Jump to content

Now that I know UserBenchmark is garbage, what are the best benchmark tests?

noahthedestroyer5
4 minutes ago, TomvanWijnen said:

 If you just look at those, and aren't looking at expensive enough chips where the maximum of 8 cores will really matter, are those values any good?

No one knows, because no one knows how they're actaully measured.

 

But here's a question:

Shouldn't you look at real-world performance in the software that you're actually using, instead of artificial benchmarks?

 

 

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, lewdicrous said:

There will be some variation in the numbers, even for similar parts, this could be due to margin of error, the silicon lottery, etc.

It's not something you should really be worried about, unless the part is severely under performing.

ok thank you. And my scores on timespy seem to match

Link to comment
Share on other sites

Link to post
Share on other sites

5 minutes ago, Senzelian said:

No one knows, because no one knows how they're actaully measured.

 

But here's a question:

Shouldn't you look at real-world performance in the software that you're actually using, instead of artificial benchmarks?

Ah, great. :P

 

Hmm, maybe you should, but if you're just quickly comparing things (generally not to buy them, but just to know), an artificial benchmark can give a good indication.

PC SPECS: CPU: Intel Core i7 3770k @4.4GHz - Mobo: Asrock Extreme 4 (Z77) - GPU: MSI GeForce GTX 680 Twin Frozr 2GB - RAM: Crucial Ballistix 2x4GB (8GB) 1600MHz CL8 + 1x8GB - Storage: SSD: Sandisk Extreme II 120GB. HDD: Seagate Barracuda 1TB - PSU: be quiet! Pure Power L8 630W semi modular  - Case: Corsair Obsidian 450D  - OS: Windows 7

Link to comment
Share on other sites

Link to post
Share on other sites

17 minutes ago, Caroline said:

userbenchmark treats stock and overclocked CPUs as the same so it might show you that your stock clocked let's say 8700K is "severely underperforming" as it's comparing it against chips that run at over 5GHz 

it relies on AVX to spit a score, therefore Intel CPUs will always score higher than AMD because of the better AVX processing

ah ok thanks. is 3dmark (timespy) trustworthy? can i be sure im performing ok if my results match up with others?

Link to comment
Share on other sites

Link to post
Share on other sites

Im looking for benchmark tests that will test my gaming pc to see if its performing up to standards through comparison with others results. What are the best tests for this? Basically, im looking for a test that will prove im performing ok.

Link to comment
Share on other sites

Link to post
Share on other sites

cinebench, firestrike... watch recent review videos and they should use several programs that will be good

desktop

Spoiler

r5 3600,3450@0.9v (0.875v get) 4.2ghz@1.25v (1.212 get) | custom loop cpu&gpu 1260mm nexxos xt45 | MSI b450i gaming ac | crucial ballistix 2x8 3000c15->3733c15@1.39v(1.376v get) |Zotac 2060 amp | 256GB Samsung 950 pro nvme | 1TB Adata su800 | 4TB HGST drive | Silverstone SX500-LG

HTPC

Spoiler

HTPC i3 7300 | Gigabyte GA-B250M-DS3H | 16GB G Skill | Adata XPG SX8000 128GB M.2 | Many HDDs | Rosewill FBM-01 | Corsair CXM 450W

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, noahthedestroyer5 said:

Im looking for benchmark tests that will test my gaming pc to see if its performing up to standards through comparison with others results. What are the best tests for this? Basically, im looking for a test that will prove im performing ok.

CPU: Cinebench R20 (compare your result to a result from a good review)

GPU: 3DMark FireStrike/TimeSpy or Superposition (these show some scores already, but same general approach as Cinebench; compare your scores to others/reviews) 

RAM: just make sure it's running at the correct speeds (usually that means turning on XMP/DOCP) and if you're really worried, run a memory diagnostic check

Storage: CrystalDiskMark (compare your score to the reviews).

"We're all in this together, might as well be friends" Tom, Toonami.

