G1S 8600M GT vs F3SV-B1 8600M GS Benchmark

Wow, at 1680x1050 with 4xAA and 8xAF and max settings you could still get ~20fps? Just drop it to 1440x900 with 0xAA and 4xAF and you could probably have 50-60fps solid.

 

It is NOT based on the newly unveiled iD engine. It is running on the Doom3 engine.

 

Yeah, a lil lower settings with lower AF and AA would go a long way in boosting those FPS..

 

Ok thanks for clearing that up I was hoping for GDDR3, but hey 512 of DDR2+TC isn't bad

 

Asus boasts 3dMark scores of 4680 for it, without overclocking.

 

FYI on S96S benchmark and Quake Wars

Setup:
T7300
3GB RAM
160GB 5400rpm
GeForce 8600M GS 256MB Stock driver no overclock

3DMark06 ORB

 

Visuals look good, fps is abit low.

I dont think personally it will get much more optimized for thefull release, but I think you will have to turn down settings a notch when the game is public because in online fps games its not the on scene stuff thats really hard its when you get all the other players in your face, it adds alot more stress to the gpu and cpu to render. In all of those shots there really isnt much going on.

Id bet it will drop down alot if you have a group fight with 3+ people on screen and others around you.

While im at it ken sort of off topic. Any word on selling c90s with a AMD 2600m in them instead of the 8600gt?? Id still have to wait for more info on the card but I have a feeling it may be stronger.

 

Hey guys, been running my F3SV-B1 with a 8600M GS gor a week now, and so far it can run any game on in its highest state.. from 1680 x 1050 on down to 1440 x 1050. running 60+ FPS on up...

The card runs wonders around the 7600 GO and smashes the 7400 GO. I have been playing CnC3 and BF 2142 a bunch and have just recently installed and played the Quake Wars ET last night.

If you don't have the extra cash for the G1S the 8600M GS will due, no matter what the 3DMark06 scores say. Live play and real world testing have proven to me that true gameplay always trumps what 3DMark tells us.

 

They will be only focus on nVidia right now.

 

Two days before reading this I bought a S96S (8600M GS) from GenTech...And this post makes me feel absolutely terrible for not reading up on GS vs. GT lol. Especially since I got a quote on the new 8600M GT Sager for just $150 more. Absolutely terrible!

 

I own the A1, and trust me, you aren't getting consistently 60+ fps. You are getting more like 20-25 fps.

 

I disagree.. how much ram you running? I can show video of my F3SV-B1 running these rates...

 

Yup, I agree with Gautam. It would be hard to get consistent 60+ fps on games like C&C3 on native resolution let alone Quake wars!!

 

Maybe with AA and AF turned all the way off.

 

Hmm, that still wont make it i think..

Anyhow, unless Sledgehammer70 posts some frap screenshots, i aint gonna believe it!

 

I guess the better question is, how do you know you are getting 60 FPS Sledge? Are you just guessing? Above 30 FPS the human brain can no longer percieve any difference, so if you are assuming, keep that in mind.

 

AlexOnFyre, you are oh so wrong and it can be found by googling... The human eye has been known to view up to over 120 FPS not 30. The main problem is if you have a CRT monitor you can only be shown as many frames as your refresh rate runs, having a LCD allows you to see more as it is hinged more on Response time rather than refresh rate. So having 30 FPS maybe good enough but in most cases the human eye can see and as you put it (perceive) more than 30 FPS. This is a proven fact by the US army and many eye doctors around the world.

Also for CnC3 I am running 1440 x 1050 and am achieving roughly 62 FPS average with no AA...

In CSS Test is set to High at 1680 x 1050 and I am achieving roughly 63.23 FPS average.

http://www.bttournaments.com/gaming/CSsource.jpg
http://www.bttournaments.com/gaming/CSsource1.jpg
http://www.bttournaments.com/gaming/CSsource2.jpg
http://www.bttournaments.com/gaming/CSsource3.jpg

will run some more screens tomorrow as I am off to bed

Also here are the full system spec's:
Intel Core 2 Duo T7500 @ 2.20Ghz
Kingston DDR2 800 PC2 6400 2 x 1GB
Hard Drive 160GB 7200RPM SATA
Nvidia - 8600m GS 256MB 767MB Shared (Total: 1023MB)
http://www.bttournaments.com/gaming/GPU.jpg
Power set to 100% no to sure what else to post?

