Metro 2033
Game » consists of 11 releases. Released Mar 16, 2010
Metro 2033 is a post-apocalyptic first-person-shooter, set in the underground community built in the ruins of a Russian municipal train system, based on the best-selling novel by Dmitry Glukhovsky.
Metro 2033 Interview: Multithreading and more
New interview published for the game Metro 2033. In it, the developer mainly discusses the workings of its engine. Rather interestingly, they say that the game has been developed entirely multithreaded from the ground up, which means it will will scale 1:1 with CPU power. This would be one of the first examples of a game engine (other than something like a chess engine) to do so. You can read the full interview here: http://games.on.net/article/7836/Metro_2033_-_Technical_QA
it pisses me off that nvidia is pushing closed standards... Its such a stupid thing to do to the industry as a whole. Because of there buisness practices I will not purchase there hardware. They wont even allow you to use physx if you have an amd/ati gpu present in your system. But that was a cool interview, I am definatly purchasing this game just a matter of whether I wait for a discount or pay the full price.. I am of the mind if developers are pushing the bleeding edge I want to support them.
" @Geno: it pisses me off that nvidia is pushing closed standards... Its such a stupid thing to do to the industry as a whole. Because of there buisness practices I will not purchase there hardware. They wont even allow you to use physx if you have an amd/ati gpu present in your system. But that was a cool interview, I am definatly purchasing this game just a matter of whether I wait for a discount or pay the full price.. I am of the mind if developers are pushing the bleeding edge I want to support them. "You should know that ATI pushed for Direct3D9 standards to be utterly closed away from Nvidia at the start of DirectX9 with the release of the ATI 9800 and 9600 series cards. It's why we ended up having three revisions of DirectX9 (A, B and C). A included the ATI standard shader compiler and it was written with ATI's defined shader length and preset library (materials) in mind because it was basically written by ATI themselves. B supported half precision shader modes for NV 57xx and 57xx but it came so late in the piece and the Nv TWIMTBP couldn't save it (ATI did their darndest to block it coming at all complaining that it would leave game devs with too much legwork and that ATI's standard was the only one which should be supported - it's the main reason why so few games worked well on NV hardware of that generation because almost none of them supported half precision shaders and the few which did, did so poorly at much performance expense) while C supported Shader Model 3 and all that came with it (which ATI had no chance to block because they had nothing in the way performance or developer driven metrics which proved that SM3.0 and NV 68xx/66xx chipsets wouldn't be accepted by game devs. All of this can be directly connected to the falling out between Nvidia and MS over the XBOX where MS didn't want to pay Nvidia the appropriate royalties they owed for NV's work on consoles GPU and Nforce(2) chipset trying to blame the expense of the R&D costs on Nvidia which were all under budget. Instead of battling it out in court Nvidia decided to go on selling future NForce chipsets without the licensed Dolby realtime encoder (which Dolby later onsold as Dolby Digital Live) even though they had been promised Dolby's support via MS (length of contract unknown).
The point is that both companies have pushed for closed standards, so don't try and claim that ATI don't play dirty. Quake/Quack anyone?
Then again I am an optimist, and that the other guy is probably an aggressive Nvidia hater. Who knows...
@ThatFrood: Indeed. I'm glad to see that some companies aren't scared to push the envelope and encourage more spend in the market. Origin did way back in 1990 with Wing Commander (386 required) as did Bullfrog a years later with Magic Carpet (support for 3D and VR headgear) and more recently so did VALVe (SMP support in engine via L4D and SM3.0 in HL2 : Lost Coast) and Crytek (Farcry, CE2, CE3 - SM 3/4 instancing, dynamic destructibility, water physics, proper IK) so it's great to see another company is willing to follow in the footsteps of giants in developing their own engine and taking advantage of the benefits offered by the PC platform.
