Clock speed determines how many calculations your processor can handle at once (or, how fast it can perform those calculations)...bus speed determines how fast your process can deal with input and output...

Now, for some applications, being able to do complex calculations very quickly is important...this would apply to say, an engineering or architectural program, which is where quad-cores become very handy...games also need this to do things like load a large map quickly...

Some applications require your computer to do many calculations very quickly, and this is where bus speed becomes important...if your game needs to keep track of the physics state for thousands of individual elements, that not only requires a power processor but also a good bus speed...

Up to this point, bus speed is usually not an issue as people find their gaming capabilities limited more by clock speed and the fact that games are not multi-threaded...if a computer game designer knows they only have one core to work with, then that severely limits what they'll do with the game and forces them to write efficient programming (in theory)...

However, if a programmer knows they have 2 or 4 cores to work with, then they don't have the same incentive to be efficient and there's nothing holding them back...for example, SC in sins will go right through objects, including planets...this was probably done because having to do pathfinding for hundreds of SC as they navigate around a planet would have required significantly more processing power, and so the game was designed otherwise...

Once games are multi-threaded, the limiting factor in processors will no longer be clock speed, but bus speed...with the advent of abundant processing power, there seems to be a growing trend in programming to rely on brute force methods instead of efficiency...to what extent this trend is occurring in PC games I'm not sure, but given the obvious lack of effort in many other areas, I wouldn't be surprised if gaming is victim to this trend...

If I'm a consumer choosing between a 2.0 GHz quad-core and a 2.8 GHz dual-core, I'm probably going to go with the quad-core...not only does it actually have more processing power but it's probably cheaper since the quality of each individual core is much less...unfortunately, that means the bus-speed of each core is probably less...

Even if it were the same, most games up till now have been designed for one core...multi-threading easily would allow a game to utilize four times more processing power...however, the bus speed still remains the same, which is where I see a potential problem...people go out and buy a cheap quad-core because they think its better only to find themselves limited by a different variable...

I don't know the specs of the Intel i7 but if they were designed for servers then their bus speed is probably twice that of most processors...normally I think processors have bus speeds of around 33 MHz so that Intel i7 is probably at 66 MHz...

The issue really isn't the hardware as much as it is consumer's tendencies and understanding of products as well as programmers being lazy and using brute force...

As for silicon technology, miniaturization is hitting its limits...a big limiting factor is "electronic noise" that forces smaller silicon designs to use significantly larger amounts of energy (obviously heat then becomes a big issue)...I don't know much about graphene but if it is a semi-conductor that conducts well at room temperature, then it could very well replace silicon...this is just a guess, but if it is found to be superior to silicon, I would imagine it is because it dissipates heat better...

I know they are working on more "biological" types of processing that try to imitate neural connections...unfortunately from what I've read, those designs only increase energy efficiency and excel at abstract applications like facial recognition...so far AFAIK they won't help with precision applications like a physics engine for a game

How did you come to conclusion that quality of cores on quadcore processor is lower than on dualcore? AFAIK they are pretty the same, sure you have many different models with different parameters like the size of cache memory for each core. In this regard, yes there could be dualcore CPU with some superior parameters compared to cheaper quads. The lower frequency of the quadcore compared to dualcore is not given by the lower quality of cores, but its the question of the CPUs TDP limit. Twice as much cores on the same frequency obviously need more power and have higher temp. This needs to be counterbalanced by lowering the frequency.

Regarding the bus, i assume you are talking about the bus connecting the CPU to the rest of your system - RAM, graphics card etc. On Intel CPUs up until i7 this bus was called FSB and run on the frequency of 266 later 333 MHz. For some reason ( i do not understand) this number was multiplied 4x, so the frequency was 1333 MHz. By overclocking the CPU you were actually increasing this speed as well. I suppose this system with the advent of multi-core CPUs started to be prohibitive because of the increased latency and low data throughput (i believe that is what you talking about).

Therefore Intel came with the i7 CPUs, which in fact were so much faster compared to Core 2 not because of advanced core design, but because of changes made to the "uncore" part of the chip and how is the chip connected to the rest of the system. Here i found something about it:

Basically with an "FSB" CPU (pre-i7 or pre-socket_754), everything goes through the FSB, which can reach up to 12.8GB/s on some CPUs, so when the CPU accesses the RAM or northbridge everything squeezes through the FSB

With a QPI(Intel) or HT(AMD) and integrated memory controller, the RAM has it's own dedicated connection of up to 21.3GB/s (AMD) or 38.4GB/s (Intel), PLUS the QPI/HT connection to the northbridge of up to 25.6GB/s (QPI) or 44.8GB/s (HT).

Basically the QPI/HT method can currently be up to about 5 times as fast as the "old" FSB way.

So as you can see huge advancement are made not only to the core design, increasing their numbers and frequency, but also to the bus, which connects them with the rest of the system. I assume those engineers at Intel/AMD are intelligent enough to see that its pointless to have hyperultra powerful CPU, if its bottlenecked by something else.

