Thread: [GD-General] Bandwidth
Brought to you by:
vexxed72
Menu
▾
▴
gamedevlists-general
|
From: Brett B. <res...@ga...> - 2004-09-21 00:08:58
|
Can anyone with a _shipped_ network-enabled (WAN, not LAN) title comment on how much bandwidth is used for your game? I'm trying to understand the network bandwidth needs for a variety of game types, so the game type (name is even better if you can disclose) and the bits/bytes per second throughput would be appreciated. I have done a lot of googling and it seems there is a big difference between the games that shipped and those that are in development in terms of usage, so I'm hoping to limit the information to actually shipped titles and real-world numbers. Thank you. Brett |
|
From: Javier A. <ja...@py...> - 2004-09-21 08:12:45
|
Memory is fuzzy, but here I go. Praetorians (RTS using standard fare lock-step multiplayer) peaked at about 8K/s on a server with 8 players (server collects input packets from all clients and resends them to every client, 5 times per second). Well, higher than that if all players decide to start frantically banging their keyboards / mice. Average bandwidth would be much lower because the low frequency of commands during actual gameplay, but we never bothered checking it; perhaps 2-3K/s. Brett Bibby wrote: > Can anyone with a _shipped_ network-enabled (WAN, not LAN) title > comment on how much bandwidth is used for your game? I'm trying to > understand the network bandwidth needs for a variety of game types, > so the game type (name is even better if you can disclose) and the > bits/bytes per second throughput would be appreciated. > > I have done a lot of googling and it seems there is a big difference > between the games that shipped and those that are in development in > terms of usage, so I'm hoping to limit the information to actually > shipped titles and real-world numbers. -- Javier Arevalo Pyro Studios |
|
From: Brett B. <res...@ga...> - 2004-09-21 08:17:29
|
Thanks! Is that kilobyte or kilobit per second? Bandwidth is usally quoted in bits so I want to be sure... ----- Original Message ----- From: "Javier Arevalo" <ja...@py...> To: <gam...@li...> Sent: Tuesday, September 21, 2004 4:23 PM Subject: Re: [GD-General] Bandwidth > Memory is fuzzy, but here I go. Praetorians (RTS using standard fare > lock-step multiplayer) peaked at about 8K/s on a server with 8 players > (server collects input packets from all clients and resends them to every > client, 5 times per second). Well, higher than that if all players decide > to start frantically banging their keyboards / mice. Average bandwidth > would be much lower because the low frequency of commands during actual > gameplay, but we never bothered checking it; perhaps 2-3K/s. > > Brett Bibby wrote: >> Can anyone with a _shipped_ network-enabled (WAN, not LAN) title >> comment on how much bandwidth is used for your game? I'm trying to >> understand the network bandwidth needs for a variety of game types, >> so the game type (name is even better if you can disclose) and the >> bits/bytes per second throughput would be appreciated. >> >> I have done a lot of googling and it seems there is a big difference >> between the games that shipped and those that are in development in >> terms of usage, so I'm hoping to limit the information to actually >> shipped titles and real-world numbers. > > -- > Javier Arevalo > Pyro Studios > > > > ------------------------------------------------------- > This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 > Project Admins to receive an Apple iPod Mini FREE for your judgement on > who ports your project to Linux PPC the best. Sponsored by IBM. > Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php > _______________________________________________ > Gamedevlists-general mailing list > Gam...@li... > https://lists.sourceforge.net/lists/listinfo/gamedevlists-general > Archives: > http://sourceforge.net/mailarchive/forum.php?forum_id=557 |
|
From: Sam M. <sa...@of...> - 2004-09-28 01:34:00
|
Savage (FPS / RTS hybrid, www.s2games.com) uses anywhere from 1k-10k per second per player, depending on how many other players are in the game and the bandwidth settings for each player (a player using 56k settings won't use more than 5k/sec). I was pretty happy with how efficient the network code turned out to be. Even in a 64 player game each player only consumes around 4k-6k per second on average, which is pretty good for an FPS. Still, to support that you need a pretty beefy server connection. FPSs generally require the highest amount of bandwidth of any genre, so if you're not developing one, you probably shouldn't use this as a benchmark. Sam McGrath www.offsetsoftware.com ----- Original Message ----- From: "Brett Bibby" <res...@ga...> To: <Gam...@li...