Mass performance benchmark


#1

Is it worth everyone (who wants to of course) using the same benchmark software on their V as the first thing they install and run and passing on the results to Eve?

Didn’t know if it might make for some useful/interesting data to look at? :slight_smile:

Of course this could be a terrible idea as let’s face it, I’ve come up with it :wink:


#2

Sounds great, which software would you or others suggest?


#3

Good idea but few things that popped into my mind.

There should be some “ground rules” how to do this. What I mean that if the box has been in a DHL/UPS/whatever courier van for like 8hours and you just unbox it and power it on -> run the benchmark the scores probably will be different than if you let it “warmup” a little first

Things like these, so that the benchmarking environment is atleast somewhat “standard” :slight_smile:


#4

Nice idea. But please make sure that the software you choose is not one which is filling up my new pc with rubbish or even adware.

It would be even better if the software will come in an USB-stick version.


#5

I’ve no idea which would be good software. My days of benchmarking and overclocking etc are well over haha

@Team might have a suggestion for software they would find the results useful from? :slight_smile:


#6

great idea and i would gladly join. Also have the same request as @FFrankHB dont wanna have a thrash 2in1 within the first day of receiving it :wink:


#7

I’m not keen on benchmarking, but i would suggest Cinebench.


#8

I think this idea is great. I’m no benchmarking Software expert either but surely it could also output CPU/Motherboard temp?
Maybe if anyone here is a Developer they could knock something up with a tiny install footprint (or even an exe) that could be made available to the community so that we’re all using an identical tool with a standard configuration.


#9

For my Part i would also do the benchmark

my Suggestion: using one of the tools notebookcheck used for theire review


#10

I’m down to do some benchmarking. Sounds like a great idea.


#11

While there are pros to benchmarking, I have a feeling it will create some hostility among those with similar machines who don’t get similar results.


#12

Your point is, we shouldn’t do it, because if we do ppl. that received a dud would know it and then complain about it here ?


#13

Tho at the same time, anyone can run a benchmark it seems and compare it to the websites that gather such data and get the same happy or sad result

End of day, it’s going to be barely noticeable I’d wager


#14

It’s true, Kee, I agree

  • tech enthusiasts tend to get caught up a lot in benchmarks and how things perform but in real world conditions, I won’t notice the split second difference it may take to load up Gmail in my web browser. Now of course if you add this split second up over a bazillion instances, sure it’ll make a difference but one didn’t get a V for the sole purpose of CPU heavy tasks.

#15

I figure if it’s barely noticeable then folk will have their minds at ease and if it’s erratic then it gives eve a heads up :slight_smile:


#16

I am down for benchmmarking!

As @inffy mentioned, a standard procedure should be made first though. That way the results are somewhat consistent.


#17

I like the idea! Perhaps we can have a “poll” for every CPU version with some bell curve distribution of expected results. People who have tested their devices could “vote” the value closest to the result they achieved. That would give us a good overview of the collective results as well as how many participated.

An example:

CPU: i7
Expected result: ~200 (arbitrary number)
Checkboxes: 80 | 120 | 150 | 170 | 180 | 195 | 200 | 205 | 220 | …

A small writeup on how to download/install/use a software like cinebench would be great :slight_smile:

Any opinions?


#18

PCMark. 3DMark, Cinebench. Whats important is to run the benchmark several times, and see how the score changes after each run. Also to prevent background program from interfering with the score, make sure the benchmark program is running at either “Above Normal” or “High” priority.

https://www.maxon.net/en/products/cinebench/

For storage we could use CrystalDiskMark.

I would go with a range, so 80-100, 100-120, etc. with even ranges so you can see the bell curve itself. (If you put smaller range towards the expected value, the graph would then look flat)


#19

if we are going to do this, someone will have to come up with step by step instructions. Also, we need to make sure the ranges are small enough to get a proper idea of how things are performing.

Great idea @Cluskey_Smith!


#20

Woop, I’m enjoying not being a massive knob for a change! :slight_smile:

I’d keep the benchmarking as simple as possible so it’s something most everyone can do without much bother and hopefully as the first thing they install and run so that all should be equal?

Happy to take advisement on this though as benchmarking isn’t something I know much about or what metrics would be useful