Question I feel like my PC is underperforming

Toggle sidebar Toggle sidebar
C

cannyy

Honorable
Jan 10, 2018
35
0
10,530
I feel like my PC has always under performed in comparison to some other similar or same builds, and I'm not sure if it's just a placebo. My PC works well, however I feel like it should work better than it does as it's very high end.

I just ran a benchmark on UserBenchmarks and it's said that my CPU is at 79.7%.

This really shocked me as I've recently OC'd my CPU and feel like I may have messed something up, any recommendations or am I just stuck?
 
Sort by date Sort by votes
Single core speed looks like it passed. Multi-core didn't go so well.

How were the temperatures and clock speeds during the test?

Your overclock might be giving you good results in what you used to test it, but poor results in other areas.
 
Upvote 0 Downvote
Eximo said:
Single core speed looks like it passed. Multi-core didn't go so well.

How were the temperatures and clock speeds during the test?

Your overclock might be giving you good results in what you used to test it, but poor results in other areas.
Click to expand...
It was hitting 100c and then dropping down I believe
EDIT: Disregard the above, it's around 70C according to my software, the 100C was the stress test software I ran previously
 
Last edited:
Upvote 0 Downvote
cannyy said:
Oh I see, what kind of benchmark software would you recommend to gauge how well my machine is doing in comparison? I'm worried I have a problem is all.
Click to expand...
The best benchmark would be running YOUR workload on different setups.
That is no so easily done.

You must decide if you want/need better performance.
Then try some experiments to see what responds better.

If, for example, you run your games but reduce the graphics settings and resolution and still get the same performance it will tell you that a graphics card upgrade will not do much.

Or, if you limit the cpu performance to 80% and you find no difference, that would suggest that the cpu was ok and more likely the graphics card could be updated.
 
Upvote 0 Downvote
geofelt said:
The best benchmark would be running YOUR workload on different setups.
That is no so easily done.

You must decide if you want/need better performance.
Then try some experiments to see what responds better.

If, for example, you run your games but reduce the graphics settings and resolution and still get the same performance it will tell you that a graphics card upgrade will not do much.

Or, if you limit the cpu performance to 80% and you find no difference, that would suggest that the cpu was ok and more likely the graphics card could be updated.
Click to expand...
Yeah I suppose it's just difficult to gauge what's the average performance of my build.
Almost all the builds that are similar to mine are using DDR5 or they are literally completely fresh builds so their performance is going to be better, without a doubt.

I don't believe the GPU is an issue, I've got a 4090 lol.

I think it's just a placebo on my end thinking that my cpu should be better but I can't confirm as I don't have anyone to compare to as only my build has all the things I have installed and the conditions I have so it's difficult to get a correct comparison.
 
Upvote 0 Downvote
cannyy said:
Yeah I suppose it's just difficult to gauge what's the average performance of my build.
Almost all the builds that are similar to mine are using DDR5 or they are literally completely fresh builds so their performance is going to be better, without a doubt.

I don't believe the GPU is an issue, I've got a 4090 lol.

I think it's just a placebo on my end thinking that my cpu should be better but I can't confirm as I don't have anyone to compare to as only my build has all the things I have installed and the conditions I have so it's difficult to get a correct comparison.
Click to expand...
Many games depend on the performance of the single master thread.
Run the cpu-Z bench on your 12900K.
You should get a single thread rating of about 835.
valid.x86.fr

CPU-Z Benchmark for Intel Core i9-12900K (1T) - CPU-Z VALIDATOR

Best CPU performance - 64-bit - August 2025
valid.x86.fr valid.x86.fr

FWIW, mine scores 831
 
Upvote 0 Downvote
Upvote 0 Downvote
You are running DDR4 3000 (with a 3200 kit, might want to look into that) according to userbenchmark. That is a little lower than most people would have with a 12900K.

DDR4 3600 or DDR5 5600+ would be more typical.

I would remove the overclock and see how it performs at stock. Then try your overclock again. If you get worse results, then your overclock is making things worse.

Been a while since I did a deep dive on overclocking, since there isn't amazing gains to be had any longer, but it is fairly easy to have impressive stats and poor performance.

