21 Programs to Analyze and Benchmark Your Hardware

nnguy2

Posts: 128   +223
I don't think you realize how much stress Prime95 puts on your cpu. Prime95 is designed to crash your system if even a little voltage is not right.

There's a reason why a prime95 stable is a thing.
 

nnguy2

Posts: 128   +223
In other words if your system is Prime95 stable, you'll never ever crash in any application outside of prime95.
 

Markoni35

Posts: 768   +268
I don't think you realize how much stress Prime95 puts on your cpu. Prime95 is designed to crash your system if even a little voltage is not right.
I don't think you realize that Prime95 doesn't even start. It doesn't come to the part to start stressing the CPU. I've done tons of things to my CPU, started all kinds of CPU stress tests and rarely any of them made my cooler come to 50% of its full speed. That's because I have excellent cooling.

Prime95 didn't even come to the part to start stressing my CPU. It just died at startup. They've got bug in the software. I'm not the only one that was unable to run the app. There are many other users. At least the competing apps (like Fritz Benchmark or wPrime) don't die so stupidly. Those apps can't use all the cores (if you have more than 4) but at least they work within those limitations.
 
These benchmark tests are mainly useful in checking whether a rig is running OK and whether config changes have improved or reduced performance. Most of these tools use a regression formula to produce scores using a synthetic test. At least of two of them have recently revised their regression formulas because the results didn't correspond with actual relative performance. In the real world of IT the benchmarks that are used always imitate the production apps that are going to be run. Applied to gaming this means that buying decisions should only be made on real game benchmarks and nothing else. But there is a major flaw in the way the usual game benchmarks are done. The flaw is that cpu and gpu loads aren't measured. Yet load data is just as important as speed data.
Perhaps the main paradox in benchmarking cpus for gaming is the heavy use of non-gaming benchmark tools.
 

grumblguts

Posts: 352   +303
SPECviewperf
RhodiumLC
RealBench 2.2
LuxMark
Cryengine_BaronHaussmann
ffxiv-shadowbringers-bench FAVORS NVIDIA...
Catzilla


These you posted
Unigine Superposition AND Basemark GPU both favour Nvidia. especially Basemark
 
  • Like
Reactions: Alfatawi Mendel

texasrattler

Posts: 988   +433
@neeyik I see you used a 9700k. Thats what I have. Have you noticed any high cpu usage with that during gaming?

I have used some of those programs, everything seems fine. Nothing out of ordinary. Even tried overclocking to 5ghz. No crashing or anything. But with games like Modern Warfare/Warzone and Apex I see cpu usage go as high as 100% during game. Downside is I can't watch streams/videos at the same time. Have you seen any issues like that or experienced anything?

I have read all kinds of things on the 9700K and there's plenty of info that talks about high usage but nothing I've tried helps. Task manager also shows modern warfare is using like 7000MB of memory with usage going to 90%. I have 16gb in dual channel. Using a msi mobo z390. Corsair h100i pro cooler, windows 10 pro 2004, Corsair rm 850x gold, gtx 1070.
 
Last edited:

brucek

Posts: 528   +612
TechSpot Elite
I'm getting too old for this sh*t.
I resemble that remark ;-)

I don't know if it exists, but my ideal tool would be a simple app that auto-identifies my hardware, and runs just enough tests to identify if there was a setup or configuration error that is costing me significant performance. Maybe with an option to run extended testing after that to check for a stability or thermal issue. Even better if I could set it to automatically run once a week to catch a future screw up or regression.
 
