Cinebench R20 released

[TheProfosist]

Get it while it's hot since it'll only be on the store soon

[Bitter]

5 hours ago, yolosnail said:

I'm genuinely impressed that my 10 year old Mac Pro can just about keep up with my laptop

 

  Hide contents

  Hide contents

8 cores is 8 cores... right

And yes, that is indeed a GT 120

 

This baffles me. My low power i5 4590S at 3.1Ghz base and all core 3.4Ghz turbo with DDR3 at 1664Mhz with tight 10-10-10 timings is scoring a low 1200?!? Does the score scale on a heck of a curve or what? I'll paste in my screen shots after I run it again to make sure I'm not absolutely insane.

 

Had a weird Windows 7 hang, the CPU graphics driver is meh. Scored 1289 and 357 for multi and single core in the i5. I'm running again to try to get a screenshot because that seems higher than it should be.

[skunkmunkey]

Anyone had an issue where when opening the folder location via the task manager that the option to open the location is greyed out? Not a huge issue as I just grabbed it from TPU anyway but did seem a little odd. I'm the admin account and opened both Cinebench and task manager as admin.

 

Anyways, I'm glad it now seems to show the correct OS, always found R15 showed the wrong OS however it still seems to only show the stock CPU speed, my 8600k is OC to 5GHz on all cores but R20 still shows it as 3.6GHz.

 

The i5-8600k got;

CPU - 2844

CPU (Single Core) - 501

 

Also ran it on my Surface Pro 4, with an i7-6650U (2C/4T) running stock at 2.21GHz it managed to scrape 763 on the multicore and 337 on the single. Quite impressed with the single core score on the tablet, it got 2/3 of the score despite running at half the speed of the newer, overclocked desktop processor.

 

[kiska3]

I am sure someone will DMCA IA

 

Enjoy

Spoiler

https://web.archive.org/web/20190308020449/http://us2-dl.techpowerup.com/files/_VvRIIogMZV-6MsUi0cMNw/1552011276/CINEBENCH_R20.0.zip

 

[Bitter]

43 minutes ago, skunkmunkey said:

Anyone had an issue where when opening the folder location via the task manager that the option to open the location is greyed out? Not a huge issue as I just grabbed it from TPU anyway but did seem a little odd. I'm the admin account and opened both Cinebench and task manager as admin.

Yes, I had that too so I found it under services and was able to right click on that and do the same thing.

[skunkmunkey]

1 minute ago, Bitter said:

Yes, I had that too so I found it under services and was able to right click on that and do the same thing.

I didn't think of that, good thinking Batman

[Tamesh16]

i would get 650 -700 for my 7700 hq and now i got 1693

[skunkmunkey]

Is anyone else finding their scores on R20 are very consistent? I've just got the same 2844 score five times in a row yet on R15 my best score was 1185 but when running it repeatedly it would fluctuate by up to about 100 points or so over the course of a few runs.

[Tamesh16]

14 minutes ago, skunkmunkey said:

Is anyone else finding their scores on R20 are very consistent? I've just got the same 2844 score five times in a row yet on R15 my best score was 1185 but when running it repeatedly it would fluctuate by up to about 100 points or so over the course of a few runs.

yeah mine is pretty consistent, however the single core seems a bit weird, a 6600hq at  base clock 2.6ghz beat my 7700hq with base clock 2.81 ghz, steady at 3.55 ghz on single core load 

[leadeater]

7 hours ago, skunkmunkey said:

Is anyone else finding their scores on R20 are very consistent? I've just got the same 2844 score five times in a row yet on R15 my best score was 1185 but when running it repeatedly it would fluctuate by up to about 100 points or so over the course of a few runs.

Should be the case, with a harder and longer test reduced variance in clocks and temperatures flattens out the results. Like R15, once we get to CPUs too fast for R20 the same thing will happen again. 

