OC3D Cinebench R20 Benchmark Thread - Page 8

KingNosser · #71 12-06-19, 05:18 PM

Quote:

Originally Posted by Warchild

change the link to [img] maybe?

I tried that obviously Idk is odd

ty for fix, the link I got didn't have the .jpg but it's handy info for another time

hmmblah · #72 12-06-19, 07:41 PM

Quote:

Originally Posted by KingNosser

I tried that obviously Idk is odd

I fixed your post. What you need to do is open the link you posted, then right click on the image and copy link address. It will have a .jpg on the end of it:

https://i.imgur.com/VJIFlrP.jpg

Then you enclose that in img tags on the forum.

FTLN · #73 22-08-19, 10:58 PM

5421cb, Intel i9-9900K, 5.1Ghz, 2666MHz DDR4, Windows 10, FTLN

Thanos · #74 08-09-19, 03:06 AM

MT
4697cb, Xeon E5 2683 v3, 3.0GHz, 1100MHz DDR4, Windows 10, Thanos

ST
284cb, Xeon E5 2683 v3, 3.1GHz, 1100MHz DDR4, Windows 10, Thanos

FWIW not seeing a difference in performance with or without fixes. If anything slightly better after patching.

FTLN · #75 27-11-19, 07:31 PM

10797cb, 7980XE, 4.7ghz, 3200mhz DDR4, Windows 10, FTLN

RobM · #76 26-12-19, 02:26 PM

Update
3655, Ryzen 5 3600, 3200mhz DDR4. Windows 10, Rob

AlienALX · #77 19-01-20, 05:37 PM

Hmm.

Thought I'd bring this up again as I moaned about it being inconsistent.

I was right, and I'll tell you how I know.

I timed a R15 run on my 14 core. I then timed a run on my 16 core. Sure enough the 16 core was faster. This wasn't just something I dreamed up, it was a thing. A noticeable thing.

Yet, the 16 core scored 200 points lower. I tried this in R20 and even though the 16 core is 200mhz slower the 16 core was still faster, yet, the score was lower.

So I thought O.K this is probably unique to me because I'm using oddball hardware.

Nope.

I watched a Jayz video and he had exactly the same issue. I've been watching his new build, and I think that led me to a TR review.

How can a CPU that renders a bench faster actually score less points? Surely the faster the better right?

So I call shenanigans, and I don't trust this bench further than I could throw it.

Davva2004 · #78 20-01-20, 10:38 PM

MT 7062, ST 495, Ryzen 3900X at stock clocks cooled by a Kraken X42, Gigabyte X570 ITX mobo, 32gb GSkill 3600mhz, 2tb Corsair MP600

Warchild · #79 21-01-20, 08:40 AM

Quote:

Originally Posted by AlienALX

Hmm.

Thought I'd bring this up again as I moaned about it being inconsistent.

I was right, and I'll tell you how I know.

I timed a R15 run on my 14 core. I then timed a run on my 16 core. Sure enough the 16 core was faster. This wasn't just something I dreamed up, it was a thing. A noticeable thing.

Yet, the 16 core scored 200 points lower. I tried this in R20 and even though the 16 core is 200mhz slower the 16 core was still faster, yet, the score was lower.

So I thought O.K this is probably unique to me because I'm using oddball hardware.

Nope.

I watched a Jayz video and he had exactly the same issue. I've been watching his new build, and I think that led me to a TR review.

How can a CPU that renders a bench faster actually score less points? Surely the faster the better right?

So I call shenanigans, and I don't trust this bench further than I could throw it.

I wonder if the weighting is being biased. Wouldnt be the first time we see that. Either that or its a genuine miscalculation. Perhaps their calculations per core are inaccurate leading to this lower score.

Dicehunter · #80 23-01-20, 02:02 AM

My ex 9900K which was overclocked to 4.80GHz vs my new 3950X using PBO in the bios to auto boost clocks, Quite impressed, Didn't do single core as I was just interested to see multi, I've seen people doing an actual overclock and getting well over 10k so I may try it later.