Does decreasing the multi for higher FSB improve or decrease frame rates in games?

[Mussels]

If the final clock is the same the higher FSB will be faster, simply because all other subsystems are faster. If you actually notice this is different per case.

timings being the opposing factor, on many systems you have to loosen (or they do it automatically) several timings or FSB options. Quite often, they negate the speed advantages. Have a look earlier in the thread - many screenies have been posted showing negligible difference.

 

[farlex85]

timings being the opposing factor, on many systems you have to loosen (or they do it automatically) several timings or FSB options. Quite often, they negate the speed advantages. Have a look earlier in the thread - many screenies have been posted showing negligible difference.

The screenies of fps in games showed practically no difference at all. The memory benchies told a different story though, w/ the higher fsb winning fairly substantially.

 

[EastCoasthandle]

Using a 9x multi for a 400 FSB clocked at 3.60GHz,
Ram at 1066MHz
PL6

Using a 9x multi for a 400 FSB clocked at 3.60GHz,
Ram at 1066MHz
PL7

Using a 8x multi for a 450 FSB clocked at 3.60GHz,
Ram at 1080MHz
PL7

Using a 8x multi for a 450 FSB clocked at 3.60GHz
Ram at 1128MHz
PL7

The difference between PL6 and PL7 @ 400x9, running COH:OP is 1.2 FPS. While HL2:LC only saw a 0.90 FPS gain. Based on these minuscule results at 400x9 PL6 and PL7 the MCH is not hindering performance nor is performance increase hindered.

 

[Mussels]

The screenies of fps in games showed practically no difference at all. The memory benchies told a different story though, w/ the higher fsb winning fairly substantially.

which is all fine if you're encoding something, but pointless if you arent. Yes this matters to the benchmark whores, but to gamers or general users, it doesnt matter at all. I'd rather use lower voltages and less tweaks for a lower FSB, so that my system gets 99.5% the same speed yet lives a longer life.

 

[FR@NK]

The difference between PL6 and PL7 @ 400x9, running COH:OP is 1.2 FPS. While HL2:LC only saw a 0.90 FPS gain. Based on these minuscule results at 400x9 PL6 and PL7 the MCH is not hindering performance nor is performance increase hindered.

Thats interesting.....

Real-World Results: What Does a Lower tRD Really Provide?

Exactly how a system tends to respond to this increase in available bandwidth remains to be seen, as this is largely dependent on just how sensitive the application/game/benchmark is to variations in memory subsystem performance. It stands to reason that more bandwidth and lower latencies cannot possibly be a bad thing, and we have yet to encounter a situation in which any improvement (i.e. decrease) in tRD has ever resulted in lower observed performance.

.....Your system must not be limited by memory bandwidth when running those two games.

 

[HTC]

Maybe using crysis can yield you a result. Let me explain:

There's a load time prior to the game, right? Will the PL lvl affect the load time?

I saw somewhere (dunno where, in TPU) that more memory in the system has a direct impact on load time so maybe, just maybe, different PL lvls will too. You would probably have to use a stop watch, though: as far as i know, there isn't any prog capable of measuring game load times!

 

[EastCoasthandle]

Maybe using crysis can yield you a result. Let me explain:

There's a load time prior to the game, right? Will the PL lvl affect the load time?

I saw somewhere (dunno where, in TPU) that more memory in the system has a direct impact on load time so maybe, just maybe, different PL lvls will too. You would probably have to use a stop watch, though: as far as i know, there isn't any prog capable of measuring game load times!

I don't have Crysis demo installed right now but interesting idea none the less. I did notice a a minor decrease in load time going from 2GB to 4GB. But I am not sure about PL7 to PL6. As for BF2 with 4GB (XP recognizes 3.25GB) it was noticeably faster entering a server. Both deleted cache and saved cache yielded faster load times into the server.

OT:
If anyone is wondering "is it worth it to get 2x2GB knowing that the OP will only recognize 3.XXGB". I have to say yes, it is worth it to consider a 2x2GB kit (be it from Corsair, G.Skill, etc as 2x2GB kits usually are Powerchip ICs). I haven't really noticed any frame rate increases but I have notice applications like PS Elements, and most games (I've played) load a lot faster then before.

 

[Deleted member 3]

timings being the opposing factor, on many systems you have to loosen (or they do it automatically) several timings or FSB options. Quite often, they negate the speed advantages. Have a look earlier in the thread - many screenies have been posted showing negligible difference.

I never mentioned any other variables, the example assumed everything else would remain the same. You're overcomplicating, on the other hand you are right that you hardly ever can keep everything else the same. So with the simple answer being "FSB" you most likely require benchmarks to know the actual difference per case.

 

[EastCoasthandle]

Thats interesting.....



.....Your system must not be limited by memory bandwidth when running those two games.

