Titan Black ?!

[PaddieMayne]

So more bloody confusion !?

Also i tell you what annoys me about there test rig, is the fact its unrealistic, how many gamers are going to have a system spec like that, they should test it on a more common set up like a 4670k/4770k and maybe throw an Ocd FX8350 rig in there for comparison, these are the rigs most of my friends online have which constitutes about 60 - 80 people in our gaming clan for Arma 3.

 

[barnsley]

To be honest I'm always skeptical of Tom's benchmarks but they come out in some serious bulk so yeah always like using them as a starting point.

Except that 7990 score is glaringly wrong. I'll see if I can get my mate to benchmark Arma 3 on his rig (he has a couple of 280x in crossfire and a i7). I definatly know he gets stupidly higher frames then that.
Tom's hardware seem to have a thing against AMD from what I've seen in the past.

So more bloody confusion !?

Also i tell you what annoys me about there test rig, is the fact its unrealistic, how many gamers are going to have a system spec like that, they should test it on a more common set up like a 4670k/4770k and maybe throw an Ocd FX8350 rig in there for comparison, these are the rigs most of my friends online have which constitutes about 60 - 80 people in our gaming clan for Arma 3.

Its because they want to show the 'raw' power of the gpus and that cpu wouldn't hold the gpus back one bit.
Just get another 780, you should be golden although the game isn't exactly your CPU's friend things should improve.

 

[SeekaX]

Except that 7990 score is glaringly wrong. I'll see if I can get my mate to benchmark Arma 3 on his rig (he has a couple of 280x in crossfire and a i7). I definatly know he gets stupidly higher frames then that.
Tom's hardware seem to have a thing against AMD from what I've seen in the past.

Wouldn't say that, they've been pretty happy with the 290x on release.

 

[PaddieMayne]

OK im gonna stick with my FX9590 as at 5ghz its not going to be far behind intel, although id be tempted to go 4930k and 2011 board its alot of money i think for not much more performance over what ive got, so im just going to get the extra gtx780 and sli them, ill let you guys know the end results in Arma 3.

 

[ahimoth]

There is a 7990 in the benchmark result, but the score is odd.

Oh my god lol! There is! I just didn't think it could be so low, lol. I thought it was odd that there was a 690 vs 7990 on everything else except this benchmark. Haha ROFL.

Except that 7990 score is glaringly wrong. I'll see if I can get my mate to benchmark Arma 3 on his rig (he has a couple of 280x in crossfire and a i7). I definatly know he gets stupidly higher frames then that.
Tom's hardware seem to have a thing against AMD from what I've seen in the past.


Its because they want to show the 'raw' power of the gpus and that cpu wouldn't hold the gpus back one bit.
Just get another 780, you should be golden although the game isn't exactly your CPU's friend things should improve.

You know... Linus (and co) and TTL especially do pretty decent bullshit busters as I like to call it. 4960x vs 4930k maybe 1 or 2 FPS? Graphics cards are the main factor in almost any game.

Premium motherboard vs shit ITX, mATX, ATX 1-2 FPS at stock clocks. So while the CPU is unrealistic it's at least consistent. You know that there's no CPU bottleneck so the only factor in performance comes down to GPU and that the GPU is the single differential in performance between benchmarks.

edit: wasn't disagreeing and that was a weird segue...

 

[SeekaX]

Oh my god lol! There is! I just didn't think it could be so low, lol. I thought it was odd that there was a 690 vs 7990 on everything else except this benchmark. Haha ROFL.




You know... Linus (and co) and TTL especially do pretty decent bullshit busters as I like to call it. 4960x vs 4930k maybe 1 or 2 FPS? Graphics cards are the main factor in almost any game.