> Sent: Monday, September 20, 2004 5:12 PM Subject: [GD-General] Bandwidth > Can anyone with a _shipped_ network-enabled (WAN, not LAN) title comment > on how much bandwidth is used for your game? I'm trying to understand the > network bandwidth needs for a variety of game types, so the game type > (name is even better if you can disclose) and the bits/bytes per second > throughput would be appreciated. > > I have done a lot of googling and it seems there is a big difference > between the games that shipped and those that are in development in terms > of usage, so I'm hoping to limit the information to actually shipped > titles and real-world numbers. > > Thank you. > Brett > > > ------------------------------------------------------- > This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 > Project Admins to receive an Apple iPod Mini FREE for your judgement on > who ports your project to Linux PPC the best. Sponsored by IBM. > Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php > _______________________________________________ > Gamedevlists-general mailing list > Gam...@li... > https://lists.sourceforge.net/lists/listinfo/gamedevlists-general > Archives: > http://sourceforge.net/mailarchive/forum.php?forum_id=557 |
|
From: mike w. <mi...@ge...> - 2004-09-29 09:36:38
|
i've done quite a bit of hosting of game servers over the past few years, and have found that it very much depended on the game engine and even the mod for the game engine, so i don't think you can really get any kind of useful information from just asking a generic 'how much does your game use' question. after monitoring and calculations, we found that a default half-life (version 1) server is about 50 kb/s per player pretty much straight up to the max players, you can run about 12 players on a standard cable/dsl before you max out and start seeing lag. as another example, battlefield 1942 used almost twice the bandwidth per player, upwards of 100 kb/s per player... mike w www.gekidodesigns.com Brett Bibby wrote: > Can anyone with a _shipped_ network-enabled (WAN, not LAN) title comment > on how much bandwidth is used for your game? I'm trying to understand > the network bandwidth needs for a variety of game types, so the game > type (name is even better if you can disclose) and the bits/bytes per > second throughput would be appreciated. > > I have done a lot of googling and it seems there is a big difference > between the games that shipped and those that are in development in > terms of usage, so I'm hoping to limit the information to actually > shipped titles and real-world numbers. > > Thank you. > Brett > > > ------------------------------------------------------- > This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 > Project Admins to receive an Apple iPod Mini FREE for your judgement on > who ports your project to Linux PPC the best. Sponsored by IBM. > Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php > _______________________________________________ > Gamedevlists-general mailing list > Gam...@li... > https://lists.sourceforge.net/lists/listinfo/gamedevlists-general > Archives: > http://sourceforge.net/mailarchive/forum.php?forum_id=557 > > |
|
From: Javier A. <ja...@py...> - 2004-09-29 10:10:48
|
mike wuetherick wrote: > after monitoring and calculations, we found that a default half-life > (version 1) server is about 50 kb/s per player pretty much straight up > to the max players, you can run about 12 players on a standard > cable/dsl before you max out and start seeing lag. > > as another example, battlefield 1942 used almost twice the bandwidth > per player, upwards of 100 kb/s per player... Wow, those numbers sound awfully high. Your "Standard DSL" outbound speed is *6* times what I have at home. :) At 50kb/s (6.25Kbytes/s), a 56K modem would barely be enough to let someone connect to a server. Maybe those numbers were measured using a heavier "LAN performance" option? (Not sure if those games support that, just wondering). -- Javier Arevalo Pyro Studios |
|
From: mike w. <mi...@ge...> - 2004-09-29 23:50:23
|
They surprised us as well, and the costs involved made us basically leave the 'game server hosting' business - the only way it was financially feasible (as far as we could see based on the bandwidth we saw going out of our servers) was to have a LOT of game servers (thus being able to get better/cheaper bandwidth rates). I think the bandwidth per player was more around the range of 32 kb/s per player for half-life until they added voice support, which added an extra 10-20 kb/s per player. BF-based games were definitely much higher. Btw, considering that people on modems are barely able to play modern games online, i wouldn't expect them to be able to host any kind of server, let alone anything close to what our servers on 100 Meg pipes could do. I mentioned cable/dsl usage based on what we found here in Vancouver Canada - we initially ran a single CS server based on a Shaw-cable modem, and found that it would barf out at between 12-16 players, we moved to a T1 line and found that it would support a single server up to about 20 players or so before it would barf