[randomhkkid]

17 hours ago, Bitter said:

This baffles me. My low power i5 4590S at 3.1Ghz base and all core 3.4Ghz turbo with DDR3 at 1664Mhz with tight 10-10-10 timings is scoring a low 1200?!? Does the score scale on a heck of a curve or what? I'll paste in my screen shots after I run it again to make sure I'm not absolutely insane.

 

Had a weird Windows 7 hang, the CPU graphics driver is meh. Scored 1289 and 357 for multi and single core in the i5. I'm running again to try to get a screenshot because that seems higher than it should be.

 

Scores seem about right to me. My 4690k at 4.6Ghz and 2400mhz memory scores 1719 and 436. Linear scaling puts your score at 1271 and 322.

 

[Iskor12]

Ryzen 5 2600 @ 4Ghz and 2666Mhz ram

Test was ran while running teamviewer remote connection.

 

[jagdtigger]

@LinusTech

 

What will you do? Keep using R15 or you also get it from outside the store?

[porina]

Just did some ram scaling tests... and in short, it doesn't look like it is very ram sensitive. Tests below were run at least 3 times in each configuration, with the best score used. CPUs were at stock settings.

 

8086k dual channel 3000 vs dual channel 2133 (same ram), practically identical at 3933 and 3934 respectively.

 

2600 I did three conditions:
dual channel 3000: 2944
dual channel 2133: 2919
single channel 3000: 2921

 

Basically either of the limited states were 0.8% slower, but I'm not going to read significance in that 2 point different between them.

 

So behaviour looks near enough same as R15. Ram bandwidth isn't significant for general running, unless you're doing it competitively where a few points could make the difference.

[skunkmunkey]

10 hours ago, porina said:

Just did some ram scaling tests... and in short, it doesn't look like it is very ram sensitive.

 

Explains why when I was playing around with my ram overclocks that it wasn't making any difference

[skunkmunkey]

Do like it when it's a brand new benchmark on HWBot, don't think I'm going to be in these positions for long haha

 

[darknessblade]

did anybody get the standalone version before they took it down, and is able to share it?

[GoodBytes]

2 hours ago, darknessblade said:

did anybody get the standalone version before they took it down, and is able to share it?

Just go at the Store and get it. Takes 2 sec.

https://www.microsoft.com/store/productId/9PGZKJC81Q7J

And as mentioned before, redistribution of software without explicit permission from the creator is illegal, and so not permitted on this forum. Such action may result in a ban.

[LAwLz]

10 hours ago, GoodBytes said:

Just go at the Store and get it. Takes 2 sec.

https://www.microsoft.com/store/productId/9PGZKJC81Q7J

Maybe he can not access the store for some reason such as region blocks, firewall settings out of his control, limited or no access to Internet on the computer he is going to run it on, etc.

Maybe he is using Windows 7 or 8 and not 10?

[GoodBytes]

4 minutes ago, LAwLz said:

Maybe he can not access the store for some reason such as region blocks, firewall settings out of his control, limited or no access to Internet on the computer he is going to run it on, etc.

The store is avail for all regions. He has control on his firewall. he has access to the internet as obviously he is accessing this forum.

If Maxon wants to region restrict their app, then that is their rights. They do the same in their server by blocking regional IPs.

 

Quote

Maybe he is using Windows 7 or 8 and not 10?

He can use the older versions of Cinebench.

[LAwLz]

5 minutes ago, GoodBytes said:

The store is avail for all regions.

The store is, but the content of the store is not the same in all regions (just like Netflix might be available in the US and EU, but they contain different content).

I am not sure if Cinebench is one of the differences, but it has been an issue in the past like with Sea of Thieves which could not be bought in Iceland, Uruguay, Slovenia, Belarus, and many other countries through the Microsoft store.

 

12 minutes ago, GoodBytes said:

He has control on his firewall.

Not necessarily. Sure he might have it over the software one on his computer, but he might be at a place where they got a network firewall between him and the Internet.