I probably agree with you about the programmers being lazy. So is the human nature anyway

Just let us know what features of Diplomacy you don't like or what options you don't want.

Please give us the ability to trade!! Ships, planets, culture?,...

Looking at the diplomatic options in elemental, thats what I am hoping for.

I'm sure whatever they do for Diplomacy will flow downhill to the other versions. It only makes sense. Though I do agree with the whole diplomacy needs a makeover along with the pirates. Just not sure if I agree with your reasoning seems how that is a choice and not a game issue.

Playing FFA with or without your friends in the game on a large map, diplomacy does go a long way to making the AI 'friendable'. I must say though that since the AI does not effectively use it's resources and has credits pile up, it is hard to win the Pirate war against them. It isn't all that much fun at times because of that. Also, the AI ALWAYS seems to choose to put bounty on the player first over other AI players.  Frustrating.

Well, i have a old motherboard with the FSB system... only difference is that i have TWO FSB at 10.7 GB/s who connect the processor to the Northbridge, good for a total of 21.4 GB/s ... Northbridge is connected by 4 serial bus to the main memory with a max speed of 32 gb/s... have very low latency because all is buffered... so, my old FSB computer is not so bad...

By the way, 1 QPI is 20 bit wide... Actual Xeon are 4 QPI ( 80-bit "flit" has 8 bits for error detection, 8 bits for "link-layer header," and 64 bits for "data" ) and can reach your 25.6 GB/s to the northbridge... unfortunaly, i7 have only one QPI to the north bridge ( 80-bit transferred in two clock cycles. four 20 bit transfers, two per clock ), good for only 6.4 GT/s in the case of a i7 extreme edition, 4.8 GT/s in the case of a usual i7...

The speed in your quote are for Xeon with 4 QPI ( multi processor ) and not for the i7 with one QPI...

i7 gulftown and bloomfield have 1 QPI but i7 lynnfield, clarksfield, Arrandale, Sandybridge ( Jan 2011 ) do not have a QPI interface but directly connects to a southbridge using a 2.5 GT/s Direct Media Interface and to other devices using PCI Express links... so, one i7 is not the other...

- 2nd Generation Intel Core i7 Processor ( no QPI ): http://download.intel.com/design/processor/datashts/324641.pdf page 10

- i7 extreme edition ( 1 QPI ): http://download.intel.com/design/processor/datashts/323252.pdf page 9

The number of transfers per clock cycle is dependent on the technology used. For example, GTL+ performs 1 transfer/cycle, EV6 2 transfers/cycle, and AGTL+ 4 transfers/cycle. Intel calls the technique of four transfers per cycle Quad Pumping. http://en.wikipedia.org/wiki/Quad_Data_Rate

"I would say for Sins2, get rid of those complicated relation ratios, diplomatic victory/ points and give players as TobiWahn suggested means to trade ships/planets/technologies/weapons."

Exactly. Diplomatic points and IMHO also culture and allegiance are more annoyances than game-enriching elements. Nothing is more frustrating than being able to colonize planets only few jumps away from your homeworld without building 3-4 culture centers in the orbit to prevent rebellion (and still getting just one third of income). I want to be able to utilize planetary bases of operations even deep behind enemy lines, if they let me, to support extended offensives and campaigns Only TEC with their Argonev construction upgrade allow me to do something like this now.

I have modded the cultural infuence on allegiance out (now decreasing only 2% every phaselane regardless of culture), and IMHO the game is far better without it, especially on large and multisystem maps.

I like Diplomacy's concepts, I just don't think they have worked out as well as the Entrenchment expansion.

Envoys

Envoy's are also not used often online, even in locked team games. Part of this is the high cost to fleet supply while maintaining envoy's. Conversely this is not a problem with Entrenchment and starbases as an Ally can help fortify an allies planet with a Starbase which does not have a negative affect on fleet supply.

In my opinion Envoy's should act very similar to the Entrenchment/Starbase concept with two parts to their goal. The first being to help establish relations for unlocked games, the second to build fortified Embassies that provide operation and tactical support (higher level abilities currently provided by Envoys).

Additionally Embassies could be upgraded to support more mid-late game support roles for an Ally (Fighter support for example). The upgraded options would need to be more for defense and not directly threaten the allies WEL for unlocked game scenarios. Several Envoy's already have planet bombardment type support that would fit nice on an Embassy. 

The Envoy AI also seems to linger in the host gravity wel sometimes instead of moving to allied planets. The AI should only need to build one AI per potential allied planet and then no envoy's once Embassies are established for Ally.

Pacts

The current pact system is alright but I would like to see some improvements to the pact system.

One specifically is modability. The current pact system is limited to the 18 alliance types currently offered. This makes it impossible to add new custom pact types for addon races. At a minimum it would be nice to have filler alliance types that could be utilized by modders.

Another problem with Pacts is they are simply Scalar improvements on mostly existing research upgrades. We didn't really get anything new with the Pact system. 

