CPU-Z Thread

Post your cpuz bench results. Clocks and voltage would be interesting as well.

My results are for a Ryzen 7 1700 @ 3.7GHz & 1.15V with stock cooler.

Attached: cpuz.png (403x402, 19K)

Other urls found in this thread:

valid.x86.fr/bench/1
valid.x86.fr/0xw6rw
en.wikipedia.org/wiki/Amdahl's_law
ibm.com/blogs/psirt/potential-impact-processors-power-family/
valid.x86.fr/nruvxb
valid.x86.fr/b0644k
twitter.com/NSFWRedditVideo

stock

Attached: 2019-05-22 12_10_30.png (400x394, 15K)

I'm wondering, since you have intel. Did your scores drop after latest round of Windows updates? Post mitigation I mean.

Btw here is a link if you want to compare, it's a mix of overclocked and stock results for each CPU so doesn't mean much if you are getting lower on stock clocks.

valid.x86.fr/bench/1

here is a test from early december.

Doesn't seem massive.

I am fully up to date with the latest 1903 W10 build just released today.

Attached: 2018-12-03 00_52_55.png (395x376, 14K)

Security doesn't matter?

huh?

It doesn't have SMT since it's the 9700k, so it's by default invulnerable to certain hardware issues of SMT.

It's fully patched and up to date with any available fixes, specter, meltdown, etc.

Yeah doesn't seem all that massive for desktop/gaming use.

I heard that full mitigation requires disabling hyper threading, but it would seem unlikely to be exploited unless he was targeted specifically, which wouldn't really happen as that level of an attack would cost money. More of issue with servers.

>disabling hyper threading
which isn't possible on my CPU anyway, since it's an 8 core CPU without hyperthreading.

I didn't realize that 9700k didn't have hyperthreading, interesting.

Here's my server. I'm unsure if running the benchmark over remote desktop changes the score much.

Attached: ServerCPUZ.png (807x769, 27K)

Yeah, it's honestly why I bought it, I had a 5820k previously that the x99 board shit the bed, so I sold the 5820k and picked up a 9700k. My time with the 5820k convinced me Hyperthreading was garbage in most of my use-cases, and real threads would serve me better. Going from 6 core 12 thread at 4.2Ghz to 8 cores at 4.6Ghz was a nice upgrade for what i'm doing.

Also here is another test I just did with less background CPU usage.

Attached: 2019-05-22 15_39_44.png (445x426, 17K)

Not bad for something I got for $80, I'll probably upgrade to an i5 9400 in the near future.

Attached: cpuz_ELJMSLJew7.png (392x323, 10K)

I have a 5 year old processor

Attached: i7-4770K.jpg (404x406, 59K)

My machine is pretty mediocre desu. Got it at launch in 2015, so not too bad for a nearly 5 year old chip.

Attached: Untitled.png (611x605, 85K)

4.8Ghz dynamic voltage, +75mV
>Did your scores drop after latest round of Windows updates?
no

Attached: c5dc8f8a.png (403x402, 15K)

2700X @ 4.2 on all cores 1.35V. My memory isn't nearly as tuned as it could be, probably.

Attached: cpuz.png (386x265, 12K)

Attached: Capture.png (480x505, 20K)

LMAOOOO

Attached: KEK.png (406x379, 14K)

valid.x86.fr/0xw6rw

>valid.x86.fr/0xw6rw

I think 1.439V is dangerous long term, chance of damage to your CPU.

Ryzen 3 2200g
3.65 Ghz according to Ryzen Master
stock voltages

Attached: 2200g.png (408x399, 23K)

It's interesting to note that single core performance is reaching limits in last few years. Most CPUs are in the 400-500+ range and it's not going much past that in most tasks/benchmarks.

Sucks we won't see much improvement in the comings years either, unless something major happens.

>certain hardware issues of SMT
and by SMT you mean HT.

SMT is the generic term. HT is intel specific.

Power9 from IBM uses SMT-4 or SMT-8 depending on the CPU.

Thank God. Single thread should die. Bring on the multicore revolution Intel fucking promised me in 2007.

>QUAD CORE
>HARD CORE
>BUY CORE 2 EXTREME GOY

Attached: goy2extreme.jpg (225x225, 8K)

SMT doesn't have h/w issues.
HT otoh....

Attached: Screenshot 2019-05-22 19.39.30.png (407x404, 18K)

nothing is going to happen in terms of per core performance and clock frequency.
all r&d has steered towards accelerators.

You're full of shit.

HT has specific issues regarding MDS that power9 doesn't have.

Specter and Meltdown however utilize a hardware deficiency in how multithreading works.

SMT is inherently flawed when talking about virtualized systems.