 

14 minutes ago, GoodBytes said:

he has access to the internet as obviously he is accessing this forum.

He might not have made the post in this thread from the computer he intents to run Cinebench on.

It is possible for people to have one machine that has access to the Internet, and another one which does not.

 

14 minutes ago, GoodBytes said:

He can use the older versions of Cinebench. 

Or, with the stand-alone installer, he can run the newest version of Cinebench.

There is not technical reason for why Cinebench R20 has to be ran on Windows 10.

 

 

My point is that there are multiple reasons why he may not be able to just "take 2 seconds" and download it from the Microsoft Store.

[AlTech]

23 hours ago, GoodBytes said:

 

23 hours ago, GoodBytes said:

And as mentioned before, redistribution of software without explicit permission from the creator is illegal

Well....Yes and No. Maxon and other creators have the legal right to sue if it is redistributed without their consent.

 

Most creators don't take legal action against legitimate websites that redistribute a free program because it actually benefits them in a bunch of ways which I won't go into here. Examples include MSI Afterburner, CPU-Z, RTSS etc.

 

In addition to this, quite a few countries have stated it is totally legal to make a backup copy of software and video games including the United States of America.

 

If the forum doesn't want to be sued then I understand but saying it's flatout illegal is not correct.

 

23 hours ago, GoodBytes said:

and so not permitted on this forum. Such action may result in a ban.

Threatening a ban over this seems quite overkill considering the above and considering the fact that posting a copyrighted image or copyrighted video on the forum isn't treated the same way.

 

To give you an analogy of how overkill this feels, it's like giving someone life in prison for borrowing a book without using their library card.

[Princess Luna]

13 hours ago, LAwLz said:

My point is that there are multiple reasons why he may not be able to just "take 2 seconds" and download it from the Microsoft Store.

I have deleted the Windows Store from all my Windows 10 machines, not only that I have never logged in with a "Microsoft account" I still only use local ADM accounts.

 

Up to this date that never mattered as I never wanted something that was only accessible through the Windows Store.

 

I'm not sure how far GoodBytes is exaggerating on this matter or not, I'm sure the fact there was a 'taken down request' on Guru3D had a big impact on this, however I do feel lucky enough for the stand alone software been floating around by other sources if I'm honest, otherwise I wouldn't have got it even though it's free.

[LAwLz]

25 minutes ago, AluminiumTech said:

Threatening a ban over this seems quite overkill considering the above and considering the fact that posting a copyrighted image or copyrighted video on the forum isn't treated the same way. 

I might be wrong here, but I have a feeling GoodBytes would not have such a threatening tone if Cinebench was on Steam instead of the Microsoft Store...

 

"You have to use the Microsoft Store or else I will ban you!"

 

 

 

Just now, Princess Cadence said:

I have deleted the Windows Store from all my Windows 10 machines, not only that I have never logged in with a "Microsoft account" I still only use local ADM accounts. 

 

Up to this date that never mattered as I never wanted something that was only accessible through the Windows Store. 

 

I'm not sure how far Goodbytes is exaggerating on this matter or not, I'm sure the fact there was a 'taken down request' on Guru3D had a big impact on this, however I do feel lucky enough for the stand alone software been floating around by other sources if I'm honest, otherwise I wouldn't have got it even though it's free. 

I wonder why they chose to only distribute it through the Microsoft Store and App Store.

The App Store I kind of get, but by not having a stand alone installer for Windows they are preventing a lot of their potential users from using the program (everyone not on Windows 10 for example). Especially since it is an absolutely arbitrary limitation.

[skunkmunkey]

Would the people right at the top end of the Cinebench charts be using Win10 or are they more likely to be using some form of Linux? I would presume that Linux could be run with far fewer background processes and allow them to squeeze out every last point, surely by limiting it to Win10 and Mac they'd be pushing away some of those that contribute to its appeal as a benchmarking tool