Performance Testing for the Common Man: Fast, Easy Benchmarks You Can Run Right Now

16

Comments

+ Add a Comment
avatar

mattman059

" Perhaps you just got a new graphics card, or doubled your installed DRAM"

HOW DID YOU KNOW!!!!! 

avatar

loyd

I'm actually not familiar with Kombuster. But Furmark can actually push a graphics card beyond its rated parameters. So that test should be used with care.

avatar

akrazyassho

Question about the Heaven benchmark and tessellation in general.  In the benchmark you can turn on and off tessellation, but there is a huge difference.  Ex: the stones on the walls all come out to a certain height/ have a unque depth, but when you turn off tes. they are all flat walls/surfaces.  My question is, that depth that the stones/rocks have, is that built into the benchmark or is that really the card deciding and rendering at its own free will due to tessellation??

avatar

loyd

The answer is both. The programmer gives some hints as to how much tessellation should be generated. In the case of Heaven, the designers have given the user some control over this through tessellation levels. If you get past the performance test stuff, you'll see that "extreme" tessellation actually looks pretty bad, while moderate degrees of tessellation adds pleasing levels of geometric detail.

avatar

maleficarus™

Stress-testing is stupid! Why "bake" your CPU like that when it will never represent anything you will ever do in real world situations. Why not just toss your CPU in the oven for 30 mins at 350 degrees? Same silly logic...

avatar

steven4570

stress testing is more designed for testing an overclocks stability not really benchmarking, in that reguard it is very useful

avatar

xXSINS OF WARXx

Idk, why did Einstein come up with the mathematical equation of e=mc^2? Why did NASA decide to send a rocket to the moon? Or why do numerous people around the globe try to break countless Guinness world records? The answer is simple; because we can. If we as a nation were to simply stop pushing the boundaries that life sets before us then the conveniences we so ignorantly take for granted would cease to exist and human nature will never continue to evolve; those types of infectious negative ideas will only plague our children…

avatar

maleficarus™

What kind of nonesense is this? Baking your CPU with prime95 or IBT is not only stupid, but useless as well. It does not in anyway shape or form represent anything we do on our PC's. Common sense logic dictates that if this isn't a real world representation, then it is not logical to do! Common sense people!!

We can all jump off a bridge. Does that mean we should because we can?

avatar

steven4570

i hate people who use the term"common sense logic" because in most cases they have no fucking clue what theyre talking about.  Some systems if we change anytthing for what every reason, need to be tested for stability. one time i had stabiity issues with everything at stock speed on this one pc i built, turns out the default voltages in the outdated bios were wrong.  it has its place, its not going to fry or die because of it, cpus like that will be working far longer than youll be using them ill guarantee.

avatar

pcfxer

You stress test for stability to simulate the life/usage of the CPU within a much shorter time span. For example; you want to know that your overclock will last for five years at five hours per day so you run a stress test for say one week at tewnty four hours per day. The goal is to design a test that will allow you to certify that your overclock is stable for normal usage.

 

AMD and Intel both do this on their CPUs. Automotive manufacturers do this as well; most iconicly Porsche did this with the Beetle. This is common practice in engineering.

avatar

maleficarus™

Yes at stock speeds. Not OC speeds that everyone uses for stress-testing. Not a single person uses prime95 stock do they? So your point is not valid at all...

avatar

JCGPZ9

Actually I do when trying out a new thermal paste or heatsink. It's a great way to put sress on the CPU while still monitoring it.

avatar

Coldrage

maleficarus™ way to make yourself look like a total tool.

I'm using a Q6600 @ 4.0 GHz since august 2007 and fold every day with it, stressed tested it countless times and it is still going strong.
 
CPU's these days are well designed to deal with heat.
Putting a CPU in an oven is just a terrible analogy.

Folding as I speak with all 4 cores at 60c

As previously said, benchmarking is an excellent way to measure the performance of a piece of hardware, it sets the bar for what you can expect in real world applications and how it will perform.

Not sure if troll or just stupid.

avatar

onisuriyuuu

He kinda is a little of both. Obviously I am forced to run at stock froob.

 

My I-5 -2500k Is on a P8H67? forgot the model. And limited to a multiplier of 33 not 55 -> ie: 4-5GHZ. versus my 3.4GHZ.

 

I Dare you to find a way to increase the clock even higher (without turbo) to a multiplier of ... 40's with this board, i will pay you a brand new i7-2300k. 

Note: My clock is even bellow stock clock..... The stock is 34 multiplier, my clock is 33 just a bit under the clock by about .... 50 mhz

 

Its impossible. So its best to know my heat limits when I try to use the most of what my clock has to offer me. And guess what.

My Temp is 55C at all 4 cores at 100 percent usage. ALSO Many times the insane multitasking that I do my CPU usage often does go upbound to 100 percent usage DAILY for hours on end. Why? Tell that to my 80 open tabs, my Photoshop, MMORPG's, ESET Nod 32 Scans, Maleware Byte Scans, Skype (amazingly takes 25% CPU for the insane convo's i get) 

 

ON THE OTHER HAND. My GPU which I had to watch intently Heated up to around 60's Celcius which is highly... dangerous on my part, Because the closer it gets to 80 C the slower it gets, and also the more wear and also 80c is the failure point anyways. SO Benchmarking allowed me to see my Maximum sustained tempature and estimated fluxuations within the tempeture based on ambient tempatrue, it allows me to clock my GPU to get the most out of it without totally blowing it to pieces for running it for longer periods.... Obviously Things don't cool down at 100% it gets hotter and hotter and hotter as the ambient tempature increases from your mini heater of a computer, untill the room gets to a point the computer can hit 80c ..... But that will never happen because I have central Air and my room is well circulated. However If i was in someplace like pueto rico..... my computer is going to be HOT HOT HOT. litterally, and maybe melt a little.

 

i5-2500k, Frio Cooler ( Cold like a fridge :O )

Radeon 5570 (win exp: 6.7 on 7.9 ) Going to upgrade this soon My CPU is cooler than it....

Corsair 800w

8GB Vengance D3 Ram running dual ch 1333mhz (board limited would have been 1600mz)

Segate 1TB 5900rpm Storage Drive.

Samsung Display? 1600x 1200 its... so... clear, and so big. Guessing around just under 30 in

 

And other Misc Accessories Like my Mamba Mouse, and smaller hard drives not note worthy ex 250GB

 

Other than any typo's i might have put or nonsense sentence at 1:00 AM in the morning, please tell me or ask any other questions, and I will be gladly to clarify anything.

avatar

triclops41

Lloyd, can you comment on why you don't linke Kombuster or Furmark for benchmarking?

avatar

onisuriyuuu

Quite honestly His theory is that You can push your GPU too far (obviously the benchmark is made to push your hardware not find a place that feels comfortable) and do some severe permanent damage and possibly void your waranty if pushed out of manufacture regulated Over clocks. ( My Radeon has a factory overclock, however if i go beyond this I can void the waranty )

Log in to MaximumPC directly or log in using Facebook

Forgot your username or password?
Click here for help.

Login with Facebook
Log in using Facebook to share comments and articles easily with your Facebook feed.