• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Post your Cinebench R23 Score

@FCG :
I see that you are using the turbo hack, mind you that older microcodes with TSX are less stable, though do turbo higher. You seem to not run undervolted, as you upped the bclk to 105, do check stability in the long term, also to avoid other devices hurting from the bclk, you may use an undervolt efi to gain same performance without bclk raising.
Can you share the efi version you are using, also the microcode revision and are you meltdown protected.
 
MrGenius said:
Ridiculous. :shadedshu:

Anyway...cat's out of the bag now. So they're pretty much screwed. :p
Apparently neither. The system requirements state Intel or AMD 64-bit CPU with SSE3.
Click to expand...
I think it has a fallback with SSE3 for older CPU's. I read on Anandtech it uses AVX though. Just not sure if 128 or 256bit
 
er557 said:
@FCG :
I see that you are using the turbo hack, mind you that older microcodes with TSX are less stable, though do turbo higher. You seem to not run undervolted, as you upped the bclk to 105, do check stability in the long term, also to avoid other devices hurting from the bclk, you may use an undervolt efi to gain same performance without bclk raising.
Can you share the efi version you are using, also the microcode revision and are you meltdown protected.
Click to expand...

I do both: undervolt (CPU0 @ -50mV, CPU1 @ -70mV) *and* increase BCLK.
There are ways to stabilize increased BCLK, which is an even bigger challenge with a dual socket system as QPI is also symmetrically effected.

I do not make use of TSX instruction set. Also, disable VMX if not needed.

I find 0x1F microcode patch level is best for performance... which is what I run.
EDIT: appears TSX (HLE) was nixed 9/3/2014 with release 0x29 for CPUID 0x306f2 (Haswell-E/EP) so even the 0x27 guys aren't "safe"
See here: https://lkml.org/lkml/2014/9/18/218

I have authored my own EFI driver, which can be found here:
https://github.com/freecableguy/v3x4

*Not* Meldown/Spectre "protected."

Trust me. ;) I know all about how the turbo hack works. :)
 
Last edited by a moderator:
Just now realizing you're freecableguy, didn't recognize from the nick, anyways, you should be teaching me, not vice versa...
 
forgot to run CPU-Z alongside CB R20 so here's a screenshot. processor is cooled with an AIO kit.
Capture.PNG
 
Last edited:
ArbitraryAffection said:
I think it has a fallback with SSE3 for older CPU's. I read on Anandtech it uses AVX though. Just not sure if 128 or 256bit
Click to expand...
I see. A quick google search reveals that it probably uses up to AVX-512. Or it can anyway(if they wanted it to).
Intel® said:
Embree Overview

Intel® Embree is a collection of high-performance ray tracing kernels, developed at Intel. The target users of Intel® Embree are graphics application engineers who want to improve the performance of their photo-realistic rendering application by leveraging Embree's performance-optimized ray tracing kernels. The kernels are optimized for the latest Intel® processors with support for SSE, AVX, AVX2, and AVX-512 instructions. Intel® Embree supports runtime code selection to choose the traversal and build algorithms that best matches the instruction set of your CPU. We recommend using Intel® Embree through its API to get the highest benefit from future improvements. Intel® Embree is released as Open Source under the Apache 2.0 license.

Intel® Embree supports applications written with the Intel® SPMD Program Compiler (ISPC, https://ispc.github.io/) by also providing an ISPC interface to the core ray tracing algorithms. This makes it possible to write a renderer in ISPC that automatically vectorizes and leverages SSE, AVX, AVX2, and AVX-512 instructions. ISPC also supports runtime code selection, thus ISPC will select the best code path for your application.
Click to expand...
https://embree.github.io/
 
ryzen 5 1600 @ 4.025ghz, ddr4 3200mhz 16-17-17-17-34-54 duel channel, arctic liquid freezer 240mm.
Capture.PNG


ryzen 5 1600 @ 4.075ghz, ddr4 3200mhz 16-17-17-17-34-54 duel channel, arctic liquid freezer 240mm. this is not my everyday overclock, i just ran it higher so that i could pass one of those r5 2600x's on the chart.

Capture2.PNG
 
Last edited:
There we go... To be honest, I was expecting it to score even less.
2019-03-07 20 21 54.png
 
My score seems a little low...

R20 AIO.jpg


Used an H105 AIO
 
ArbitraryAffection said:
yeah that seems a bit low. What's your AVX offset? I would've expected the 6 real cores to beat the 8 logical cores on the 7700K. Unless this bench just really likes threads lol. My 2700X at stock (3.9-4.0 in this bench) is getting nearly twice your score.
Click to expand...

set at 0
 
IDK then. Sure nothing running in the background etc?

ALso here is my HP Envy x360 with 2500U, plugged in and set to performance mode (12W)
cbr20.jpg


And again for academic purposes, here it is on battery, set to optimal mode.

IMG_20190308_012231.jpg

cbr20 battery.jpg


so it gains around 30% performance from being plugged in. I didnt check the clock speeds during each run but it seems to hit 3 to 3.1 all core for a short time plugged in then drops to 2.7 or so after a while, then probably further down as it heats up. IDK what it was doing here on battery. probably sub 3 ghz though.
 
Last edited:
@T4C Fantasy -- Sorry bub, I ran the test again after a restart and having a gander over my power settings

R20 AIO.jpg


I have no idea what the hell went wrong with the first run. I ran the test TWICE the first time just to be sure there wasnt something wrong
 
EDIT:: closed some things.

Cooling H2O

1552011176031.png
 

Attachments

  • 1552010988594.png
    1552010988594.png
    2.2 MB · Views: 326
Last edited:
T4C Fantasy said:
put your cooling in the post im replying to
Click to expand...

Sure im kinda bored though, I think this weekend ill try to beat the xeon that holds the top.
 
ArbitraryAffection said:
Don't make me get a 2990WX, I'll do it I swear! Just for the cinebench! Oh wait, I can't afford one :C
Click to expand...

iv built one, really cool chips, super large. like the palm of my hand kind of big.
 
Solaris17 said:
iv built one, really cool chips, super large. like the palm of my hand kind of big.
Click to expand...
Next year maybe, i would like to get the top of the line TR 3rd gen. 3990WX or something. Maybe it will be a 48 core :) just to own such a powerful processor would be so cool.
 
You must log in or register to reply here.
Back
Top