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

[noahthedestroyer5]

Ive noticed that so many people say to not use userbenchmark, why? Doesnt it give you comparisons to similar hardware so you can see if youre ok in performance? I stopped using it because so many people hate it. And what makes 3Dmark so trustworthy to see if youre performing ok?

[lewdicrous]

Afaik, userbenchmark is biased towards Intel, so their benchmarks are calculated/made in such a way that favors Intel. Could be wrong though.

[Mateyyy]

Because many of those figures you see on Userbenchmark are about as reliable as tossing random numbers in random.org.

 

[Haro]

because userbenchmark is not reliable, they alter results afaik, and is inaccurate. 

3Dmark is more trust worthy to say the least.

[noahthedestroyer5]

Just now, lewdicrous said:

Afaik, userbenchmark is biased towards Intel, so their benchmarks are calculated/made in such a way that favors Intel. Could be wrong though.

ah ok. So it could be trustworthy if it wasnt?

[noahthedestroyer5]

1 minute ago, Mateyyy said:

Because many of those figures you see on Userbenchmark are about as reliable as tossing random numbers in random.org.

 

so i just should stick to 3dmark to test performance?

[lewdicrous]

Just now, noahthedestroyer5 said:

ah ok. So it could be trustworthy if it wasnt?

Userbenchmark is not considered to be trustworthy by many. Their recent actions certainly burned whatever trust they had to begin with.

[VeganJoy]

3 minutes ago, noahthedestroyer5 said:

Ive noticed that so many people say to not use userbenchmark, why? Doesnt it give you comparisons to similar hardware so you can see if youre ok in performance? I stopped using it because so many people hate it. And what makes 3Dmark so trustworthy to see if youre performing ok?

userbenchmark engages in a lot of unprofessional practices and you can find plenty of documentation about all that. buuuuuut the single thing i continue to use their site for is same-hardware comparisons. keep in mind that many benchmarks are heavily overclocked systems that will make your scores look like shit but it's nice to see how your specific hardware is performing. 3dmark/passmark do this better though, but i usually use a variety of test suites. for anything other than comparing your system to other identical systems, ub is shit lol

[noahthedestroyer5]

2 minutes ago, TofuHaroto said:

because userbenchmark is not reliable, they alter results afaik, and is inaccurate. 

3Dmark is more trust worthy to say the least.

so 3d mark is really trustworthy?

[lewdicrous]

Just now, noahthedestroyer5 said:

so i just should stick to 3dmark to test performance?

Are you looking for something to benchmark your rig or are you looking to compared between hardware?

[Haro]

10 hours ago, noahthedestroyer5 said:

so 3d mark is really trustworthy?

they are.

[noahthedestroyer5]

Just now, VeganJoy said:

userbenchmark engages in a lot of unprofessional practices and you can find plenty of documentation about all that. buuuuuut the single thing i continue to use their site for is same-hardware comparisons. keep in mind that many benchmarks are heavily overclocked systems that will make your scores look like shit but it's nice to see how your specific hardware is performing. 3dmark/passmark do this better though, but i usually use a variety of test suites. for anything other than comparing your system to other identical systems, ub is shit lol

ok lol. so is 3dmark good?

[noahthedestroyer5]

Just now, lewdicrous said:

Are you looking for something to benchmark your rig or are you looking to compared between hardware?

both. to test my system and compare

[TomvanWijnen]

2 minutes ago, Mateyyy said:

Because many of those figures you see on Userbenchmark are about as reliable as tossing random numbers in random.org.

 

But how about if you ignore those total scores, and instead look at the actual values? (so ignoring the two values with a red dot in front of them, and only looking at the values with the yellow line)

 

Are those any good?

[VeganJoy]

Just now, noahthedestroyer5 said:

both. to test my system and compare

passmark will do that for every component which is nice. theyre not too shabby

[noahthedestroyer5]

1 minute ago, TomvanWijnen said:

But how about if you ignore those total scores, and instead look at the actual values? (so ignoring the two values with a red dot in front of them, and only looking at the values with the yellow line)

 

Are those any good?

im not sure. I guess not maybe?

[noahthedestroyer5]

1 minute ago, VeganJoy said:

passmark will do that for every component which is nice. theyre not too shabby

true i do have passmark but i prefer 3dmark since its on steam a=

[lewdicrous]

Just now, noahthedestroyer5 said:

both. to test my system and compare

For testing, you have 3DMark, Heaven, Cinebench, Blender and some others. Some games also have builtin benchmarks.

[VeganJoy]

Just now, noahthedestroyer5 said:

true i do have passmark but i prefer 3dmark since its on steam a=

pretty sure 3dmark only scores your cpu and gpu which can be compared with other users' benchmarks

[noahthedestroyer5]

Just now, lewdicrous said:

For testing, you have 3DMark, Heaven, Cinebench, Blender and some others. Some games also have builtin benchmarks.

ah ok so if my score is the same as others with the same hardware im good?

[noahthedestroyer5]

Just now, VeganJoy said:

pretty sure 3dmark only scores your cpu and gpu which can be compared with other users' benchmarks

it does

[noahthedestroyer5]

Just now, BlueScope819 said:

"effective speed" is defined as 40% single-core, 58% quad-core, and 2% multi-core.

 

Do you think anything above a quad core should be weighted 2%?

 

That big 9% weighting at the top of the average score is based off of that. Neither of those are accurate you need to look at the actual 1 core, 2 core, 4 core, and 8 core numbers but notice that it doesn't go beyond 8 core? 8 cores are very consumery these days.

ah ok. im new to computers so idk this stuff

[TomvanWijnen]

4 minutes ago, BlueScope819 said:

"effective speed" is defined as 40% single-core, 58% quad-core, and 2% multi-core.

 

Do you think anything above a quad core should be weighted 2%?

 

That big 9% weighting at the top of the average score is based off of that. Neither of those are accurate you need to look at the actual 1 core, 2 core, 4 core, and 8 core numbers but notice that it doesn't go beyond 8 core? 8 cores are very consumery these days.

 

(not to mention a 10900k would score the same as a 10900k with two of its cores disabled)

Like I said, I don't care about the "Effective Speed" and "Average Score" values, I just care about the 1, 2, 4, 8 core speeds. 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?

[noahthedestroyer5]

2 minutes ago, BlueScope819 said:

Observe.

(remember that those CPUs are about $200 more expensive compared to their AMD counterparts when you factor in the cost of the motherboard and cooler)

ok thank you im watching now. so basically is 3dmark trustworthy to see if im performing ok?

[lewdicrous]

6 minutes ago, noahthedestroyer5 said:

ah ok so if my score is the same as others with the same hardware im good?

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.