Working on my first benchmarks, could use some feedback

[MGsubbie]

I don't know if this is the right place to post this... If not, feel free to move it to a more appropriate forum mods.

 

Tomorrow I'm switching from a 3770k to an 8700k. I've spent the past 2 days on some benchmarking, and this is what I have so far. I'm planning on adding frames per second, higher is better and the likes, as well as the full specs. I'll bench the 8700k at stock speeds (all core turbo is 4.3 matching my 3770k, so I can do a clock per clock comparison) and whatever overclock I can achieve. The non-3770k numbers are just placeholders. How is this looking so far? Any tips to improve what I'm working on?

[PokeCatz]

Welp, time to upgrade my 3770.

[Aidanlockett1]

6 minutes ago, PokeCatz said:

Welp, time to upgrade my 3770.

i think you might be ok, im running a quad core rn 

[PokeCatz]

1 minute ago, Aidanlockett1 said:

i think you might be ok, im running a quad core rn 

I have to upgrade my GPU before I upgrade my mobo or processor. The 650ti is kind of a POS.

[phongle123]

9 minutes ago, PokeCatz said:

Welp, time to upgrade my 3770.

Welp, time to downgrade my 1700x to a 8700k. Dat Single Score.

[Aidanlockett1]

Just now, PokeCatz said:

I have to upgrade my GPU before I upgrade my mobo or processor. The 650ti is kind of a POS.

Riiiip, I have a 1060 its amazing for 1080p

[MGsubbie]

Just in case you missed my edit, only the 3770k numbers are real, I haven't tested the 8700k yet.

[DocSwag]

You may want to add 1080p as well or use just 1080p. At 1080p the bottleneck is more on the GPU; unless your going 1080 ti sli or something the difference at 1440p will probably be very small, and won't be likely to illustrate a large difference.

 

I would also consider adding a few more synthetics, but other than that it looks pretty good. You seem to have a good range of different games that you're testing with.

32 minutes ago, PokeCatz said:

Welp, time to upgrade my 3770.

The 8700k numbers aren't real, they're place holders  

[MGsubbie]

On 10/18/2017 at 4:15 AM, DocSwag said:

You may want to add 1080p as well or use just 1080p. At 1080p the bottleneck is more on the GPU; unless your going 1080 ti sli or something the difference at 1440p will probably be very small, and won't be likely to illustrate a large difference.

 

I would also consider adding a few more synthetics, but other than that it looks pretty good. You seem to have a good range of different games that you're testing with.

Thanks for the advice. I decided not to go with 1080p because it's not a realistic scenario for a 1080ti user. I decided not to include too many synthetics either for the same reason. There are a ton of benchmarks by people who are much better at it than me that look at all of that. Those people tend to focus on tests where everything other than CPU power is as close as possible. For example Digital Foundry tends to pair their Sandy/Ivy Bridge CPU's with 2133-2400Mhz RAM, rather than what I think the average user has, being 1600Mhz. (When single threaded use reaches maximum, faster memory can help a great deal.)

 

I want to create a more realistic starting point for 2nd/3rd generation Intel users, Everybody knows that the theoretical gains from Coffee Lake are huge. I want to focus more on the gains that people get in games right now, to help them decide whether to upgrade now or keep their current CPU a generation longer/wait for Ryzen refresh.