GeForce Benchmarks On The SteamOS Beta

H

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
The alternative OS gurus at Phoronix have benchmarked the SteamOC beta using a handful of GeForce video cards today. Definitely worth the click to check out.

A comprehensive performance comparison is underway at Phoronix that pits SteamOS against other desktop Linux distributions, but for those anxious to see some performance numbers, here are benchmarks done so far this weekend from seven NVIDIA GeForce graphics cards on the public SteamOS 1.0 Beta operating system.
Click to expand...
 
Still pretty much useless considering the lack of games for Linux.

Call me when I can load up COD or Batman.
 
It's nice that they have benchmarks, but useless because they don't compare it to an identical benchmark on Windows or other flavors of Linux.
 
Am I the only one wondering WTF that guy is wearing in that picture on page 2? :confused:
 
chiablo said:
It's nice that they have benchmarks, but useless because they don't compare it to an identical benchmark on Windows or other flavors of Linux.
Click to expand...

Word. I'm at work and won't have time to come back to this. I skipped to the conclusion, which said nothing, so went to see the benchmarks, and got absolutely nothing out of the whole article.
 
Yeah I am equally disappointed in the fact there is no comparisons to....anything really.
 
King Icewind said:
But it looks like the whole 500GB requirement is garbage? They used a 160GB hdd.
Click to expand...
Pretty sure the 500GB requirement is just to make sure everyone has a decent user experience.

Not only does it insure everyone has enough space for plenty of game installs, it helps assure a certain level of performance. Someone trying to run SteamOS and modern games off an old 160GB IDE HDD they just happened to have kicking around is NOT going to be as fast as any 500GB HDD.
 
chiablo said:
It's nice that they have benchmarks, but useless because they don't compare it to an identical benchmark on Windows or other flavors of Linux.
Click to expand...

That was my thought exactly!


King Icewind said:
I was hoping for game benchmarks compared between the SteamOS and Windows.
Click to expand...

Yeah, that's what I expected as well. I mean without any sot of baseline comparison the numbers loose much of their meaning.
 
I'm going to echo everyone here and say that no comparison = waste of time. Kyle & Co. are you guys gonna try and load it up?
 
I took the extremely easy and basic work it takes to look up other unigine results, ya lazy bums. The results are essentially 1:1 with windows.

As well, the article mentions that their results were very similar to the results from other linux builds they tried TF2 with when that first came out. Looking back at those, TF2 fps was higher in linux compared to windows (not a huge amount, but it was there), so there's that too.
 
Oh man, this will be the last time I'll ever look phoronix "benchmarks". Talk about useless.

Also, that svg on page 3 with OpenArena makes no sense. The game should run with 1k fps at a minimum. Not that the q3 engine doesn't brake at ~900 but oh well.
 
lilbabycat said:
I took the extremely easy and basic work it takes to look up other unigine results, ya lazy bums. The results are essentially 1:1 with windows.

As well, the article mentions that their results were very similar to the results from other linux builds they tried TF2 with when that first came out. Looking back at those, TF2 fps was higher in linux compared to windows (not a huge amount, but it was there), so there's that too.
Click to expand...
Problem with that is that they way they port it to linux is though an openGL interpreter although one would think it would be slower it's porting it into a newer version of opengl. So you have DirectX9 games with huge overhead vs newer opengl with little overhead. If they ported the game into DX11 probably get even more fps but they don't. Similar issues come with steams comparison they often use a 32 bit version of linux but a 64 bit version of windows which is an oddity.
 
I read it earlier. One 8 year old game (Doom 3), one sucky game (Xonotic) and a bunch of useless synthetic benchmarks.

What is the point of it being Steam OS if he's not going to benchmark any of the 200+ Steam titles which run under Linux and Windows? Instead, it's a couple of lame timedemos of 2 games virtually no one plays.
 
The Avalon said:
Thanks for your valuable (actually worthless) contribution to this thread.
Click to expand...

Actually I thought he summed up the situation rather succinctly, wtf did you contribute?

Unknown-One said:
Someone trying to run SteamOS and modern games off an old 160GB IDE HDD they just happened to have kicking around is NOT going to be as fast as any 500GB HDD.
Click to expand...

I don't really see how old IDE drives come into play. In these benchmarks it was a 150GB Raptor drive, not an old IDE drive. Most people with a <500GB drive these days are going to be using an SSD.

Someone choosing between something like a 256GB SSD or a 500GB Hard Drive is going to be making a legit choice, and hopefully no one would be telling them to choose the 500GB drive in that case due to it being "faster"...
 
GotNoRice said:
I don't really see how old IDE drives come into play. In these benchmarks it was a 150GB Raptor drive, not an old IDE drive. Most people with a <500GB drive these days are going to be using an SSD.
Click to expand...
He mentioned 160GB drives. More people have old 160GB drives kicking around than 150GB raptors or SSDs :rolleyes:
 
Unknown-One said:
He mentioned 160GB drives.
Click to expand...

The person you quoted said that they used a 160GB drive in the article, but they didn't. They used a 150GB Raptor, as clearly listed in the specs of their benchmark system on page 2. Nowhere in the article is anything mentioned about a 160GB drive.

Unknown-One said:
More people have old 160GB drives kicking around than 150GB raptors or SSDs :rolleyes:
Click to expand...

Even if that's true, it still doesn't support the blanket statement you tried to make about the 500gb requirement being about a minimum level of performance. I would contend that despite the number of creaky old IDE drives out there, SSDs would still be more relevant in that capacity range for anyone considering building any sort of relevant system today.
 
So people looking for a good comparison bench might be a bit disappointed. I fired up the linux version of Metro LL on the OS a few minutes ago and it ran fine, but the game has zero options (resolution included) beyond a a single vague slider for "Quality" which at the highest level doesn't appear at first glance to look as good as the Windows version. You can't even run stuff in full range RGB yet so say hello to black crushing and wash out for the time being.
 
You must log in or register to reply here.
Back
Top