 

mini eLiXiVy: my open source 65% mechanical PCB, a build log, PCB anatomy and discussing open source licenses: https://linustechtips.com/topic/1366493-elixivy-a-65-mechanical-keyboard-build-log-pcb-anatomy-and-how-i-open-sourced-this-project/

 

mini_cardboard: a 4% keyboard build log and how keyboards workhttps://linustechtips.com/topic/1328547-mini_cardboard-a-4-keyboard-build-log-and-how-keyboards-work/

Link to comment
Share on other sites

Link to post
Share on other sites

Just now, Cyracus said:

cinebench, firestrike... watch recent review videos and they should use several programs that will be good

ok thank you

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, minibois said:

CPU: Cinebench R20 (compare your result to a result from a good review)

GPU: 3DMark FireStrike/TimeSpy or Superposition (these show some scores already, but same general approach as Cinebench; compare your scores to others/reviews) 

RAM: just make sure it's running at the correct speeds (usually that means turning on XMP/DOCP) and if you're really worried, run a memory diagnostic check

Storage: CrystalDiskMark (compare your score to the reviews).

thanks. can i use timespy for my cpu too? will it still show my processor is fine?

Link to comment
Share on other sites

Link to post
Share on other sites

5 minutes ago, noahthedestroyer5 said:

What are the best tests for this?

i use cinebench R20 for CPU, game FPS for GPU (or just clock speed and temps really)

 

but 3DMark timespy/firestrike is also good for GPU scores, not too sure about their CPU scores for cpu with many cores though

-sigh- feeling like I'm being too negative lately

Link to comment
Share on other sites

Link to post
Share on other sites

Just now, noahthedestroyer5 said:

thanks. can i use timespy for my cpu too? will it still show my processor is fine?

Sure.

Personally I prefer Cinebench as a CPU benchmark, as there is just much more content specifically targeted to its CPU scores.

"We're all in this together, might as well be friends" Tom, Toonami.

 

mini eLiXiVy: my open source 65% mechanical PCB, a build log, PCB anatomy and discussing open source licenses: https://linustechtips.com/topic/1366493-elixivy-a-65-mechanical-keyboard-build-log-pcb-anatomy-and-how-i-open-sourced-this-project/

 

mini_cardboard: a 4% keyboard build log and how keyboards workhttps://linustechtips.com/topic/1328547-mini_cardboard-a-4-keyboard-build-log-and-how-keyboards-work/

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, noahthedestroyer5 said:

thanks. can i use timespy for my cpu too? will it still show my processor is fine?

Time Spy also includes a CPU physics test, though when you look up benchmarks for CPUs, you'll more often than not see Cinebench results rather than 3DMark, so I find that it's easier to compare CPU results using Cinebench.

Desktop: Intel Core i9-9900K | ASUS Strix Z390-F | G.Skill Trident Z Neo 2x16GB 3200MHz CL14 | EVGA GeForce RTX 2070 SUPER XC Ultra | Corsair RM650x | Fractal Design Define R6

Laptop: 2018 Apple MacBook Pro 13"  --  i5-8259U | 8GB LPDDR3 | 512GB NVMe

Peripherals: Leopold FC660C w/ Topre Silent 45g | Logitech MX Master 3 & Razer Basilisk X HyperSpeed | HIFIMAN HE400se & iFi ZEN DAC | Audio-Technica AT2020USB+

Display: Gigabyte G34WQC

Link to comment
Share on other sites

Link to post
Share on other sites

19 hours ago, minibois said:

Sure.

Personally I prefer Cinebench as a CPU benchmark, as there is just much more content specifically targeted to its CPU scores.

true

 

19 hours ago, Mateyyy said:

Time Spy also includes a CPU physics test, though when you look up benchmarks for CPUs, you'll more often than not see Cinebench results rather than 3DMark, so I find that it's easier to compare CPU results using Cinebench.

ok thank you ill download cinebench

 

19 hours ago, Moonzy said:

i use cinebench R20 for CPU, game FPS for GPU (or just clock speed and temps really)

 

but 3DMark timespy/firestrike is also good for GPU scores, not too sure about their CPU scores for cpu with many cores though

will timespy be proof that my system is performing ok?