@SeriouslyNow:
Just like half the stuff in DX10 was thrown away because nvidia couldnt support it at the time.. thus dx10.1. I am not replying because I hate nvidia but comparing an api which is not a closed standard to a program like physx to me is not the same thing. I am not an nvidia hater at all, I just dont like the way they are dealing with this situation. Its all good though because the red side is working with bullet physics and pixelux on something called DMM which will be an open physics program. They will have to pony up some cash though and get the ball rolling on this feature if its ever gonna take off.
http://www.rage3d.com/board/showthread.php?t=33960673
And to the guy who said physx will take off and eventually become an open standard ... thats most likely the land of make believe. Yeah there is a work around by using outdated drivers.
"First ; rage3d as evidence. I lol heartily. You are most certainly an nvidia hater, don't make the pretense of balanced opinion and then post a rage3d thread because two are mutually exclusive. Non ATI folk don't frequent the board.@SeriouslyNow:
Just like half the stuff in DX10 was thrown away because nvidia couldnt support it at the time.. thus dx10.1. I am not replying because I hate nvidia but comparing an api which is not a closed standard to a program like physx to me is not the same thing. I am not an nvidia hater at all, I just dont like the way they are dealing with this situation. Its all good though because the red side is working with bullet physics and pixelux on something called DMM which will be an open physics program. They will have to pony up some cash though and get the ball rolling on this feature if its ever gonna take off.
http://www.rage3d.com/board/showthread.php?t=33960673"
And to the guy who said physx will take off and eventually become an open standard ... thats most likely the land of make believe. Yeah there is a work around by using outdated drivers.
Second : There's a difference between MS mediating standards so that vendors can offer product with their own specialisations if they wish to (10.1 is about as useful as dogs balls) as opposed to ATI being arseholes and literally trying to push NV out of the market in the case of DX9. Karma got them in them in the end though and ATI struggled for two more generations to match Nvidia's SM3/SM4 implementations at almost every level.
Third : DMM is one of many physics solutions and it's not yet ratified enough to gain any exposure and no amount of cash will make it worthy now as people are either going non middleware, such in CryEngine, XRAY and SeriousEngine or they are opting for Havok and PhysX CPU driven middleware and occasionally offering PhysX for GPU higher precision stuff. PhysX is already pretty widely supported in its non GPU version so I think that, at least for the next two-five years it will be in a neck and neck race with Havok and other inhouse solutions which use the CPU. I can't see many other physics middleware taking much of that ground from either of the two I just mentioned with the small exception of Euphoria for its AI driven IK.
Forth : which guy talked about physX becoming an open standard?
" @Turien said:"First ; rage3d as evidence. I lol heartily. You are most certainly an nvidia hater, don't make the pretense of balanced opinion and then post a rage3d thread because two are mutually exclusive. Non ATI folk don't frequent the board. Second : There's a difference between MS mediating standards so that vendors can offer product with their own specialisations if they wish to (10.1 is about as useful as dogs balls) as opposed to ATI being arseholes and literally trying to push NV out of the market in the case of DX9. Karma got them in them in the end though and ATI struggled for two more generations to match Nvidia's SM3/SM4 implementations at almost every level. Third : DMM is one of many physics solutions and it's not yet ratified enough to gain any exposure and no amount of cash will make it worthy now as people are either going non middleware, such in CryEngine, XRAY and SeriousEngine or they are opting for Havok and PhysX CPU driven middleware and occasionally offering PhysX for GPU higher precision stuff. PhysX is already pretty widely supported in its non GPU version so I think that, at least for the next two-five years it will be in a neck and neck race with Havok and other inhouse solutions which use the CPU. I can't see many other physics middleware taking much of that ground from either of the two I just mentioned with the small exception of Euphoria for its AI driven IK. Forth : which guy talked about physX becoming an open standard? "@SeriouslyNow:
Just like half the stuff in DX10 was thrown away because nvidia couldnt support it at the time.. thus dx10.1. I am not replying because I hate nvidia but comparing an api which is not a closed standard to a program like physx to me is not the same thing. I am not an nvidia hater at all, I just dont like the way they are dealing with this situation. Its all good though because the red side is working with bullet physics and pixelux on something called DMM which will be an open physics program. They will have to pony up some cash though and get the ball rolling on this feature if its ever gonna take off.
http://www.rage3d.com/board/showthread.php?t=33960673"
And to the guy who said physx will take off and eventually become an open standard ... thats most likely the land of make believe. Yeah there is a work around by using outdated drivers.
