CatZilla beta goes public - test it now!

As we reported a couple of weeks ago, CatZilla is a brand-new DirectX11 and OpenGL benchmark developed in Poland. Originally planned to be launched in November, the benchmark has seen a couple of delays. But since yesterday you can download a public beta and perhaps test the new benchmark yourself. Actually, no. You must test this new benchmark! I've been following up closely with the development of this benchmark and as far as I'm concerned this is the best alternative to the Futuremark benchmark series since Aquamark3 launched in 2003! Great graphics, awesome story, multiple subtest, 'awesome' in short.

By the way, as some of you might already know, we are working on direct integration with the CatZilla benchmark. This will allow you to submit your results to HWBOT as well as join CatZilla based HWBOT competitions. Catzilla will come to HWBOT, that's for sure, so you better gear up!

Download link: http://www.allplayer.org/Download/ALLBenchmark.exe


47

Belgium Massman says:

I'm genuinly excited about this new benchmark. Best alternative since Aquamark (and we all know how awesome that one was back in the day!)

United States Bobnova says:

I'm downloading it now. Very, very slowly.

GENiEBEN says:

Massman said: I'm genuinly excited about this new benchmark. Best alternative since Aquamark (and we all know how awesome that one was back in the day!)


More interesting than 3DMARK "i-have-no-version-but-call-me-12" that's for sure.

Australia Dinos22 says:

Ok will test

Italy Crash01ita says:

downloading me too, i will try later

Belgium CornerJack says:

No effect SLI :)

K404 says:

Downloading :)

Australia JJJC says:

Downloading now, looks interesting. :D

United Kingdom borandi says:

Had a small play with it yesterday.

Tiger (1080p):
4333: i7-3930K and GTX560, 1600C9, all stock
6665: A10-5800K and HD7970, 1600C9, all stock
8014: A10-5800K and 2xHD7970, 1600C9, all stock
7214: 2xE5-2690 and HD7970 (925/1125), 1600C11
7901: 2xE5-2690 and HD7970 (1000/1687), 1600C11
8455: 2xE5-2690 and HD7970 (1100/1687), 1600C11
8982: 2xE5-2690 and HD7970 (1200/1687), 1600C11

K404 says:

5870 and I3 2100 (all stock) 72xx 62xx 39xx

Belgium CornerJack says:

Processeur : i7 3970x à 5GHz
Carte mère : Asus R4E
Carte graphique : Asus GTX660 Ti 1100/1800
Kit mémoire : 16GB 2133MHz 9-11-10-28
Système d'exploitation : Win 7
Score : 8768

Processeur : i7 3970x à 5GHz
Carte mère : Asus R4E
Carte graphique : Asus GTX660 Ti 1100/1800 + EVGA GTX660 Ti 1100/1800
Kit mémoire : 16GB 2133MHz 9-11-10-28
Système d'exploitation : Win 7
Score : 8724 ???

Processeur : i7 3970x à 5GHz
Carte mère : Asus R4E
Carte graphique : Gigabyte 680GTX OC 1200/1800 4GB
Kit mémoire : 16GB 2133MHz 9-11-10-28
Système d'exploitation : Win 7
Score : 10295

K404 says:

Maybe it would be good to have a good GPU benchmark that doesn't scale with SLI/CF. Maybe this could become a GPU benchers "Spi32M?"

United Kingdom borandi says:

K404 said: Maybe it would be good to have a good GPU benchmark that doesn't scale with SLI/CF. Maybe this could become a GPU benchers "Spi32M?"


I could make a GPU 2D benchmark that doesn't use multi GPUs :)

Belgium FunSkilZZ says:

downloading :P actually pretty fast

Colombia saint19 says:

Donwloading. Could anyone test it with a i5 or i7 and GTX 680? I will test this with GTX 680 and AMD to see if CPU is important on this...

United States Bobnova says:

Tiger, 3770k@4ghz, GBT 670 OC at 980/1500, RAM at 1600-8-8-8: 9801
Tiger, 3770k@4ghz, GBT 670 OC at 1050/1750, RAM at 1600-8-8-8: 9878
Tiger, 3770k@4ghz, GBT 670 OC at 1086/1750, RAM at 1600-8-8-8: 10052

I love the benchmark. Decently long continuous load, amusing, pretty, and a physics test with halfway decent physics instead of the 1990s travesty that is the 3d11 "physics" test. And the whole thing doesn't take three days, like Heaven for instance.

I would pay for this.

United Kingdom topdog says:

It's interesting to see how much the loading time affects the final score

Here I had a 5870 and a 2600K at 4.5ghz running at the same settings, ran three times




http://img.photobucket.com/albums/v133/scoobytopdog/loadingtime2_zps8af9c7b9.jpg

United States Gunslinger says:

slowest































download








































ever

Belgium CornerJack says:

Score very strange...

Colombia saint19 says:

CornerJack said: Score very strange...




That second test i7-377k and GTX 680 Lightning is what I need to compared with 1090T and same GPU model...

Could you run it with CPU at 3.8GHz? I will try to set the GPU at the same clock for the test.

GENiEBEN says:

Gunslinger said: slowest
download
ever


U

S

I

S

P

:)


Took 98sec.

says:

Bobnova said: I'm downloading it now.
Very, very slowly.


seems alright on my end


Netherlands rsnubje says:

Multi gpu detection isn't quite right.
http://www.allbenchmark.com/component/comprofiler/userslist/Single%20Card%20Tiger%20Results?Itemid=312

I'm in second place here, but is ran with 3x HD 7970 lol. Owyeah, if your SLI/CF doesn't work, try AFR(2)(friendly) mode;)

United Kingdom borandi says:

The loading time difference is interesting. Someone install to Ramdisk and see if that does anything?