  • Like
Reactions: KernTUG

Nobina

Posts: 2,663   +2,292
For hardware info HWiNFO is pretty much all you need. As for benchmarks, if I'm interested how a GPU/CPU performs in a certain game I'll just look at a YouTube video of that GPU/CPU running that game. It is the most accurate way of seeing how it performs.
 

geogan

Posts: 16   +12
  • Like
Reactions: sumeetrana

kmo911

Posts: 241   +31
Userbenchmark.com said it all https://www.userbenchmark.com/UserRun/33914056. allready benchmarked and 78 th % for pc gaming. when 11 th 12 th comes out I must beat this 10 th low end pc gameing laptop alot. so youre 4k goes over 500 fps are a nice one. LIKE
top end best amd pcie 4.0 https://www.3dmark.com/3dm03/654668...munqBVWPkRqB6_r6rS4nr6zRxB-HQAGq7qX7D_31O8adA so waiting for rx 6000 to go by this little lovley one too. then pcie 5.0 would arrive and so on.

https://www.userbenchmark.com/UserR...79XDVSdDuo82WQyqSfFiSnJKkh_ZMWxZDY9hCNy7zoh6Q over 100 with better cpu gpu nvme ssd pcie 4.0-6.0 into the future.
 
Last edited:

neeyik

Posts: 1,331   +1,416
Staff member
  • Thread Starter Thread Starter
  • #38
@neeyik I see you used a 9700k. Thats what I have. Have you noticed any high cpu usage with that during gaming?
Not especially. The thing to bear in mind with CPU usage figures is that the system can't tell the difference between whether a thread is active or stalled: both will be reported as 'usage'. So a high percentage value doesn't necessarily mean the processor is being worked all the time and if there are other aspects, either within hardware or software, that's holding up threads from being completed, then one will be left with the impression that it's the CPU that's at fault.
 

texasrattler

Posts: 988   +433
Not especially. The thing to bear in mind with CPU usage figures is that the system can't tell the difference between whether a thread is active or stalled: both will be reported as 'usage'. So a high percentage value doesn't necessarily mean the processor is being worked all the time and if there are other aspects, either within hardware or software, that's holding up threads from being completed, then one will be left with the impression that it's the CPU that's at fault.
Anything I could do to improve cpu usage or am I just screwed? Ive tried everything and then some. I keep getting told by most I just need to replace the processor.
 

neeyik

Posts: 1,331   +1,416
Staff member
  • Thread Starter Thread Starter
  • #40
Anything I could do to improve cpu usage or am I just screwed? Ive tried everything and then some. I keep getting told by most I just need to replace the processor.
My situation might be because I game at 1440p or 4K on the highest settings, with G-Sync on a 60 Hz monitor, so the performance bottleneck is nearly always going to be shifted off the CPU. But if one is wanting higher frame rates at lower resolutions, for example, then the CPU is always going to be hit a lot harder.

You could, perhaps, look into using dual channel, dual rank RAM, or using nvme drives, if you're not already doing so - these will help reduce the amount of time threads are stalled for, which in turn will lower the CPU usage.
 

texasrattler

Posts: 988   +433
My situation might be because I game at 1440p or 4K on the highest settings, with G-Sync on a 60 Hz monitor, so the performance bottleneck is nearly always going to be shifted off the CPU. But if one is wanting higher frame rates at lower resolutions, for example, then the CPU is always going to be hit a lot harder.

You could, perhaps, look into using dual channel, dual rank RAM, or using nvme drives, if you're not already doing so - these will help reduce the amount of time threads are stalled for, which in turn will lower the CPU usage.
yea already using all that. I can go to 60hz but that defeats why I bought a msi 27 monitor 144hz 1080p. I just cant figure out why the game is using so much cpu usage and memory.
 

Alfatawi Mendel

Posts: 70   +122
It is plainly clear this article is not for you. You are not an enthusiast nor an overclocker.

For those who do research and build their own rigs, usually you don't need to upgrade. That being said:

It is easy to see how you can take what hardware you have and compare if you swapped out other hardware. Most people throw money at it and have no idea it's a waste. I found a couple excellent resources you can compare against a single upgrade versus a whole new rig:

userbenchmark.com
gpucheck.com
I think there was a reason Userbenchmark wasn't included in the article.
The reason is that it's pretty much BS.
 
  • Like
Reactions: ColdSoup