I go to guru3d, rage3d and widescreengaming forums all the time. You are making alot of assumptions and false blame.... Its not amd/ati's api, its Microsofts so your beef should be with them but I can see you are clearly wearing green colored glasses. Can you see the future in terms of opencl's success and its impact on AI and physics in games and all the possibilities? I am quite sure theres a movment to do physics and other cpu intensive tasks on the gpu (not including nvidia's solution current physx solution), I think thats right around the corner. Hell you know nvidia hopes so. I wasnt claiming that DMM was going to overtake havok or physx in the near future, I just said I am happy to see that there could be a better solution in the works.. but apparently you know so much about DMM and its non impact on the industry.
quote: Karma got them in them in the end though and ATI struggled for two more generations to match Nvidia's SM3/SM4 implementations at almost every level. :endquote:
Unfortunatly for amd/ati they did alot of extra work in vein at the start of dx10 so yea sm4 was big let down (2000 series). But they had alot of features that were not put in because nvidia could not support it. I noticed how you left sm4.1 off your list? HD4000 series hurt nvidia bad even though there single gpu solutions couldnt quite match nvidia's they really hurt nvidia's profit margins.. I am pretty sure thats why they stopped production of gt200 well before there latest failure (so far) has hit the market. Fermi is most likely going to fail.. they wont be able to supply the nvidia fanboys.. prices are going to be sky high.. I just hope they figure it out the second time around. Competition is good and so far amd/ati has none.
" @SeriouslyNow said:I didn't leave anything off my list I talked specifically about two generations. Either you can deal with the facts or you can't. There's no need to try and make it sound like I was twisting the truth because I plainly wasn't, 4xxx and 5xxxx ATI series are third and forth gen hardware. It's true that Nvidia hasn't had anything to answer ATI's DX11 hardware and in some cases pre DX11 performance, but that hasn't stopped them being more financially successful than ATI throughout the same quarter. ATI didn't do extra work through the DX9 phase, they just took advantage of MS's awful treatment of Nvidia so I'm personally glad that the extra work they did do in the HD2xxxx series worked against them because frankly ATI deserved some bad karma to come their way in kicking Nvidia when they were down." @Turien said:I go to guru3d, rage3d and widescreengaming forums all the time. You are making alot of assumptions and false blame.... Its not amd/ati's api, its Microsofts so your beef should be with them but I can see you are clearly wearing green colored glasses. Can you see the future in terms of opencl's success and its impact on AI and physics in games and all the possibilities? I am quite sure theres a movment to do physics and other cpu intensive tasks on the gpu (not including nvidia's solution current physx solution), I think thats right around the corner. Hell you know nvidia hopes so. I wasnt claiming that DMM was going to overtake havok or physx in the near future, I just said I am happy to see that there could be a better solution in the works.. but apparently you know so much about DMM and its non impact on the industry. quote: Karma got them in them in the end though and ATI struggled for two more generations to match Nvidia's SM3/SM4 implementations at almost every level. :endquote: Unfortunatly for amd/ati they did alot of extra work in vein at the start of dx10 so yea sm4 was big let down (2000 series). But they had alot of features that were not put in because nvidia could not support it. I noticed how you left sm4.1 off your list? HD4000 series hurt nvidia bad even though there single gpu solutions couldnt quite match nvidia's they really hurt nvidia's profit margins.. I am pretty sure thats why they stopped production of gt200 well before there latest failure (so far) has hit the market. Fermi is most likely going to fail.. they wont be able to supply the nvidia fanboys.. prices are going to be sky high.. I just hope they figure it out the second time around. Competition is good and so far amd/ati has none. ""First ; rage3d as evidence. I lol heartily. You are most certainly an nvidia hater, don't make