Scrapyard Wars Episode 1b

[sgloux3470]

Since the cpu's they going to use is not very powerfull

 

They were good enough to run games extremely well what, 6-7 years ago?  

 

You guys are seriously overestimating the importance of the CPU in most of these games.  If they were benching Total War or Arma you may have a point, but for most stuff they'll be fine.  Even if you wanted to argue about CPU power, they more or less (clock speed dependent) have nearly identical performing CPU's so it will still be a GPU arms race anyway.

 

https://www.youtube.com/watch?v=NXg-mCPUgHQ#t=89

 

Even a 290 isn't going to push enough frames on ultimate settings to reach a CPU bottleneck scenario in Tomb Raider.  It tends to sit around ~60-70 FPS in Tomb Raider on those settings which is more than manageable for either CPU.

[vt3c]

Final part should come tonight!

[vt3c]

i

 know it's a quad core but each of those cores are like a slightly under clocked and upgraded version of luke's cores right?

This is a strange debate, 2 wolfdale cores vs 4 yorkfield. 

Holy shit I'm hyped for this next episode. 

[SSOB]

I utterly dislike episodes this short. As I've mentioned in the previous topic: ~20 minutes is the sweetspot. Money isn't everything and you can't keep pushing the limits because once you cross a certain threshold people won't even bother to watch the videos anymore.

[Sharkyx1]

They were good enough to run games extremely well what, 6-7 years ago?  

 

You guys are seriously overestimating the importance of the CPU in most of these games.  If they were benching Total War or Arma you may have a point, but for most stuff they'll be fine.  Even if you wanted to argue about CPU power, they more or less (clock speed dependent) have nearly identical performing CPU's so it will still be a GPU arms race anyway.

 

https://www.youtube.com/watch?v=NXg-mCPUgHQ#t=89

 

Even a 290 isn't going to push enough frames on ultimate settings to reach a CPU bottleneck scenario in Tomb Raider.  It tends to sit around ~60-70 FPS in Tomb Raider on those settings which is more than manageable for either CPU.

The dumb thing about TRs ultimate settings is its medium shadows and FXAA 

 

but yeah, TR is basically unaffected by CPU speed

 

EDIT: Especially in the benchmark, since there are like no objects, i imagine they wont use the built in one since they don't normally use those

[Terminashunator]

Can we get a final spec list and price breakdown for both systems?

[Diceair]

Since when did Tomb Raider give one lick about CPU power?

 

Sir you just got served. I told you cpu will play some part

[Flamewalker]

I do wonder though if the gap would have been like this if they had an nvidia gpu ,the dx11 driver overhead is so much lower on nvidia cards then on amd cards 

[sgloux3470]

I do wonder though if the gap would have been like this if they had an nvidia gpu ,the dx11 driver overhead is so much lower on nvidia cards then on amd cards

I think that is only specifically in regards to Star Swarm.

Sir you just got served. I told you cpu will play some part

Touche. I was rather surprised. Old Phenom II's tend to hold their own in Tomb Raider. I think clock speed played more of a role than core count though.

[Flamewalker]

I think that is only specifically in regards to Star Swarm.

It's not, its pretty much in every dx 11 game atm (the more cpu limited you become the more noticable it becomes), star swarm just makes it blatently obvious how bad the driver overhead is on amd compared to nvidia 

 

 

Just a quick example of what i mean, theres alot more about it (http://www.overclock.net/t/1528559/directx-driver-overhead-and-why-mantle-is-a-selling-point-bunch-of-benchmarks first page here has alot of screenshots showing you how bad the dx11 overhead is on amd cards on several games)

[sgloux3470]

It's not, its pretty much in every dx 11 game atm (the more cpu limited you become the more noticable it becomes), star swarm just makes it blatently obvious how bad the driver overhead is on amd compared to nvidia 

 

 

Just a quick example of what i mean, theres alot more about it (http://www.overclock.net/t/1528559/directx-driver-overhead-and-why-mantle-is-a-selling-point-bunch-of-benchmarks first page here has alot of screenshots showing you how bad the dx11 overhead is on amd cards on several games)

 

Wonder what it looks like with non-gameworks/non-mantle games.

 

I would imagine that Mantle supporting games get less driver focus for DX11 mode, and gameworks will be gameworks.

 

I can say that with the games I get a CPU bottleneck with (Rome 2, Arma 3) I had zero difference going from 270xCF to a GTX 980.  So I'd be hesitant to blame their drivers entirely when it could be game specific.