Loading time will either be SSD -> Mem (dependant on SATA bus) or Mem -> GPU (dependant on PCIe bus)

Also I was able to run the 1080p Tiger mode with a 1440x900 monitor.

United States Bobnova says:

RAMDisk has no meaningful effect on loading time score. 7.5s with a V3 240GB MI or with a 2GB RAMdisk. Interestingly it loaded in 4.2s a couple of times earlier this morning, other than those it's been 7.5s every time.

Belgium Massman says:

As far as I remember, the developer said this was more of a system benchmark than 3dmark is. Loading times, multi core, frequency, memory, vga ... it's all important for the final score apparently. Interesting, huh :)

says:

im not sure i find it interesting if it's a hard drive bench as well as 3D bench.....

United Kingdom borandi says:

Team.AU said: im not sure i find it interesting if it's a hard drive bench as well as 3D bench.....


Same here.

Given that RAMDisk did nothing, it does mean it's more a PCIe bus thing. This means that BCLK comes more into play. 106*50 > 100*53.

But I won't like the fact that if one run out of three loads correctly, that's the better score.

Belgium Massman says:

Team.AU said: im not sure i find it interesting if it's a hard drive bench as well as 3D bench.....


I don't think it's really a hard drive benchmark. It's taking the loading times into account for the final score, reflecting how slow loading times could affect your gaming experience. As in: if you're still using that old 5400rpm hdd, you're gonna have a bad gaming experience.

It doesn't scale with ramdisk, it seems.

United States Bobnova says:

I have a mechanical HDD with an identical windows install around here somewhere, I'll try it later tonight and see what kind of times it manages. Loading time might be partly decompressing textures into RAM or something too, hard to say what's really happening. Loading doesn't peg the HDD activity light either.

zeropluszero says:

just downloaded and installing now. should be good for something new.

Canada Vinster says:

Downloded it, will install later and test. thanks Vin

Canada Vinster says:

Here are my findings and tests.

Bench with 2xGTX580 in SLI, PhysX set to Auto



Bench with 2xGTX580 with SLI Off, PhysX Set to Auto



Bench with 2xGTX580 With SLI Off and GPU2 Set as Dedicated PhysX



Bench with 2xGTX580 with SLI Off and Physx Disabled



The other thing is I was able to get my GPU OC much higher in this bench then I could in 3DM11 or Heaven... so I don't think this bench is as taxing on the system as other benches that are comparable.

Here is the catzilla test... SLI Enabled, PhysX on Auto.



Vin

Netherlands rsnubje says:

Looks like SLI isn't working well because your single score is almost the same? Maybe another SLI mode?

rsnubje said: Multi gpu detection isn't quite right. http://www.allbenchmark.com/component/comprofiler/userslist/Single%20Card%20Tiger%20Results?Itemid=312 I'm in second place here, but is ran with 3x HD 7970 lol. Owyeah, if your SLI/CF doesn't work, try AFR(2)(friendly) mode;)
Still think of this benchmark as more for 3D instead of the whole system tho.

United States Johnksss says:

Same here, just started downloading. Says 6 hours left...Oh goodie. "facepalm"

Bulgaria I.nfraR.ed says:

Was downloading with 4.0+ MB/s yesterday, but I have 75Mbit/45Mbit download/upload.

Canada Vinster says:

I have a 15Mbit/1Mbit connection and I got it at 1m/sec... so it was fine for me.

I found that in SLI the bench used my 2nd GPU and not my 1st GPU, you can see that from my screenshots in the MSI chart, then once I disabled SLI the bench used my 1st GPU...

and In SLI the bench doesn't use the 1st GPU at all in my case... GPU usage stayed at 0%-1%..

Vin

EDIT: any plan on addressing the SLI/xfire issue? I can't help but think of the peeps that have a GTX690/590/295 and the AMD guys with their x2 cards... This bench does no justice to them...

United States cowgut says:

afr2 is the deal



Netherlands rsnubje says:

I downloaded the update last week but after that crossfire wouldn't work anymore. The benchmark simply crashes with more than 1 videocard.

Belgium Massman says:

Anyone else having issues getting 4Way SLI to run properly? I'm getting about 6K Catzilla with 4x 580 ... :-/

United States Johnksss says:








Canada Vinster says:

SLI is non functional for me... 2x580 gets me the 37xx catzilla score above and you can see in the GPU usage on GPU#1 is non existant.

I downloaded the newest version and still no difference. Using 306.97 drivers. will try to update to 310 in the net day or so... but frustrating seeing other people with near same HW get double my score.

Vin

United States Johnksss says:

you need to set it to afr2 under nvidia control panel.

Belgium Massman says:

AFR2 is enabled, but the benchmark doesn't seem to scale past 2 GPUs here ...

United States Johnksss says:

Massman said: AFR2 is enabled, but the benchmark doesn't seem to scale past 2 GPUs here ...


you may need to set the one for 4 gpus..and would need to use nvidiainspector for that...speculation of course.

Canada Vinster says:

johnksss@say-it-loud.com said: you need to set it to afr2 under nvidia control panel.


Thank you that did the trick for me... see new screenies... :)




Vin

Colombia saint19 says:

Here are two results, both with the following at stock settings on GPU.

Please log in or register to comment.

Leave a Reply: (BBCODE allowed: [B], [QUOTE], [I], [URL], [IMG],...)