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, noahthedestroyer5 said:

will timespy be proof that my system is performing ok?

as an overall, yes

but i would trust it's GPU score more than CPU score, as i've mentioned

 

sounds like you're being a bit paranoid haha

-sigh- feeling like I'm being too negative lately

Link to comment
Share on other sites

Link to post
Share on other sites

5 minutes ago, Moonzy said:

as an overall, yes

but i would trust it's GPU score more than CPU score, as i've mentioned

 

sounds like you're being a bit paranoid haha

im naturally really anxious lol

Link to comment
Share on other sites

Link to post
Share on other sites

12 minutes ago, Moonzy said:

as an overall, yes

but i would trust it's GPU score more than CPU score, as i've mentioned

 

sounds like you're being a bit paranoid haha

so i just ran unigine heaven with my 2080 super and ryzen 3700x. Is 4,270 a good score? I cant seem to find much online

Link to comment
Share on other sites

Link to post
Share on other sites

13 minutes ago, noahthedestroyer5 said:

will timespy be proof that my system is performing ok?

In order for any benchmarking test to prove that, we first need a rigorous definition of "performing ok".

We have no idea what that means. That could mean anything from "the machine boots and I can log in" to "I'm setting world overclocking records", to grossly specific things like very specific targets for hard drive performance or achieving a certain framerate in a certain scene in a specific game configured a special way.

When trying to prove that performance has surpassed a certain target level, we need a very rigorous definition of what that performance should be. This helps inform us as to what benchmark tooling would be the best, and helps inform you on what results you will deem acceptable.

So, without further ado:

What does "prove that my system is performing ok" mean, exactly?

ENCRYPTION IS NOT A CRIME

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, straight_stewie said:

In order for any benchmarking test to prove that, we first need a rigorous definition of "performing ok".

We have no idea what that means. That could mean anything from "the machine boots and I can log in" to "I'm setting world overclocking records", to grossly specific things like very specific targets for hard drive performance or achieving a certain framerate in a certain scene in a specific game configured a special way.

When trying to prove that performance has surpassed a certain target level, we need a very rigorous definition of what that performance should be. This helps inform us as to what benchmark tooling would be the best, and helps inform you on what results you will deem acceptable.

So, without further ado:

What does "prove that my system is performing ok" mean, exactly?

like it means will my performance in game look like what it should 

Link to comment
Share on other sites

Link to post
Share on other sites

4 minutes ago, noahthedestroyer5 said:

so i just ran unigine heaven with my 2080 super and ryzen 3700x. Is 4,270 a good score? I cant seem to find much online

it'll be hard to find scores to compare with unigine heaven, especially with a system that's similar to yours

3DMark makes that easier, there's a button that directly shows u similar setup scores iirc

-sigh- feeling like I'm being too negative lately

Link to comment
Share on other sites

Link to post
Share on other sites

3 minutes ago, noahthedestroyer5 said:

like it means will my performance in game look like what it should 

What should your ingame performance look like?

ENCRYPTION IS NOT A CRIME

Link to comment
Share on other sites

Link to post
Share on other sites

18 hours ago, Moonzy said:

it'll be hard to find scores to compare with unigine heaven, especially with a system that's similar to yours

3DMark makes that easier, there's a button that directly shows u similar setup scores iirc

ah ok thank you

 

18 hours ago, straight_stewie said:

What should your ingame performance look like?

I dont know lol. like on rainbow 6 with my set up i expect around 150

Link to comment
Share on other sites

Link to post
Share on other sites

Just now, noahthedestroyer5 said:

I dont know lol. like on rainbow 6 with my set up i expect around 150

Ok. So install Rainbow 6, play around, and see if you get what you are expecting.

ENCRYPTION IS NOT A CRIME

Link to comment
Share on other sites

Link to post
Share on other sites

5 minutes ago, straight_stewie said:

Ok. So install Rainbow 6, play around, and see if you get what you are expecting.

ok i just played it. I got around 160s

Link to comment
Share on other sites

Link to post
Share on other sites

Just now, WI-FIultrasnoop said:

Did you try the benchmark test that R6S allows you to do.

I get a higher fps in the benchmark around 190s i believe

Link to comment
Share on other sites

Link to post
Share on other sites

27 minutes ago, noahthedestroyer5 said:

ok i just played it. I got around 160s

Then there you go. You wanted to prove that your machine could play rainbow six at 150FPS or greater, and you did.

 

Task complete.

ENCRYPTION IS NOT A CRIME

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×