out and then we moved to a 100 Meg coloc (which allowed us to finally use the machines to what they were able to, got 4-5 CS servers on a machine, 1 Natural Selection server per machine (extremely server-processor intensive). The bandwidth usage seemed fairly consistent across the various pipes we plugged the servers into. The LAN game server options for HL/BF were completely seperate types of servers (as far as the game-launching goes), these were entirely 'internet' servers as far as the game in concerned, but i'm not sure if the game itself handles the network bandwidth differently, or just handles the server browsing and game launching differently (ie only looking on the local network for servers, NOT calling home over the net to authenticate, etc). Again, this is just my experience based on the servers that we ran. Cheers Mike W www.gekidodesigns.com Javier Arevalo wrote: > mike wuetherick wrote: > >> after monitoring and calculations, we found that a default half-life >> (version 1) server is about 50 kb/s per player pretty much straight up >> to the max players, you can run about 12 players on a standard >> cable/dsl before you max out and start seeing lag. >> >> as another example, battlefield 1942 used almost twice the bandwidth >> per player, upwards of 100 kb/s per player... > > > Wow, those numbers sound awfully high. Your "Standard DSL" outbound > speed is *6* times what I have at home. :) At 50kb/s (6.25Kbytes/s), a > 56K modem would barely be enough to let someone connect to a server. > > Maybe those numbers were measured using a heavier "LAN performance" > option? (Not sure if those games support that, just wondering). > |
|
From: Noel L. <nl...@co...> - 2004-09-30 14:17:33
|
On Wednesday 29 September 2004 03:22 am, Javier Arevalo wrote: > At 50kb/s (6.25Kbytes/s), a 56K modem > would barely be enough to let someone connect to a server. Is anybody making games with 56K modems in mind anymore? I thought even the crappiest of broadband connections was assumed these days for most online games. --Noel Games from Within http://www.gamesfromwithin.com |
|
From: Javier A. <ja...@py...> - 2004-09-30 17:35:37
|
Noel Llopis wrote: > Is anybody making games with 56K modems in mind anymore? But the game we're talking about is 1998's Half Life, which many people played over modem lines. The requirements have increased with voice and whatnot, but still... -- Javier Arevalo Pyro Studios |
|
From: mike w. <mi...@ge...> - 2004-10-10 06:59:17
|
Seeing as it's pretty quiet around here lately, figured i'd post about an interesting set of articles recently posted discussing Game Design and the 'Big Picture'(tm): http://www.ludonauts.com/index.php/Developers:_Missing_the_'Whole'_Point Often it seems that most games are designed by dozens of individuals, and end up feeling this way - basically completely fail in the 'big picture' department... Any thoughts on the article? I personally agree with the author, in that games are unique in that every single piece of the development puzzle is, in fact 'a design decision', something that often has no input or control by the designer...or so it seems this way in the final product that is produced... Where in films, no matter how many hundreds of people were involved, often you hear people discussing how 'every frame is perfect', and how a film lives up to the directors' 'vision'... How do film's manage this when games often seem like there was no 'vision' to begin with in the first place (whether this is true or not is a different story)... Cheers Mike Wuetherick |
|
From: Brian H. <ho...@bo...> - 2004-10-10 14:51:22
|
> How do film's manage this when games often seem like there was no > 'vision' to begin with in the first place (whether this is true or > not is a different story)... Cultural differences, which will erode over time. In a movie, there is NO question that the director is charge. The actors, set designers, script writers, ALL respond to the director except in very anomalous situations (e.g. a star driven movie where the star has considerable weight with the studio). In games, no such similar structure happens. Fiefdoms and personal domains are far more common, and rarely does anyone have the authority to quash that. As games become more assembly-line oriented, this will naturally diminish. Budgets and teams are becoming far too big for the ad-hoc method of game design to work much longer. Some companies, such as EA, are getting this process down very well, much to the ire of the individual contributors. In Japan it's already this way to a large degree, e.g. the situations where Kojima or Suzuki are clearly the defining visionaries for their products, and everyone else is there primarily to assist, not to have their own miniature visions of how things need to get done. Brian |
|
From: Jamie F. <ja...@qu...> - 2004-10-11 10:04:11
|