I have no reason to lie just posting my results and what I am seeing with my systems performance.

 

I didn't say the eye, I said the brain. You can see almost 300 FPS, but you can't tell the difference between framerates (fluidity) after 30 FPS. The only differences you will percieve is the headache you get over 100 FPS (which most monitors don't do.) Unless it is absolutely fluid, meaning real life, you will get a headache with in a few hours above 120FPS.

But this isn't about biology, it is about the F3SV, and while I would like to believe you, it is hard to believe it outperforms the G1s (which is shown to get 25-30 FPS on CnC3 with AA and AF down and at Mid settings.)

 

Yeah at 1680 x 1050 with mid AA & AF it would be hard... but I am running 1440 x 1050 with no AA and no AF... you will see a huge performance increase with these settings...

not to sure how I can prove it if you won't accept direct screenshots from the game...

 

You said "highest settings...up to native res" which most assume to mean full AA and AF (or at least some) as well as native res included (not excluded in some cases, as we now know).

That all known, it is pretty feasible to get over 30, even up to 60 (maybe), but over 60 I still doubt. Though with those settings and getting over 30 is still great!

 

I realize my benches are only on 1GB of ram, but I considered that before I wrote a response to your post. Even if you tack on another 1GB, and get dual channel, let's be liberal about it and say you get 2x FPS (although that's not what happens). That means 20x2 = 40. So you'll get around 40-45FPS, not 60+ like you quote.

Thanks. You saw the keyword of "consistent" referring to average. I don't doubt that sledge might get 60+ on a load screen or while booting into a game, or even during a spike in gameplay. But that isn't consistent, it's just a statistical outlier.

Perhaps 40-50, but not consistently 60+ on average. Remember that sledgehammer was saying he could play games "at the highest" state, so we need to assume some AA (2x?) and maybe AF at HD resolutions.

Sounds improbable.

 

I thought Command & Conquer 3 had a framerate cap@30. Is that the cap has been removed in the latest patch?

http://www.xbitlabs.com/articles/video/display/gf8800gtx-roundup_10.html

 

I am afraid you are arguing on the totally wrong side here. The higher the FPS, the more fluid the motion is, and lesser the strain on the eye.
If you want an in-depth article, read this:
http://amo.net/NT/02-21-01FPS.html
Thank meaker for the link.

 

Once again I am only posting what my system is showing and telling me... I have no reason to lie to anyone... I will post more screenshots with my findings...

 

honestly, it doesn't have to do with the mechanics of the eyeball. In between 120FPS and lifelike motion (tending towards infinite frames per second, or 300FPS for the eye (approx.)) there is a phenomenon called staggering, where the brain is filling in an image (predicting what happened in between seeing one frame and the next) and is interupted by another signal from the eyes. It is sort of like tearing in real life, but with no visual after effects. You simply start getting a headache from conflicting actions going on in your brain.
In reality, there are no frames per second, your eyes and brain interpret light as it shows up, but the sodium-sodium chain can only move so fast, and when it is broken by a preemptive signal (another flash of light before being done with the first.) you have some extra floating around that eventually cause a headache to get you to stop looking and let the brain even itself out. Watching a series of light flashes works differently than looking around in the real world. In addition it is a condition which humans have only lived with for less than 100 years. So most likely any information we have now is based on limited data sets and may end up being entirely false. I am basing my assumptions on a sister lab to my own at Georgia Tech (Psychological Optics lab), which concluded that above 30 FPS, the human brain can no longer accurately gauge FPS (within 10 frames) when there is motion blur. Above 60 it can no longer gauge even without motion blur. In any case, here is the best site I have read on the matter.

 

Wow, that is a lot of interesting info. I hear/read about these kinds of things all the time, but they always have differing opinions.

 

Here is a stupid question that I sorta need answered. can I install any mac OS on a home built computer (not mac)?

 

Simple answer is No. There was a project a while back to get it to run, but it is just a big hassel and doesn't work too great either. I also think the project was shut down.

 

Not legally. Let's say that I, hypothetically, have done this. There might be issues with the w-lan (intel 3945abg is not supported). Also there might be problems with the sound (which is somewhat easily fixed). And if you have a 512MB go7700 (or almost any nVidia card), you're pretty screwed. Hypothetically speaking of course.

