Processor GFlops Compilation

[nascasho]

I hate my life.

And yes, w/o pumping soooo much Vcore, I couldn't get it stable at 3.6Ghz for the life of me (Was at 1.36V non-stable) I would swap boards... but SLI...

Right now I like these clocks because I'm running on stock Volts, 1.25.

The RAM I keep synced with the FSB. They're both rated for 800Mhz each.

 

[hoss331]

Yea if you want to keep SLI your kinda limited, but better ram should help though.

 

[nascasho]

How does GFlops work in correlation with like Gaming. Lets say a game shined at 3.4Ghz, I'm seeing normal benches, but you guys must see much more since your CPU can push out more?

GFlops is the amount of info a CPU can pop out?

I'm confused here.

Sry for being a noob.

 

[FordGT90Concept]

FlOPS weigh in heavily in terms of player positions, collision detection, and physics.

FlOPS = FLoating point Operations Per Second

A floating point basically means the decimal position is important (e.g. 1 is just an integer while 1.1 is a floating point value). An operation is like multiplying two floating point values together. A GFLOP (billion floating point operations per second), for instance, means the processor could multiply two floating point values together one billion times in a second.


FLOPS are most important in terms of scientific work. Usually the graphics card will fall short before the CPU.

 

[nascasho]

So low GFlops would affect performance in gaming?

I'm about to fling this thing out the window because I have no idea why I'm getting such a low score... Hell, I don't even notice any difference.

 

[FordGT90Concept]

So low GFlops would affect performance in gaming?

I'm about to fling this thing out the window because I have no idea why I'm getting such a low score... Hell, I don't even notice any difference.

Yeah. I mean, you wouldn't want to play a game on an Atom. XD

My Core i7 920 only got 26.6 average. A dual-core E8600 beat it, albeit with a clockspeed almost double mine...

The way I see it, Linpack is mostly meant for stability, not benchmarking (unless you are benchmarking stability).

 

[nascasho]

Wow, makes sense...

I guess 3.4Ghz is equiv to your i7 920 stock considering I think they have about 40% diffence at the same clocks...

I get 27.9ish

 

[lemonadesoda]

hoss331 | Q9650 @ 4.25GHz | 59.4010

http://i131.photobucket.com/albums/p293/hoss281/Untitledc.jpg

Great results! But bug in your stats. Check your MHz. No way a Q9650 can pull 59.4010 GFlops @ 4.25GHz. I think your clock must be higher.

 

[Melvis]

Add Me

Melvis | AMD | FX-57 @ 2.8GHz | 01.6395 | 0.5855357 | 32-bit | Old School

Thanks for updating mine

 

[rflair]

Why are some people only doing 4 runs, when the test defaults to 5 and the original Ops message was also defaulting to 5?

I mean the difference is under 1 minute to run 5 (default) instead of 4.

Lets keep consistency.

 

[Darknova]

Why are some people only doing 4 runs, when the test defaults to 5 and the original Ops message was also defaulting to 5?

I mean the difference is under 1 minute to run 5 (default) instead of 4.

Lets keep consistency.

Run the IntelBurnTest(yes, it works on AMDs), set it to run at least 4 times and run the standard stress level.

and newtekies first run was 4 times.

 

[rflair]

and newtekies first run was 4 times.

Opps my mistake. Was just looking at the pretty pictures.

 

[Darknova]

Opps my mistake. Was just looking at the pretty pictures.

I know some people have done 5 runs, and newtekie did say "at least 4 runs", more just gives a more accurate reading.

 

[yogurt_21]

Here we go... my 4.6GHz bench settings. I never intended it to be Linpack stable:

infrared | Intel E8400 @ 4.6GHz | 31.67638

http://img.techpowerup.org/090519/Capture051.png

I need a better cooler for this kinda thing! ^^

he's Alive! it's a miracle!

seriously where you been?

nice overclock there, but they again you always were able to push a cpu to it's max. (5GHZ p4)

 

[hoss331]

Great results! But bug in your stats. Check your MHz. No way a Q9650 can pull 59.4010 GFlops @ 4.25GHz. I think your clock must be higher.

WTH, I missed a number, supposed to be 4.425

Thanks

 

[newtekie1]

Table Update

I'd like to see some Athlon X2 results, and maybe we can get some people from the low cache overclocking club to post some results here, I'd really like to get an idea how cache affects this test.

I can already tell the memory bandwidth drastically affects performance. Time to up my RAM speed and see what I can do.

Why are some people only doing 4 runs, when the test defaults to 5 and the original Ops message was also defaulting to 5?

I mean the difference is under 1 minute to run 5 (default) instead of 4.

Lets keep consistency.

I said at least 4 runs, instead of 5, for people running older setups. It takes longer and longer to run the weaker the processor. Look at my Atom run, it would have taken another 45+ Minutes to do one more run, and it had already taken almost 3 and a half hours just to run the 4. Doing more runs just adds accuracy to the scores since we are taking the average of all the runs.

4 was the number I decided on that would still ensure pretty good stability, yet not make the test a total drag to run on lower end hardware.

 

[rflair]

rflair | AMD | Opteron 146 @ 2.8GHz | 4.305

This was run with only 512megs in the test, this PC only has 1 gig and is running Win7 x64.

 

[newtekie1]

Add Me

newtekie1 | Intel | Celeron 325 @ 2.53GHz | 02.6307

 

[lemonadesoda]

^^ You need a "zero" in front of that score!

 

[Naekuh]

Naekuh | i7 975 D0 @ 4.3ghz HT OFF | 59 ish im too lazy to bust out calculator.

Does HT on even matter?

The old LinX didnt even use HT threads, but i could run it with HT on.

Erk.. old cpu-z tho, looks like it cant read my D0 properly.

 

[CH@NO]

add me

CH@NO | e8200 @ 3.8GHz | 23.1021

 

[lemonadesoda]

Correct Me

Please edit/correct my old entry #7 with the following data:

lemonadesoda | 2x Xeon E5420 @ 2.50 GHz | 58.67 | CPU at stock, FB-DIMMs at 667

 

[newtekie1]

Table Updated

 

[Naekuh]

actually i need another update.

Guys with i7.. DISABLE HT ON.

HT ON makes your results worse.

 

[DanishDevil]

Add Me

DanishDevil | Intel | Intel Xeon X3210 @ 2.13GHz (Stock) | 26.95774