Problem is that it's very hard to make everything multithread and there are diminishing returns after a certain amount of cores/threads.

en.wikipedia.org/wiki/Amdahl's_law

Ryzen 7 2700 at 3.35 GHz, I guess it makes sense that the single core score is a bit lower here

Attached: Ryzen 7 2700 bench.png (904x908, 205K)

Isn't it already known that accelerators suffer the same brick wall as multicore.

So we are pretty much fucked in the next few decades, unless something revolutionary comes about?

It sucks on one hand as shit doesn't get much faster, but on the other hand you can buy a computer to last your 5-7+ years.

dudes who bought a six-core LGA1366 chip are still laughing. Wish I would've invested in one back then, shit would still be just fine.

This slav poorfag hates all of you.

Attached: Annotation 2019-05-23 014003.png (402x400, 18K)

I invested in a 6 core LGA 2011-v3 (x99)

The motherboard shit itself and a comparable replacement was $400+.

Ended up dropping it for Z390 and a 9700k.
At least with consumer chipsets the motherboards are way fucking cheaper.

You should overclock, single core will improve. If you keep voltages low enough (assuming you got a decent unit), you will not have significant temperature or energy increase and just use default cooler.

Also 2700 boosts to 4.1GHz, but I don't think it does it on all cores, when you overclock it lets you run all cores at set frequency so you get better results. If I'm not mistaken.

You're correct about the boost clocks, and I have a decent after market cooler on there. However, whenever I up the clocks and the voltage, my PC goes blank and the LED on my GPU (AORUS RX580) starts blinking. I have an MSI B450 Tomahawk, do you think that the mobo can't handle the voltage increase?

4.9 GHz Spectre and Meltdown patches applied, MDS Windows update applied. Intel's singlethreaded performance really has not improved over the last 5 years, and we all know how much they're struggling to increase core counts.

Attached: 7700k.jpg (402x403, 64K)

I'm full of shit?
You just erased 60 years of perfect SMT implementations, just because your favorite company did a huge deliberate oopsie in order to squeeze a bit of performance out of their shitty reheated architecture.
HT is intel's version of SMT, and it's fucked up beyond recognition.
It's blatantly false to call SMT flawed, when everyone else got it right.
That's like calling RSA flawed, because you can't catch all the security holes in your program, or calling C an insecure language because your stupid ass cannot keep your buffers/input safe.

Oh yeah that sounds like a common issue, unfortunately a lot of b350/b450 motherboards are shit, have bugs or you just got super unlucky with the CPU. Make sure you have latest bios and read if there are issues for your specific motherboard. VRM cooling issue is another possibility.

How much are you increasing? Try very small increments and see where it starts fucking up.

>when everyone else got it right.
Yeah, IBM totally just released patches cause they got it right and didn't actually need to.

ibm.com/blogs/psirt/potential-impact-processors-power-family/

shut the fuck up retarded shills

I can live with this.

Attached: cpu-z-bench.png (407x403, 31K)

Here we are at 1.3v @ 3.8Ghz, which I figured was a conservative clock speed at a high voltage, and I am stressing it now in CPU-z so I guess I can fiddle with it a bit more.

Attached: R7 2700 3.8 Ghz.png (890x901, 203K)

it seems to me that you are totally ignorant on the topic.
The article that you just posted, doesn't patch any SMT "vulnerabilities, you fucking faggot.
IBM has unverified security flaws on their speculation. Nobody found a vulnerability, they patched it beforehand to avoid any chance there's a vulnerability.
The other patches had to with branch prediction.

The latest security holes that where found on Intel processors, along with several other attacks the last 2 years, have to do with HT and the way intel manages data on the DRAM.

let's start again. there's SMT and there's intel's HT shitty implementation.

by your logic, Tomasulo is flawed, because your shitty company cannot design a robust arch.

gotta support Jow Forums's only skill.... screenshot posting.

Speculative execution is how most implementations of SMT are done, including intel, AMD, and ARM, and IBM Power ISA.

Spectre, meltdown, and foreshadow are ALL examples of vulnerabilities in speculative execution, severely limiting SMT performance when patched under certain workloads.

Ryzen 7 2700 running at 4.0 Giggle hertz, 1.3 volts. Not sure if it's going to be stable for stuff like video rendering or not, but it passed the stress test in Ryzen master and this CPU Z bench.

Attached: 4.0 Ghz 1.3volts.png (905x905, 205K)

This scaling and font rendering looka ugly as all fuck

you are THAT incompetent.
SMT has nothing to do with speculation and vice versa.
SMT doesn't define the way your CPU does branches, renames registers, unrolls loops, handles superblocks and hyperblocks, does OoO or in-order execution, shares data between cores(concistency and coherency)
You are an uneducated piece of shit, wasting my time. you didn't even take the take to at least verify your full of shit posts with the other kind of shit that wikipedia has.
your whole post summaries to this
>SMT is bad because Intel's CPU design has flaws
the next step is to blame the security holes on the sand that you get the silicon from.
>hurr durr, all vulnerable CPUs are made of silicon. THAT'S IT. I am switching to GaAs.
GaAs the kikes.