 

Nah, it's still up. And for some (mostly desktops) it worked quite well. Someone did, however, hack their site and it's down presently.

 

Well the project always lives on, but for a period of time the Darwin project was receiving funding from a major company whose name I absolutely cannot remember.

 

AAH. Wait, what? It's called the OSx86-project. Or are we talking about the same thing, but with different names?

 

The fact of the matter is if the FPS on a laptop shows and 30 FPS and than the same scene at 100 FPS you can clearly see the difference regardless of what your saying the brain can do. If what your stating is true than 30 FPS should be good enough no matter what, but in general that is a false statement as most games will either not run smoothly or just look like crap at 30 FPS.

Take my 7800GT desktop and run a game at full settings at 1680 x 1050 in a standard game getting me 45 FPS average. now throw my 8800Ultra in the same system and run the same game at 90 FPS the difference is clearly seen as a night and day difference. My brain must be processing something more right? or is it an illusion?

Until I see some concrete facts or findings that the brain can only break down 30 FPS I am going to call Bull.

Some more information on this can be found here, based on hardware and the human interface. mostly hardware side of things...

http://www.tweakguides.com/Graphics_1.html

 

Ah, I just googled it, originally it was called Darwin, after funding was cut, it changed to the OSX86 project, so I was wrong. The project is still running and there are still OSx86 images floating around the web. Thanks for clearing this up!

 

Yup, and it's not merely old one floating around, it's new versions and new versions of the kernel (which is legal to tinker with i believe).

 

And until I see FRAPS screens with the stats you mention I am going to call bull, so I guess that makes us even.

 

I'll have to try the new 10.6 version, I remember wanting to do it a year or so ago, but my CPU(Atholon 3000+) at that time didn't have SSE3.

 

You mean 10.5 (leapord). It's not really released yet, so you'll have to wait til october at least.

 

Installing Fraps Now... This will be fun showing you how I am just reporting what I am seeing...

 

CSSource Fraps 70FPS

http://www.bttournaments.com/gaming/CSsourcefraps.jpg

 

It isn't CS that I doubt, you already showed us that. It is CnC and FEAR.

 

that CSS pic, AA/AF is off for sure but what about the other in game settings? all high? (See the floor and those woods)

 

Call of Juarez SP Demo DX10, Thanks to AlexOnFyre for File Planet download account.

Setup:
G1S-A1
stock driver no overclock
3GB RAM
160GB 5400 on eSATA

 

Looks excellent and thankyou very much alex and ken for this.

Request: Can you pull the settings down so that you get a decent framerate for playing it and then take screen shots in the same area. I want to compare what image quality we lose by downing the overall quality.

 

Sure, I'll find a playable settings and post pictures later.......or tomorrow.

 

Call of Juarez for some reason makes me think "oregon trail 2008" lol

I remember the original back in the day, then I have a new version that came with my first computer that was modernized that looks alot like this.

We need sombody to start on a patch right away, so we can go buffilo hunt'n and ride the trails once again


for those who need a visual reminder

 

I want FRAPS on those Vicious!

EDIT: Also, I think we could all stand to take a hit in "Foliage Quality" even though Call of Jaurez is this year's leading candidate for "Best Shrubbery Rendering in an Action Game"

 

I am with Alex here, it's not that you DON'T get 70+ on FEAR, Lost Planet, Halo 3, heck, even games that are in production a year from being made whatever at highest settings, it's just that it's a SPIKE, and not an average. Spikes are of no use for our discussion.

 

An avg 60 fps does not tell the whole story, the game could go as low as 5 fps at one point and 150 fps at the other point. Anyway, consistent 60 fps means the framerate should never drop below 60..so avg 60 fps is 100% =/= consistent 60 fps.

Consistent 60 fps = minimum fps greater or equal to 60 fps.

That HL2 ss does not really look like at highest settings...and I want to know how he gets 62 fps in CNC3 because the game is cap @ 30 fps max.

 

Thats why all in game benchmarks should be like FEARS built in one.

It tells min fps, max fps, and average fps.

The ontop of that it tells you a % of what FPS was in what lvl

like 0% fps below 35
12 % fps below 50
82% fps above 50

with that data you can very clearly tell just how well the game is running and how much it "spikes" into a high or low fps.