Well, results can vary from one setup to another, from one game to another. I only wish more gaming developers included benchmarking tools in other games where more tests results can be made. Fraps IMO is a bit too cumbersome (having to gather data points, then convert them into a graph, etc). Grid demo, Medal of Honor: Airborne and Assassins Creed would have been great test candidates.

 

[niko084]

By general rule higher fsb means more performance because it affects more than your processor.
"BUT" only if your ram and such is faster, and you may or may not see the difference in a game.

 

[EastCoasthandle]

By general rule higher fsb means more performance because it affects more than your processor.
"BUT" only if your ram and such is faster, and you may or may not see the difference in a game.

Sure if you don't change the multi. But if you lower the multi is there a real gain in games? In the 2 games tested I didn't see it.

 

[niko084]

Sure if you don't change the multi. But if you lower the multi is there a real gain in games? In the 2 games tested I didn't see it.

Yes 3.2 @ 7x would be faster than 3.2 @ 8x for say.
Weather it is noticable depends on a lot and the difference may be really small.

 

[EastCoasthandle]

Yes 3.2 @ 7x would be faster than 3.2 @ 8x for say.
Weather it is noticable depends on a lot and the difference may be really small.

Well I did a comparison @ 3.6 using 8x and 9x multi in 2 games and results were within margin of error. If there is a game or program that actually shows a difference I am open for suggestions. What program/game did you use to come to that conclusion? Unless you are referring to specific memory benchmarks?

 

[niko084]

Well I did a comparison @ 3.6 unsing 8x and 9x multi in 2 games and results were within margin of error. If there is a game or program that actually shows a difference I am open for suggestions. What program/game did you use to come to that conclusion? Unless your are referring to specific memory benchmarks?

No its simple fact that a higher fsb affects more than just your processor so in turn it will indeed make more difference in speed.

Memory bandwidth is by far the largest thing you will see a difference in.
Like I said it is faster "technically", but weather you will ever see a difference worthy of calling a difference is another story.

It's like 2x1gb sticks of ram is "technically" faster than 4x512 sticks, but will you ever see the difference? Probably not.

Now if you took it a step further and dropped to 7x and still at 3.6, then you might see more evidence to the fact that its faster.

Best way to really test, would be to do the same test 10 times or more and take an average.

The way I see it beyond a straight memory benchmark, its not worth it to drop your multi if you need to raise any voltages.

 

[EastCoasthandle]

No its simple fact that a higher fsb affects more than just your processor so in turn it will indeed make more difference in speed.

Memory bandwidth is by far the largest thing you will see a difference in.
Like I said it is faster "technically", but weather you will ever see a difference worthy of calling a difference is another story.

It's like 2x1gb sticks of ram is "technically" faster than 4x512 sticks, but will you ever see the difference? Probably not.

Now if you took it a step further and dropped to 7x and still at 3.6, then you might see more evidence to the fact that its faster.

Best way to really test, would be to do the same test 10 times or more and take an average.

The way I see it beyond a straight memory benchmark, its not worth it to drop your multi if you need to raise any voltages.

Nothing is simple fact unless it's proven to be a fact first! Again, if you have a game that can be used to support this by all means share it with us so we can see for ourselves. You haven't shown us anything to substantiate this claim. Only implying common knowledge through the suggestion that its a simple fact using no evidence. IMO, so far it's not worth lowering the multi (based on my setup) for the same CPU frequency because it doesn't yield you anything.

 

[Silverel]

I got no differences using 16x200 and 13x247 for gaming...

 

[HTC]

For the FSB increase to be possible, your RAM must be able to tolerate it or you'll have to slow it down and, by doing so, you'll be defeating the purpose of increasing FSB.

To get an accurate reading on this, one must be able to increase the FSB without touching the RAM timings (suggest a low OC for this to be possible): this way, the only difference will be in the FSB and, by consequence, the RAM speed, which is the whole point!

 

[niko084]

Nothing is simple fact unless it's proven to be a fact first! Again, if you have a game that can be used to support this by all means share it with us so we can see for ourselves. You haven't shown us anything to substantiate this claim. Only implying common knowledge through the suggestion that its a simple fact using no evidence. IMO, so far it's not worth lowering the multi (based on my setup) for the same CPU frequency because it doesn't yield you anything.

Well obviously you need to lesson in computers and how they operate no offense.
It's fairly well known that a higher fsb is better than a lower fsb with a lower multi.
I'm very well also saying that its probably not going to be noticable to any real degree in anything other than a benchmark of various types, as said before memory being the main one.

Read a bit about how everything works off your fsb and you will come to the conclusion yourself that it is factually faster to have a higher fsb.

Remember you wont see any benchmarking difference unless you are using the extra speed on the bus for more than your processor also.