It's not entirely true, of course. I'm sure your average film director doesn't go to the camera manufacturers and tell them how to build a camera. It's all about finding the right boundaries between skill sets. We're getting there, slowly.... Jamie -----Original Message----- From: gam...@li... [mailto:gam...@li...]On Behalf Of Brian Hook Sent: 10 October 2004 15:51 To: gam...@li... Subject: Re: [GD-General] Missing the 'Big Picture' > How do film's manage this when games often seem like there was no > 'vision' to begin with in the first place (whether this is true or > not is a different story)... Cultural differences, which will erode over time. In a movie, there is NO question that the director is charge. The actors, set designers, script writers, ALL respond to the director except in very anomalous situations (e.g. a star driven movie where the star has considerable weight with the studio). In games, no such similar structure happens. Fiefdoms and personal domains are far more common, and rarely does anyone have the authority to quash that. As games become more assembly-line oriented, this will naturally diminish. Budgets and teams are becoming far too big for the ad-hoc method of game design to work much longer. Some companies, such as EA, are getting this process down very well, much to the ire of the individual contributors. In Japan it's already this way to a large degree, e.g. the situations where Kojima or Suzuki are clearly the defining visionaries for their products, and everyone else is there primarily to assist, not to have their own miniature visions of how things need to get done. Brian ------------------------------------------------------- This SF.net email is sponsored by: IT Product Guide on ITManagersJournal Use IT products in your business? Tell us what you think of them. Give us Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more http://productguide.itmanagersjournal.com/guidepromo.tmpl _______________________________________________ Gamedevlists-general mailing list Gam...@li... https://lists.sourceforge.net/lists/listinfo/gamedevlists-general Archives: http://sourceforge.net/mailarchive/forum.php?forum_idU7 |
|
From: Javier A. <ja...@py...> - 2004-10-13 08:13:43
|
This is what I posted on my own weblog: I almost stopped reading when Walter Kim states that "It's my belief that the problem of story and game fusion is part of a larger problem in videogame development: object-oriented programming". Thankfully, he quickly clarifies that he's trolling his way to make a point, and ends with "Production pragmatics are gratuitously dictating the course of design before any design work is even done." Which is dead on, of course. He doesn't offer much in the way of enlightment, but he provides a very useful point of view. Recommended reading. The emphasis on experience, emotion (and crying as the ultimate proof) and storytelling are quite fashionable. But, while important, I doubt they're really the crux of the problem. It's easy to say "many games have bad narrative" and "many games fail", and then establish a correlation. But no. Many successful games do not have a meaningful story. Not every game needs one, or even benefits from having one. (Minor pet peeve: Facade keeps being brought up as a significant example of sucessful game narrative - I do not even consider it a game!) Now, specific to your question: As long as games are considered "software development", it will always be a bit like that. To me that's like saying "moviemaking is carpentry" because you have to build the sets. But in all honesty, the biggest problem is the lack of experience and maturity of the creative, technical and production leads, as well as the craft itself which still doesn't have a solid cultural background (terminology as well as truly established methodology). Anyway I posted on the forum there as well. mike wuetherick wrote: > Seeing as it's pretty quiet around here lately, figured i'd post about > an interesting set of articles recently posted discussing Game Design > and the 'Big Picture'(tm): > > http://www.ludonauts.com/index.php/Developers:_Missing_the_'Whole'_Point > > Often it seems that most games are designed by dozens of individuals, > and end up feeling this way - basically completely fail in the 'big > picture' department... > > Any thoughts on the article? I personally agree with the author, in > that games are unique in that every single piece of the development > puzzle is, in fact 'a design decision', something that often has no > input or control by the designer...or so it seems this way in the > final product that is produced... > > Where in films, no matter how many hundreds of people were involved, > often you hear people discussing how 'every frame is perfect', and > how a film lives up to the directors' 'vision'... > > How do film's manage this when games often seem like there was no > 'vision' to begin with in the first place (whether this is true or not > is a different story)... -- Javier Arevalo Pyro Studios |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X