Premium motherboard vs shit ITX, mATX, ATX 1-2 FPS at stock clocks. So while the CPU is unrealistic it's at least consistent. You know that there's no CPU bottleneck so the only factor in performance comes down to GPU and that the GPU is the single differential in performance between benchmarks.

edit: wasn't disagreeing and that was a weird segue...

i7 4960x vs i5 4670 makes next to no difference either, for gaming anything better than an i5 is a complete waste of money. But everybody knows the i7 brand and that's why Alienware and friends get away with putting overspec'd CPUs with badly underspec'd GPUs into gaming rigs.
It's not really a bust, it's common knowledge around the forums, but it's very hard to get people to understand who are new to PCs. It's always a tragedy when you spend hours explaining to a guy that his 4930k is completely wasted because the GT630 will bottleneck in games and then the advice is ignored because 'Alienware have never disappointed me'.

 

[PaddieMayne]

Just thought id let you guys know, that i went with SLI GTX780ti in the end as i came into a little money so splashed out also got a 500gb Samsung Evo SSD, ill let you guys know the Arma 3 results shortly...

 

[tinytomlogan]

Looks nice fella - shame about that 120mm rad at the top though.

 

[PaddieMayne]

Why? Do you think it ruins the aesthetics ?
Maybe i should get rid as its all hooked up to a 420mm rad outside, but not sure if a single 420mm rad with 6 x 140mm fans could cope with the TDP of 2 x 780ti and an 5ghz ocd 9590??

 

[SuB]

woh plottwist much, rad OUTSIDE?!

 

[PaddieMayne]

Yeah its on the outside wall under my car port. Been runing this set up for over 18 months now, and although i had initial concerns regarding condensation, the good news is i have never suffered any problems, thankfully.
More pics..

 

[barnsley]

I've been looking into it more and it seems that for some odd reason the 8350 is only a couple of frames behind the i7 3770 in arma3 in single player. Online performance is almost certainly crap across the board due to severs not having enough grunt (they tend to be wildly underspecced, you ideally want a dedicated monster for a big A3 server, we're talking six core xeons with a headless AI client server, when that gets implimented) meaning they slow everything down for everyone.

 

[SuB]

Have they got Arma 3 servers using more than 1 cpu yet? I was under the impression it still wasn't multi-core aware.

I cry when I join Altis servers with 6 fps serverside need somewhere to play domi without the ensuing death of the server.

 

[PaddieMayne]

Im personaly into Waistland, its a great game try it out and server seems not to bad.

 

[barnsley]

Have they got Arma 3 servers using more than 1 cpu yet? I was under the impression it still wasn't multi-core aware.

I cry when I join Altis servers with 6 fps serverside need somewhere to play domi without the ensuing death of the server.

Last I heard they were using ~4 cores dependent on setup (virtual servers are still used en masse and probably only use one core). Some gamemodes actually do some very cleaver stuff to get around the low sfps issue.

 

[SuB]

Im personaly into Waistland, its a great game try it out and server seems not to bad.

Aye wasteland is alright, but there isn't enough content to make me want to play it more than 2 at the minute and servers are still quite sluggish.

Can you do me a small favour dude, can you reduce your signature image height by about 50% or so, we're trying to move to a more wide than tall view on sigs. I hadn't spotted yours before now

Last I heard they were using ~4 cores dependent on setup (virtual servers are still used en masse and probably only use one core). Some gamemodes actually do some very cleaver stuff to get around the low sfps issue.

Interesting indeed.. how do you mean?

 

[barnsley]

Interesting indeed.. how do you mean?

Well, not THAT interesting just disabling bits and pieces, vigorous despawning of objects (e.g. weapons), completely disabling animals , focusing AI in areas that have players in. Supposedly (I've been talking to some people who run servers) things like that do get the FPS up.
Basically, not THAT smart but as with anything to do with arma, its irritaitingly simple for such a complex problem you could call it smart.
In terms of the actual server itself, (much like A2) -noCB and disabling Hyperthreading (if relevent) seems to boost performance ALOT. Also, choosing a server host which advertises server specs will save alot of bother.

 

[SuB]

Who you got yours with?

 

[barnsley]

Who you got yours with?

I've used Jest servers in the past as they have decent hardware/price ratio (proper dedicated servers, yum). Now I don't use any hosts as I cba to run a server for the time being.
My friends currently use jest or host their own.