With previous Intel chips, getting the clock speed too far from the cache speed could cause performance issues. With these newer chips memory Gear is a major factor. With the huge core counts, sometimes you have to tweak by core, or control the efficiency cores to prevent overheating there or reign in the performance cores so the efficiency cores can run higher.

Not sure what would cause poor single core performance outside of thermal or power limits though, maybe background applications?
 
  • Like
Reactions: cannyy
Upvote 0 Downvote
Seems nominal to me; not a significantly different result.

Score vary, and those proud of their scores via overclocking are eager to post.

Your pc may feel sluggish if any significant work is done on hard drives.
Moving anything you actively use to a ssd, even a sata ssd will make things quicker.
 
  • Like
Reactions: cannyy
Upvote 0 Downvote
Eximo said:
You are running DDR4 3000 (with a 3200 kit, might want to look into that) according to userbenchmark. That is a little lower than most people would have with a 12900K.

DDR4 3600 or DDR5 5600+ would be more typical.

I would remove the overclock and see how it performs at stock. Then try your overclock again. If you get worse results, then your overclock is making things worse.

Been a while since I did a deep dive on overclocking, since there isn't amazing gains to be had any longer, but it is fairly easy to have impressive stats and poor performance.

With previous Intel chips, getting the clock speed too far from the cache speed could cause performance issues. With these newer chips memory Gear is a major factor. With the huge core counts, sometimes you have to tweak by core, or control the efficiency cores to prevent overheating there or reign in the performance cores so the efficiency cores can run higher.

Not sure what would cause poor single core performance outside of thermal or power limits though, maybe background applications?
Click to expand...
Just firstly addressing the 3200mhz running at 3000mhz, I have updated this to 3200mhz now, it wasn't working previously so that's why it was lower.

Would DDR5 5600 or DDR4 3600 make a massive difference or would it not be worth the cost?

My power supply is a Corsair TX850M 80+ Gold, would I be better with a larger PSU or is that not going to make a difference?
My temps are high (70/ high 80s, low 90s under load[depends on the game really]) but as far as I'm aware, the 12900k runs hotter than most anyway.
 
Upvote 0 Downvote
DDR5 would mean a new motherboard, how much that is worth it is up to you. Same with swapping in slightly faster memory, DDR4 is pretty cheap right now though.

More PSU doesn't change the rest of your hardware, no. Only so much a give CPU and GPU can draw. As long as your system isn't randomly shutting down or crashing, you are fine with that PSU.

I would first try to get what you have now benchmarked to a rough equivalent of other systems with similar components. Seek out 12900k reviews using DDR4 and see if you can replicate their results.
 
  • Like
Reactions: cannyy
Upvote 0 Downvote
geofelt said:
Once you have sufficient ram, ram speed means little for games.
Here is one study:
www.pcgamer.com

Don't waste money chasing RAM speed for gaming on AMD or Intel

Revisiting the question that every gamer considers when building a new gaming system: does RAM speed matter?
www.pcgamer.com www.pcgamer.com

If budget is not a problem, you could always upgrade to 13900K.
Your motherboard would need a current bios level.

What is your cpu cooler?
Click to expand...
I could but it's a lot of money lol.

For cooler I have a Corsair ICUE H100i Elite Capellix.
 
Upvote 0 Downvote
cannyy said:
www.3dmark.com

I scored 29 891 in Time Spy

Intel Core i9-12900K Processor, NVIDIA GeForce RTX 4090 x 1, 32768 MB, 64-bit Windows 11}
www.3dmark.com www.3dmark.com

That was one I ran previously, let me know what you think.
Click to expand...
Definitely something going on with the CPU side of things as your graphics score is right where it should be. It might be something with the way the memory is running (like if your system is in Gear 2 despite running slow memory) or it could be the slow memory period (I don't think it would have quite this impact). What do the logged graphs look like from that run? I think only the submitter can see those, but I'm not in front of my PC to check.
 
Upvote 0 Downvote
thestryker said:
Definitely something going on with the CPU side of things as your graphics score is right where it should be. It might be something with the way the memory is running (like if your system is in Gear 2 despite running slow memory) or it could be the slow memory period (I don't think it would have quite this impact). What do the logged graphs look like from that run? I think only the submitter can see those, but I'm not in front of my PC to check.
Click to expand...
At least it's not just me that's thinking this!

I can't seem to figure out where the graphs are, any ideas?
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom