivy bridge. also quad channel ddr 1600

STOP

Attached: 1544223628835.png (653x726, 35K)

You can get a 16-core Opteron for $10, I'd rather go with that

Bandwidth and latency are directly related.
You'd have none this if you got an education as opposed to jacking off and shitposting with tendies slewn around your room.
> I was made aware of this computer spec that influences performance while watching my favorite e-celeb on youtube
> I think I actually know what design trade-offs are made in designing a processor
Sorry dumbfag, you're just an idiot.

DDD3 latency < DDR4 latency < DDR5 latency
Bandwidth goes up.. latency goes up

9900k is dual channel ddr4 and still has much lower latency lmao

no ddr4 @ speed has less latency than ddr3.
So, what's you're point beyond deflecting from reality. Ultimately, all that matters to you (idiot) who knows nothing about compute architecture and micro-architectural specs is whether it performance better.. namely in vidya. so, what exactly are you mumbling yet again? Some stupid shit that fell out of your pea-brain? ..
> stop

Cope

Yes you are.. for you lack of formal education and knowing a single thing about what you speak of beyond some slide-deck or pinhead on youtube. For this, you are a brainlet incapable of refuting anything stated beyond verging into a deflection or posting a cringey reply.

> Tendies are ready dear
That's for you

Attached: roast.jpg (1440x1299, 100K)

Or some 香肉火鍋 , whatever floats your boat

The point was that both the 9900k and Zen 2 utilize dual channel DDR4 and the former has no regression but even an improvement in latency compared to previous generations. This disrupts the claim in the post I replied to regardless of your mudslinging.

>香肉
What is this, 新年?

Your point was uninformed, misguided, misdirected and based on a shitty myopic view.

L3 on Ryzen is double what it is on even threadripper ringing in at 72MB. Find an Intel processor with that much L3. Such an L3 causes latency to go up significantly. It also results in greater overall performance such that the Zen 2 smokes the 9900k.

What matters is how it performs during real-world data flows.. Hint : It kicks the shit out of 9900k. RAM latency is almost double yet you have 72MB of L3 that it will constantly be hitting vs ram. So, during a legit real-world flow, it will outperform a 9900k.

Something you'd know if you had any understanding of processor design. However, you're a fucking idiot who thinks their idiotic snarky posts of retard talking points has some point to it. They don't. An educated person sees through your shitpost and sees your rotting brain. Stop trying to get a win here. You're embarrassing yourself.... idiot

OH NO NO NO NO AMDBROS WE GOT TOO COCKY

>shifts the goalpost to copecache size
>'legit real-world flows' are only those compatible with a bunch of disjointed L3 slices with garbage access times and interconnect congestion when there's an L3 miss on the local one
it just gets better and better

and it's 64mb L3 btw, and some xeons have even bigger L3s and still have superior RAM latency so you're still wrong :^)

>Xeon X7560 8 core/16 threads
you mean 8core/8threads?

No AVX or AES support.

Good luck on the mobo search