HD5870 and HD5850 Reviews

[Hornet331]

I think you guys are still kind of high... a core 2 around 2.6ghz is enough to handle a lot... but why run a core 2 under 3ghz?

Depends purly on the engien, TF2 carves for every MHZ. Even with a 3.8ghz ci7 and that thing still scales with every mhz more.

[Jamesrt2004]

AMD Demos Dual Cypress Graphics Card - Hemlock
http://vr-zone.com/articles/amd-demo....html?doc=7718

priced below $500.... so the SINGLE cypress for 370 seems a bit steep.

[Dante80]

Except if they were talking about the 5850X2....

[jmke]

Depends purly on the engien, TF2 carves for every MHZ. Even with a 3.8ghz ci7 and that thing still scales with every mhz more.

ah yes, the venerable TF2, the source engine, known for being a system crusher, and completely unplayable at anything below 50 FPS.



ow... what now it seems that at 2560x1600 4xAA/16xAF you'll get 50fps with a HD 4770...
so with a 2.4Ghz C2D you'll be quite safe here; no need for a faster CPU. you missed the point where even a slower CPU will still be fast enough, sure a faster CPU might give you higher FPS, but those are wasted

[Hornet331]

ah yes, the venerable TF2, the source engine, known for being a system crusher, and completely unplayable at anything below 50 FPS.



ow... what now it seems that at 2560x1600 4xAA/16xAF you'll get 50fps with a HD 4770...
so with a 2.4Ghz C2D you'll be quite safe here; no need for a faster CPU. you missed the point where even a slower CPU will still be fast enough, sure a faster CPU might give you higher FPS, but those are wasted

play on a 24+ player, best on a 32player server and then show me the min fps.

You will get dips in the high 30s regardless of which resolution your using, guaranteed and not just for a second but rather quite long (a few seconds, as long as the action lasts).

Syntesitc testing of TF2 means jack.

[Manicdan]

play on a 24+ player, best on a 32player server and then show me the min fps.

You will get dips in the high 30s regardless of which resolution your using, guaranteed and not just for a second but rather quite long (a few seconds, as long as the action lasts).

Syntesitc testing of TF2 means jack.

like when people say they get 100fps in WoW, but that was done in the middle of no where, go to dalaran (or whatever the popular city is at that time) and you can see single digit fps on almost any hardware with max settings.

[Sgt.McRuff]

About tf2, core i7 helps a lot, set in launch options -threads 8 . By default tf2 will only use 3 with multi-core turned on. The engine can support up to 16.

Overclock i7, those blasted particles will never see it coming

[W1zzard]

tf2 is not usable as benchmark for modern graphics cards .. next revision of our vga benchmark setup will most probably not have tf2 anymore.. way too cpu limited

[jmke]

play on a 24+ player, best on a 32player server and then show me the min fps.

33 player server, 1920x1200 8xAA/16xAF, always >60FPS; avg, no dips under 40fps. And that with 9800GTX+

way too cpu limited

in the sense that it will keep on scaling with higher CPU speeds; but that doesn't necessarily mean anything
if the lowest FPS you get with a entry level system is already playable, there's no return in investment if you go for a faster CPU

[natty]

sure a faster CPU might give you higher FPS, but those are wasted

not if you have a 120hz monitor

[jmke]

ok, for that 0.1% go OC your i7 ;-)

[kl0012]

if the lowest FPS you get with a entry level system is already playable, there's no return in investment if you go for a faster CPU

Hmm... Well, then HD5870 also isn't good investment since you get playable FPS(>50) with GTX 260, 1920x1080 res. almost in every game according to techPowerUp. Some games (like Crysis) are unplayable but HD5870 also gets pretty low score there (<50 FPS).

[STaRGaZeR]

tf2 is not usable as benchmark for modern graphics cards .. next revision of our vga benchmark setup will most probably not have tf2 anymore.. way too cpu limited

Never ever remove FEAR.

[informal]

Has anyone seen the CPU scaling review with these 5000 series cards? The only one I saw that closely resembles that is this one and it shows that 5870 works equally well on both Deneb and Bloomfield at the same clocks(ie. there is no CPU bottleneck),while the dual GPU cards tend to scale a bit different than this.

[Chickenfeed]

not if you have a 120hz monitor

Let me remind folks that many games use tick based netcode online and high framerates DO result in better game play online (regardless of refresh rate )

The Quake 3 engine ( for relevance COD through MW2 all use Quake code at their cores ) runs ideally at 125FPS. In source based titles 100tick rate is common in competitive servers / games ( all but Left 4 Dead basically ) Even though these "extra" frames can be argued as being wasted from a visually stand point, any remotely competitive Counter Strike / Day of Defeat / Team Fortress or Call of Duty player can and WILL notice a difference between 60 and 100+fps.

Now again this mainly relevant in a handful of multi player first person shooters. It obviously doesn't matter if all you play is single player or MMOs ( most don't benefit from higher framerate as far as netcode goes )

I just like to mention this as all this talk of frame rates in excess of 60 being wasted is often invalid. I for one can't play CS or DOD unless I'm getting 100fps constant ( it is just what I've been used to ever since 1.6 and 1.3 )

[smdude]

Exactly, the physics of Q3 feel so off below 125 FPS

[villa1n]

Let me remind folks that many games use tick based netcode online and high framerates DO result in better game play online (regardless of refresh rate )

The Quake 3 engine ( for relevance COD through MW2 all use Quake code at their cores ) runs ideally at 125FPS. In source based titles 100tick rate is common in competitive servers / games ( all but Left 4 Dead basically ) Even though these "extra" frames can be argued as being wasted from a visually stand point, any remotely competitive Counter Strike / Day of Defeat / Team Fortress or Call of Duty player can and WILL notice a difference between 60 and 100+fps.