the pretense of balanced opinion and then post a rage3d thread because two are mutually exclusive. Non ATI folk don't frequent the board. Second : There's a difference between MS mediating standards so that vendors can offer product with their own specialisations if they wish to (10.1 is about as useful as dogs balls) as opposed to ATI being arseholes and literally trying to push NV out of the market in the case of DX9. Karma got them in them in the end though and ATI struggled for two more generations to match Nvidia's SM3/SM4 implementations at almost every level. Third : DMM is one of many physics solutions and it's not yet ratified enough to gain any exposure and no amount of cash will make it worthy now as people are either going non middleware, such in CryEngine, XRAY and SeriousEngine or they are opting for Havok and PhysX CPU driven middleware and occasionally offering PhysX for GPU higher precision stuff. PhysX is already pretty widely supported in its non GPU version so I think that, at least for the next two-five years it will be in a neck and neck race with Havok and other inhouse solutions which use the CPU. I can't see many other physics middleware taking much of that ground from either of the two I just mentioned with the small exception of Euphoria for its AI driven IK. Forth : which guy talked about physX becoming an open standard? "@SeriouslyNow:
Just like half the stuff in DX10 was thrown away because nvidia couldnt support it at the time.. thus dx10.1. I am not replying because I hate nvidia but comparing an api which is not a closed standard to a program like physx to me is not the same thing. I am not an nvidia hater at all, I just dont like the way they are dealing with this situation. Its all good though because the red side is working with bullet physics and pixelux on something called DMM which will be an open physics program. They will have to pony up some cash though and get the ball rolling on this feature if its ever gonna take off.
http://www.rage3d.com/board/showthread.php?t=33960673"
And to the guy who said physx will take off and eventually become an open standard ... thats most likely the land of make believe. Yeah there is a work around by using outdated drivers.
All of that rest I can't bothered responding to until you formulate it in such a way where you either say yes, you're staunchly and irrationally anti Nvidia or restructure it in some way to show actual balance.
Dude I simply brought up the fact that I didnt agree with the way nvidia is doing buisness because it hurts the pc gaming industry as a whole. You brought up the whole DX9 thing and pointed your green colored fingers at ati.. when in fact that was Microsoft. Because of this I guess you wanted ATI to stick of up for its competitor, which is laughable. "No Microsoft dont do DX9 this way that would hurt nvidia's overall performance with there current graphics cards"... You had absolutly nothing to counter my disagreement with so you went digging for some bs counter example because your blinded fanboy rage got in the way of logic. Then you went off on a tangent about sm3/4 and how Karma got amd/ati.. So I guess Karma left when dx10.1 and sm4.1 came out lol. Whether DX10.1 is relevant is irrelelvant. In fact you just typed up a bunch of bs and I played your game.
"that didnt stop nvidia from being more successful in the same quarter" 5xxx series is clearly a better choice for any gamer to take. Q4 there were shortages of the 5xxx series cards they were not readily available until the end of December.. which is the end of the quarter so please... Take a look at this quarter that will give you a much better idea of who was more successful. There are alot of stubborn nvidia fanboys out there. With all that said I do actually use nvidia products just not on my high end gaming rig. I have no brand loyalty so to speak. Brand loyalty is not a smart thing to take practice in.
Please Log In to post.
This edit will also create new pages on Giant Bomb for:
Beware, you are proposing to add brand new pages to the wiki along with your edits. Make sure this is what you intended. This will likely increase the time it takes for your changes to go live.Comment and Save
Until you earn 1000 points all your submissions need to be vetted by other Giant Bomb users. This process takes no more than a few hours and we'll send you an email once approved.
Log in to comment