Good score! It seems that PCIe3 hasn't done much of anything for FSX. Long live SB!

 

Ben don't you think maybe what it might do is add smoothness, freedom from micro stutters? I would think any gains might be in this domain. Just a guess though!

Ben don't you think maybe what it might do is add smoothness, freedom from micro stutters? I would think any gains might be in this domain. Just a guess though!

 

I wouldn't think so. Through the vsync test, it has been proved that microstutters are not the thing of a CPU or FPS, but rather something else. Call it a driver...

The only hope now that we might see was advantage of more throughput on the video side for autogen, but that didn't help, as obviously PCIEx3 didn't help for FPS at all.

obviously PCIEx3 didn't help for FPS at all.

 

We need some more testing in super dense autogen before we jump to that conclusion.

Ben don't you think maybe what it might do is add smoothness, freedom from micro stutters? I would think any gains might be in this domain. Just a guess though!

 

Nope, it's all in the numbers, and nothing but the numbers, so help me geek.

I wouldn't think so. Through the vsync test, it has been proved that microstutters are not the thing of a CPU or FPS, but rather something else. Call it a driver...

 

I would agree with something else, but I am suggesting the something else is a timing issue. Not a CPU issue, not an FPS issue, but a timing/sycronization issue. This would go far to explain why specific systems perform differently, why a lower end box like mine may be freer from micro stutters in some situations than experienced by folks w/ more potent platforms. Since there are a near infinite number of combinations of clock speed, memory bandwidth, memory latency, FSB speed, drivers, etc it is compelling to look at system timing as relevant. I think this is the basis of FSGS's tuning concept.

Noel,

 

There have been a lot of people over the years refer to another indefinable coefficient or attribute called smoothness, or some similar concept, but in the end, it is all in the speed, as expressed in the regular frequency of FPS calculations, and nothing else. The sequence, spacing and rate of frames rendered accurately per second defines, describes, qualifies and quantifies smoothness as far as the machine and science is concerned. Any such esoteric concept above or beyond that is only in the imagination of the beholder.

 

Kind regards,

Noel,

 

There have been a lot of people over the years refer to another indefinable coefficient or attribute called smoothness, or some similar concept, but in the end, it is all in the speed, as expressed in the regular frequency of FPS calculations, and nothing else. The sequence, spacing and rate of frames rendered accurately per second defines, describes, qualifies and quantifies smoothness as far as the machine and science is concerned. Any such esoteric concept above or beyond that is only in the imagination of the beholder.

 

Kind regards,

 

Just because you haven't defined 'smoothness', or 'freedom from micro stutters', or 'micro stutter rate' that doesn't mean it's esoteric. It could be defined. The external frame rate limiter is an example of starting w/ raw horsepower, then controlling surge/slow thru a gate. Nothing esoteric about it, and many have concurred there was real value in how you set the sim up to run on your particular platform. Relegating the entire experience to 'all in the speed' is not accurate I'm afraid. It certainly helps to start w/ raw speed, but then there's the tuning.

Relegating the entire experience to 'all in the speed' is not accurate I'm afraid.

 

Sure it is. Like I said:

 

"The sequence, spacing and rate of frames rendered accurately per second defines, describes, qualifies and quantifies smoothness as far as the machine and science is concerned." How you get there is by tuning, testing, tweaking and for a near lifetime, or so it seems. Noel, you are singing to the choir. Most everyone here is a master at it.

 

Point is, there is no extra "smoothness" or mystically better experience to be found in an IB doing the same frames as a SB. It's in the evenly spaced, properly rendered sequential frame count that brings the special magic, nothing else, and that accomplishment is as a result of hard earned knowledge learned and applied, first in Sandy, and now Ivy.

 

Anyway we are saying the same thing, but in a different way.

 

Kind regards,

in the end, it is all in the speed, as expressed in the regular frequency of FPS calculations, and nothing else

 

Sorry to step in but the FPS counter indicates the average number of frames per second, not the distribution of frames within each second.

 

That is the sense in which I think some people experience stutters which are "micro": by which they mean, lack of even frame distribution within the second.

 

Tim

That is the sense in which I think some people experience stutters which are "micro": by which they mean, lack of even frame distribution within the second.

 

I agree completely. That's why I wrote:

 

It's in the evenly spaced, properly rendered sequential frame count that brings the special magic, nothing else,

 

...all of which is a result of machine capability as designed and directed by each of us.

 

Ah, nevermind... :hi:

Sure it is. Like I said:

 

"The sequence, spacing and rate of frames rendered accurately per second defines, describes, qualifies and quantifies smoothness as far as the machine and science is concerned." How you get there is by tuning, testing, tweaking and for a near lifetime, or so it seems. Noel, you are singing to the choir. Most everyone here is a master at it.

 

Point is, there is no extra "smoothness" or mystically better experience to be found in an IB doing the same frames as a SB. It's in the evenly spaced, properly rendered sequential frame count that brings the special magic, nothing else, and that accomplishment is as a result of hard earned knowledge learned and applied, first in Sandy, and now Ivy.

 

Anyway we are saying the same thing, but in a different way.

 

Kind regards,

 

Stephen, the entire discussion began when I posed the question to Ben, who commented, "It seems that PCIe3 hasn't done much of anything for FSX . . ."

 

You give a workable definition of smoothness which is quite reasonable: It's in the evenly spaced, properly rendered sequential frame count that brings the special magic, nothing else.

 

And prior to this reasonable comment, you state: ". . . in the end, it is all in the speed, as expressed in the regular frequency of FPS calculations, and nothing else."

 

I'm only suggesting perhaps PCIe 3 could impact the in your words, that special magic, and in Word Not Allowed's, something special. Ben's comment was that since FPS wasn't impacted, PCIe 3 was not meaningful, or something along those lines.

 

Ahh . . . never mind!

Ahh . . . never mind!

 

LOLOLOL! :friends:

Finally got FSX setup last night, been a long long week, just wrapped up my best friend's wedding so now I have time to start messing w/tweaking the system, overclocking, and testing FSX.

 

Quick, initial impressions:

 

SSDs are great, load times are tremendously reduced in everything, including FSX. I can turn my computer on and be in the air in under a minute. Granted, we're not talking about Orbx PNW + PMDG 737 NGX yet, but that will all be loaded in short order.

 

Ivy Bridge: mixed bag. Super fast, undeniably. Media transcoding is one of two primary purposes of this build so even if FSX performance ends up being slightly less than Sandy Bridge I'll still be happy, and it will still be one of the fastest FSX machines money can buy ;) Temps are high, even at stock. IBT & OCCT temps at stock are in the low to mid 50's peaking around 60 every so often on an H100, definite downer. I'm shooting for an OC north of 4.6, if I hit that it's about the same as a 4.8 SB which is plenty fast. We'll see how things run once a PCI-e 3 GTX 680 is added to the mix, hoping to smooth out those super demanding high autogen scenarios that were never smooth on any previous system for me.

 

Will post more later, off to tweak now!

Stock FSX Mark 11 results: 14, 32, 22. Pretty crappy, but I blame the inadequate graphics card which is only here to tide me over until a suitable GTX 680 can be sourced.

Ewww, poor results for such a mighty chip. Waiting for that 680!

 

hoping to smooth out those super demanding high autogen scenarios

 

Yes, would like to know everything about that!