Attached: 1544283003778.jpg (226x170, 13K)

SMT is bad because it's been shown to be vulnerable across every major architecture that currently utilizes it.

good cpu

Attached: tach.png (612x607, 228K)

don't hate me slavman

Attached: helpmin.png (251x211, 72K)

kek, other ryzenfags don't know how to oc

Attached: new cpu z.png (405x406, 18K)

Not you, user.

nohomo

Attached: 13n22x[1].jpg (500x461, 18K)

2600X letting PBO do all the work with a -0.1V offset.

Attached: UzjEi9S[1].png (403x402, 15K)

and yet still gets btfo by my 1600x manual OC, what a scrub

ty based slavman

valid.x86.fr/nruvxb

Attached: cpuz.png (407x407, 14K)

Single thread performance seems a little low

Attached: cpu-z2.png (1946x1734, 560K)

valid.x86.fr/b0644k

Not bad for a $50 placeholder

Attached: file.png (399x353, 30K)

@4.3Ghz now
Was 4.5 before intel sourmilkā„¢ aging

Attached: inturd.png (403x403, 16K)

i need a new tablet

Attached: cpuz.png (810x402, 36K)

Superior system coming through

Attached: 1537293031613.jpg (403x402, 59K)

This is my Acer AspireOne from 2012. It was given to me for free a few years ago because it was too slow. I upgraded its RAM from 2 GB to 8 GB using a spare SODIMM, and installed Windows 8.1 because 10 was unusably slow on it.

Nah mang it'll be fine. You might get better performance on Windows 8.1 though.

Attached: best_cpu.png (417x414, 17K)

is yours OC? i havent done it to mine yet bc im waiting on some fans

Attached: stock 9600k.png (403x402, 15K)

What cooling/memory do you have paired with your 2700X? I can never manage to get that high of a single core score.

Attached: chika.jpg (992x1403, 205K)

the cpu is stock with a noctua d14. memory is cl16 3200 so the timings aren't that great.

>OC
Stock speed is good enough.

Attached: cpuz.png (407x401, 14K)

x48 on all cores, could get it to x49 on all but don't like the power draw it produces. x49 gets very close to 600 single core score iirc.

i9 9900k @ 5.0GHz 1.32v

Attached: bench.jpg (400x400, 56K)

>AMD/Poozen

Attached: (((((mascots))))).png (1600x1223, 1.1M)

i5 2500K, it's old but it gets the job done, waiting for Zen 2 to do full platform upgrade.

Attached: i52500kcpuzbenchmark.png (814x400, 56K)

its PBO shitting voltages up. i have a voltage offset now.

stock

Attached: bench.png (403x402, 16K)

>valid.x86.fr/0xw6rw

Attached: Selection_074.png (1318x1256, 327K)

are these bdie? how much for that memory?

Hynix AFR. I got them for 130$ per 2x 8GB, so 260$ in total. this is what I run daily, tFAW at 30 and tRFC at 480

Attached: afr.png (945x787, 415K)

nice.

Disable HT and do 5.2GHz

You got good ram for such timings on four sticks.
My TridentZ 2x16GB 3200C14 B-Die only did 3000C14. sucks.

@ 4GHz all core

Attached: 1950x.png (405x402, 18K)

T-Topology board.

I take objection to this post. Mine is not Ryzen, but C-60 APU. While shitty and weak, it's way better than the old Atom CPUs it competed with.

>While shitty and weak, it's way better than the old Atom CPUs it competed with


you can do it c-60 i believe in you

Attached: jpg.png (88x68, 11K)

Yikes looks like I lost the silicon lottery

Attached: received_674575022981162.png (404x401, 24K)

Or you just compared your stock score to an average that includes scores that are inflated by industrial coolers.

Based and Redpilled

Attached: cpuz.jpg (400x399, 38K)

here is the worlds worst 4 core with ht

Attached: Capture.png (407x404, 16K)

Asus C60 ITX board with 4GB RAM here. Anything past XP is painful. It works as a sandbox/plug-random-stuff-into device however

Mine with 8 GB RAM and C-60 in a netbook came with 7 preinstalled but I'm running 8.1 right now. Seems to be the lightest and smoothest Windows for low-end systems that's newer than XP.

Marvel at the power of a 1st gen mobile i5

Attached: Capture.png (403x402, 12K)

2700x pe1, pbo, uv, tuned ram,

Attached: unknown (1).png (509x505, 89K)

and i thought my q 720 was bad...

8700k @4.9GHz on latest windows 10
Offset vcore fluctuates between 1.232v and 1.248v during the benchmark

Attached: win1903.png (403x402, 20K)

show memory timings

i5 6500 @ 4.5 1.31v

Attached: CPU Z Bench.png (403x402, 10K)