I'm not sure if trading ships/planets is feasible while maintaining balance, but I do know sharing technology could be possible. For example, Vasari phase pact would have been much better had it shared phase gate technology allowing other races to build phase gates (albiet without dark fleet upgrades). Biggest problem with sharing technology to build is the limited number of build slots (9 per category). Maybe a new tactical structure or starbase with hybrid technology or improvements to existing technology with hybrid upgrades.

Other thoughts on existing technology to share...

TEC sharing improved destabilization, repair bay upgrades, turret upgrades, ...

Vasari sharing phase gates, turret upgrades, ...

Advent sharing shield projection, turret upgrades, ..

@Frogboy Regarding Diplomacy:

Firstly, I always really appreciate that the developers here actually jump into the forums and ask for opinion and feedback. I have never joined an online community predicated on a game until Sins arrived. When I saw the opportunity for direct interaction with players, modders, and developers, joining the community seemed a worthy venture in co-creation.

A few points on Diplomacy:

Right of the bat, I did like Diplomacy. I thought it was a cool micro-expansion and touched on areas of the game I hoped would be added. But it was not quite exactly what I thought it would be. With a few of my RTS mates above I must agree. It was a bit disappointing that a trade system was not created for Sins similar to diplomatic options in GalCivII. It would have been great to be able to trade territory, tech, etc. I am not a programmer nor do I have experience balancing a game (a huge part of making sins what it is), so I am not certain if a trade element to Sins would be feasible, but this was a feature of Diplomacy I was hoping would be available. 

I did, however enjoy the "formula" or "score card" that added up all the elements that would make another player like or dislike you. I felt this added a great deal of character to the gameplay and made the relationships more believable. If I were to make any changes to this, I would only recommend that it be more difficult to form relationships or that perhaps the scoring be a variable in the game. For shorter games, the points add more quickly so that relationships can be formed before end of game, for longer games, relationships with a positive slant are harder earned.

There are a few elements of Diplomacy that I found didn't work quite as they may have been intended to. First, the diplomacy score card (which determines the overall breakdown of another player's relationship to you...I am not sure what the technical term is for it) seems to have a glitch. I noted that rejecting missions never seems to affect the "rejected missions" score. I have also never seen the neighbouring territory score change (although I may not have been paying attention on that one.) Secondly, the threshold at which you are offered pacts by other players appears to be too rigid. With the variable bouncing as frequently as it does, more often than not I have players offering and negating pacts on and off repeatedly over a short period of time as that number bounces over the line of the "yes" or "no" value regarding the agreement of the pact. This can get annoying and also detract from the immersion of the diplomatic experience. Thirdly, the pact visual layout is tricky. There isn't an obvious indication that you have been offered or are currently engaged in a pact in the primary diplomacy screen. Unlike the trade agreement or peace treaties that show a hand icon to indicate a pact, the "manage pacts" button gives no indication that a new pact has been offered or that one is in play. Often I found that I had been offered treaties without even realizing it. Finally, I found that the bonuses given by envoys was almost too great. When I first played Diplomacy, I was amazed at how much money I was gaining from TEC envoys, for example, or the rate the Vasari were increasing my mining output. As a Human vs AI, once that kind of diplomatic bonus was achieved and maintained I felt I was too easily able to overpower the other AI. 

Thanks, Frogboy. Great game, great fun. And thank you again for the invitation to comment. 

@ShotmanMaslo:

I have to disagree on this one. Removing these limitations on the game dissolves a whole dimension of strategy. Your comment is reminiscent of other suggestions I've seen to remove phase lanes completely or multiply the power of cap ships. Ultimately the game would no longer be a strategy game. If anything, the game needs more barriers to rapid expansion and construction. It's already too easy to build a massive fleet with only a handful of planets at your disposal.  

Today I used the diplomacy tree more than ever.

While it was nice to play an us vs them type game having from the start worked on building an alliance with another Vasari while regarding the other non Vasari AI races as enemies or at least non allies.

But having said that i hate giving up all that fleet capacity for the benefits. It to me seems to steep a cost.

More diplomacy options including affecting the attitude towards a 3rd party. ie Bribe another race to back off or become more aggressive towards a different 3rd party race

I never play PVP. So a good single player game and balance is very important to me.

One major problem is pirates hang around way to long for no profit and to the point of it becoming a suicide mission. I must say i totally agree with pirate scaling. I posted a suggestion in another thread about pirates. Give them a couple capital ships a bit later in the game as they use more fleet capacity with funny names like Blackbeard’s Planet Killer with a high planet damage or Sally Longshanks which has high missile damage pirates should be using ships from all the races and with unique capital ships (heavily modified "captured" capital ships)

Planetary Militia forces.

Distant Stars mod does a very interesting take on these forces at neutral planets with defense forces. They use ships from all three races, not just the TEC. Likewise, perhaps the pirates, truly being pirates, would be able to do this as well? I like the idea of militia forces as well as Pirate forces having a mix of all three of the races ships.