Cinebench

[stasio]

CINEBENCH 11.5 is released.

New Enhancements Provide Latest Technology to Test CPU and Graphics Card Performance Under Real-World Circumstances.

[Raiden]

And results with it?

Post

[stasio]

And results with it?

Post

Look at Gulftown bench.

[Particle]

Awesome. This really ought to shine under high-core systems.

[KILLER_K]

Good deal as i have been waiting for a new one for a while now. Thanks for the post.

[jcool]

Yes, I already saw this yesterday...finally! Here is a sample score, scales really well with cores and especially HT.

[SimpleTECH]

Here's what I got:

[mattkosem]

This version seems to benefit more from hyperthreading than 10 did.

--Matt

[mAJORD]

hmmm but can a Program really benefit from HT? :P

Depends what angle you look at it I guess hehe

Anyway great to see a new one. You should make it a Ranking thread Stasio!

[mattkosem]

hmmm but can a Program really benefit from HT? :P

Depends what angle you look at it I guess hehe

Anyway great to see a new one. You should make it a Ranking thread Stasio!

That's a silly question. Of course it can. Any multithreaded application that is prone to thread stalls will be helped by hyperthreading. I'm not sure that the angle at which you look at it matters at all. Run the test once with HT on and once with HT off at the same clock and you'll see the difference

--Matt

[mAJORD]

That's a silly question. Of course it can. Any multithreaded application that is prone to thread stalls will be helped by hyperthreading. I'm not sure that the angle at which you look at it matters at all. Run the test once with HT on and once with HT off at the same clock and you'll see the difference

--Matt

Right, HT makes a big difference to the performance.

How do you purposely program to take more advantage of Hyperthreading specifically? I'm not a programmer so this isn't a statement, more how I understand it:

Any program has a purpose in life, and if you were trying to make it more efficient, regardless of the processor, you'd want it to make full use of a core, for each of its threads.. Obviously it's impossible to make any program use 100% of a non SMT cores resources, so hyperthreading is almost always a benefit to the end user..

So my understanding (wrong or not) is The benefit from hyperthreading is a side effect of the type of program. Perhaps this version scales better 8 threads? be it HT or otherwise.

..

[V I P E R]

Here is one quick bench: 920 @ 4410Mhz

[-=DVS=-]

It does have some quirks doesn't detect processor speed correctly shows 4.1ghz and doesn't make use of SLI.

[=[PULSAR]=]

Hmmm I'm only getting 6.70 with i7 920 @ 4.15ghz???

[Sam_oslo]

@4.22GHz(222x19) 7.20 pts
@4.51GHz(215x21) 7.70 pts

[WeldZilla]

My 260GTX when clocked up and running in conjunction with a fast processor aquits itself nicely.
But when it comes to cpus this proves out the old saying, "The more the merrier"



WZ

I was only running at 4500mhz on the cpu and my memory was at 1200mhz 5-5-5-6

[shemer77]

gonna dwnld this and post my results. I should do ok.

[mAJORD]

Some non overclocked reuslts on "lesser" CPU's

Core i5 750 @ 2.8Ghz (Turbo) 3.72
Core i3 530 @ 2.93Ghz: 2.43
Phenom II X4 965@ 3.4Ghz: 3.99
Athlon II X4 630 @ 2.8Ghz: 3.20

[I.nfraR.ed]

With my 965. I expected to do around 4.4, judging from the results from my Asus board. Anyway...


And one OpenGL result with 24/7 clock on the cpu. It seems that it doesn't like ATi cards, or they have poor OpenGL performance.

[Tiltevros]

here is mine

[mattkosem]

Here's what I get at 4.4 with HT on my 920.

--Matt

[Gaul]

I'm in..........

[doonsy]

Here is what I get on an old Q6700 @ 3.6 and a stock 5870

[Der_KHAN]

Here is what I get on an old Q6700 @ 3.6 and a stock 5870

strange, i get 4.03 with my Q6600 also at 3.6ghz. looks like one of your cores was loaded with sth else

[doonsy]

strange, i get 4.03 with my Q6600 also at 3.6ghz. looks like one of your cores was loaded with sth else

You were right, I killed off a ton of background processes etc.