It's like expecting a higher 3dmark06 score with a P4 1.8 and a set of sli 8800ultras when you overclock the video cards, your not going to get jack more simply because the processor is holding you back.

 

[Silverel]

Well obviously you need to lesson in computers and how they operate no offense.

psst... that's offensive...

It's fairly well known that a higher fsb is better than a lower fsb with a lower multi.
I'm very well also saying that its probably not going to be noticable to any real degree in anything other than a benchmark of various types, as said before memory being the main one.

Read a bit about how everything works off your fsb and you will come to the conclusion yourself that it is factually faster to have a higher fsb.

Remember you wont see any benchmarking difference unless you are using the extra speed on the bus for more than your processor also.

It's like expecting a higher 3dmark06 score with a P4 1.8 and a set of sli 8800ultras when you overclock the video cards, your not going to get jack more simply because the processor is holding you back.

I guess the point of the thread was to find out if its worth upping your FSB for gaming. It's really not noticeable. My rig OC's better with a higher multi and lower FSB in MOST tests I've run. Not just Gaming. There's more than just basic understanding of what the FSB does for system performance going on here.

Results speak much louder than theory...

 

[EastCoasthandle]

Well obviously you need to lesson in computers and how they operate no offense.
It's fairly well known that a higher fsb is better than a lower fsb with a lower multi.
I'm very well also saying that its probably not going to be noticable to any real degree in anything other than a benchmark of various types, as said before memory being the main one.

Read a bit about how everything works off your fsb and you will come to the conclusion yourself that it is factually faster to have a higher fsb.

Remember you wont see any benchmarking difference unless you are using the extra speed on the bus for more than your processor also.

It's like expecting a higher 3dmark06 score with a P4 1.8 and a set of sli 8800ultras when you overclock the video cards, your not going to get jack more simply because the processor is holding you back.

I believe your problem is not your misdirection in saying I am not familiar with computer but your lack of evidence in games.

 

[HTC]

psst... that's offensive...



I guess the point of the thread was to find out if its worth upping your FSB for gaming. It's really not noticeable. My rig OC's better with a higher multi and lower FSB in MOST tests I've run. Not just Gaming. There's more than just basic understanding of what the FSB does for system performance going on here.

Results speak much louder than theory...

True, IMO. With my previous board (P5K-SE), i managed the same OC with lower voltage using an 8 multi as opposed to a 9 one.

In games, most people won't note any difference but the same can't be said by programs such as encoding programs.

EDIT

Please read post #42: i think it's the only way to prove or disprove this!

 

[niko084]

I believe your problem is not your misdirection in saying I am not familiar with computer but your lack of evidence in games.

Sorry if it was offensive... But now it sounds more like you are misunderstanding what I am saying..

"yes" its faster...

Is it worth it, no in almost all situations, and will you see a difference in a game benchmark, most likely not anything worth anything, maybe on 20-30 tests you may average .5-1 fps more maybe......

Now if I were to clock my ram to like 666 which would be 1:1 at 2.66, bench supreme commander "memory instensive game", and then drop my multi to 6x and ran my fsb to 443, to bring me back to 2.66 and give my ram 886, you would probably see a slight more definate difference. Further more if you dropped to 5x and 532 bus.

When I get home tonight, I'll do exactly that, I don't think I can push a 532 bus without quite a bit of work, but 443 will be easy.

 

[EastCoasthandle]

Sorry if it was offensive... But now it sounds more like you are misunderstanding what I am saying..

"yes" its faster...

Is it worth it, no in almost all situations, and will you see a difference in a game benchmark, most likely not anything worth anything, maybe on 20-30 tests you may average .5-1 fps more maybe......

Now if I were to clock my ram to like 666 which would be 1:1 at 2.66, bench supreme commander "memory instensive game", and then drop my multi to 6x and ran my fsb to 443, to bring me back to 2.66 and give my ram 886, you would probably see a slight more definate difference. Further more if you dropped to 5x and 532 bus.

When I get home tonight, I'll do exactly that, I don't think I can push a 532 bus without quite a bit of work, but 443 will be easy.

Less talk and more results would benefit this thread. But as I said before, based on my setup the lower multi, higher FSB does not make games faster. Although I didn't post any results I did try the aforementioned games (in a previous post of mine) and saw no difference.

 

[HTC]

Less talk and more results would benefit this thread. But as I said before, based on my setup the lower multi, higher FSB does not make games faster. Although I didn't post any results I did try the aforementioned games (in a previous post of mine) and saw no difference.

Did the memory timings suffer any change when you increased the FSB and lowered the multi?

 

[EastCoasthandle]

Did the memory timings suffer any change when you increased the FSB and lowered the multi?

No it didn't from what I observed. Except for when I used a PL6 instead of PL7 but I provided both. If you look at my results the timing didn't change.