Now again this mainly relevant in a handful of multi player first person shooters. It obviously doesn't matter if all you play is single player or MMOs ( most don't benefit from higher framerate as far as netcode goes )

I just like to mention this as all this talk of frame rates in excess of 60 being wasted is often invalid. I for one can't play CS or DOD unless I'm getting 100fps constant ( it is just what I've been used to ever since 1.6 and 1.3 )

+1 ~~~ I dont play less than 125 in any fps!!

[billdavis]

nope nope nope Doom 3 is the best

I'd listen to Wizzard

annnnd if you can see faster than 60fps...like your house(in the states) lights flicker @~60 let me know any supplments/vitamins your taking, that would be awesome

[onethreehill]

Tested in DX11!

http://pclab.pl/art38674-8.html

[Leeghoofd]

it's not about seeing the fps, it's the feeling; the usual smoothness you loose when the fps drops below a certain level... any hardcore online fps gamer knows what I'm talking about...and will notice directly via the game input/ouput that the game is acting up... don't argue bout that, there's a world of difference between being a single or muliplayer experience...

[gamervivek]

But it isn't. Everyone who played Q3 seriously knows about the com_maxfps 120 and what it does for them. Having graphics and physics 1:1 allows for the unique trick jumping possibilities, and not only that, it is the ONE and ONLY engine that has no solid limits, enabling skilled players to move much faster and to jump much further.

Any lucky dude can score against a professional player in ANY OTHER GAME, but Q3 is the only game where even minor differences in skill result in a whole dozen of cans of whoopass for the one who is less experienced. I've played with people that thought of themselves as of GOOD GAMERS, having good skills in other FPS games like CS or COD, and when I play Q3 with them it is a major ownage. It is something very special to play with someone who thinks he knows FPS because he is good at some game, and in Q3 in a quick 15 minutes game the end result is like 40:-4, yes thats negative 4 for the "pro gamer" who thought he is any good at quake for being good at some other FPS.

No other game can do it. As old and outdated Q3 is, in my book it has the best gameplay physics up to date. I don't mean pure eye candy physics like nvidia does, but GAMEPLAY physics.

well said,extremely.

[villa1n]

But it isn't. Everyone who played Q3 seriously knows about the com_maxfps 120 and what it does for them. Having graphics and physics 1:1 allows for the unique trick jumping possibilities, and not only that, it is the ONE and ONLY engine that has no solid limits, enabling skilled players to move much faster and to jump much further.

Any lucky dude can score against a professional player in ANY OTHER GAME, but Q3 is the only game where even minor differences in skill result in a whole dozen of cans of whoopass for the one who is less experienced. I've played with people that thought of themselves as of GOOD GAMERS, having good skills in other FPS games like CS or COD, and when I play Q3 with them it is a major ownage. It is something very special to play with someone who thinks he knows FPS because he is good at some game, and in Q3 in a quick 15 minutes game the end result is like 40:-4, yes thats negative 4 for the "pro gamer" who thought he is any good at quake for being good at some other FPS.

No other game can do it. As old and outdated Q3 is, in my book it has the best gameplay physics up to date. I don't mean pure eye candy physics like nvidia does, but GAMEPLAY physics.

Q3 is not even in the same category as CS or CoD... its a SPAM fps, not a tactical fps. Which is why most players who are good at tactics , not bunny hopping lazer parties would get owned by people who are. Its preference, but they are not the same just because they are first person..

Quake Spam Finals

cod4 clutch

[gamervivek]

Q3 is not even in the same category as CS or CoD... its a SPAM fps, not a tactical fps. Which is why most players who are good at tactics , not bunny hopping lazer parties would get owned by people who are. Its preference, but they are not the same just because they are first person..

Quake Spam Finals

cod4 clutch

Bunny hopping laser parties
but you can't deny that Q3 is a game of skill,you can get a kill over a far better player by luck in CS but in Q3 there's simply no chance.
And even if they ain't the same game it boils down to what you'll play as if it was the last game you got to play-
1)A game where your tactics were lost on someone who camped the whole round only to shoot you in the head as you went past him.
3)comboing someone with rocket and rail at spawn....... three times in a row.

I know what I'll choose.

[villa1n]

Bunny hopping laser parties
but you can't deny that Q3 is a game of skill,you can get a kill over a far better player by luck in CS but in Q3 there's simply no chance.
And even if they ain't the same game it boils down to what you'll play as if it was the last game you got to play-
1)A game where your tactics were lost on someone who camped the whole round only to shoot you in the head as you went past him.
3)comboing someone with rocket and rail at spawn....... three times in a row.

I know what I'll choose.

But thats assuming your playing in Pubs.. versus scrims / matches.. which for the most part avoid camping noobs. Not denying it takes practice to become that good at q3..but thats the same with anything, 10 000 hours and your an expert, i m just saying, there is more mental tactics in games like cod4 which are more team based, have base of fire, rushing, flanking etc, q3 reminds me more of thunderdome, 2 men enter 1 man leaves.

Edit: Dont want to Derail a Review Thread ^^ everyone has preferences, just wanted to point just because its a First Person view doesn't mean Game play is comparable.

[jmke]

Hmm... Well, then HD5870 also isn't good investment since you get playable FPS(>50) with GTX 260, 1920x1080 res.

yup; never said it was a good investment if you all play is TF2