Jump to content

Some Performance Tests


295 replies to this topic

#41 Bill Lumbar

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 2,073 posts

Posted 21 December 2014 - 10:08 AM

View PostFlapdrol, on 21 December 2014 - 06:31 AM, said:

We're not talking "many games", we're talking THIS game, and in this game you're better off with the anemic pentium compared to any amd chip, because it only scales well to 2 cores (with settings low-ish). You can blame the code, but that doesn't improve performance.

And nobody is drinking "intel juice", amd's FX cpu's were a massive dissapointment for gamers, in most games the FX8150 was slower than the phenom II x4 and x6, because of reduced per core performance, and those phenom II's were only slightly faster than core2. With the higher uncore and clocks of the x3xx it's a bit faster, but still dissapointing. FX lauched after 2500K, and none of them match the 2500K when it comes to gaming. Competition in the cpu space is dead, the 4790K wastes space on an igp, but is still half the size of an FX, and sells for twice the price. :(

Look Frapdrol, I completely agree that Amd has had some pretty big flops in their attempts to bring us CPU's that can keep up and surpass Intel's line up. You guys are right in that aspect of what you have said, and are saying. This really is nothing new from Amd in the last 6 years. Case in point, look at the flop they gave us with the first generation Phenom CPU's, the flaws with them, and even a bug in the silicon that people had to install or what was it update their bios to deal with it. Its been so long now I forget what they called it....but it caused a serious performance drop. These first Phenom's was Amd's answer to Intel's C2D and Intels Quad line up. Amd failed that round too, and I would have to agree with you that those first Phenom's were a waste of time and money to buy. I remember I was one of the ones that waited to buy into the Phenom's until they fixed the bug and I picked up a 9850 i believe and later sold it off and upgraded to a Phenom II. My advice to anyone would have been skip them, don't buy them.

Then Amd brought us the Phenom II line, and IMO, redeemed themselves to a point. These Phenom II's wasn't cleaning house with Intel's higher end offerings, and I as much as we all would of liked them to, they did put up good numbers, and could over clock like crazy if one had proper cooling. I bought into that generation, 2 for myself, a 940 B.E. and a 955 B.E. both was very decent CPU's. I think one of Amd's biggest problems as a company is the hype that was sold to us all about the first Phenom's, and the hype we all put on them our selves, to be the new killer of Intel's CPU's and taking back the crown. This is the very same thing that happened with the 8100's series, BD, and when it was released there was even more hype put on the new BD's and as it turned out, they was a flop like the first Phenoms, and should be avoided.

While the **'s and the 8300's series never took the crown from Intel like many wanted to see, or even keep true pace with mid and higher end offerings from Intel, yet another blow to the Amd fan crowd, the 8300's line up is a decent processor for gaming, and many other things. However, it never was the I7 killer many wanted it to be, and felt it needed to be for Amd to stay in the race. What I find funny about the Intel 2 core you brought up is that it in fact also keeps up with I5's and I7's in gaming when over clocked to 4.7-4.8 ghz, so once again, that isn't saying much for the I5's and I7's or it is telling us a bigger picture of the lack of optimization this game and software in general has for ALL CPU's that are more then 2 cores. Its a joke, and once that catches up to the hardware, things should become much clearer on just how good or bad some of the CPU's really are, or are not.

I used to be into reading and studying the very clouded and long books of information on, bench marks, performance reviews, and used to visit Toms often to discuss hardware on their forums. This would of been around 4 years ago. One day after many long debates and threads regarding the Intel vs. Amd topic, which tends to bring out the strongest fan boys, I just walked away from it, and all the very long research it took to shift through the lines of BS, in the race for the "BEST" Cpu and the most fps money can buy. Coming back and seeing the talks on Amd vs. Intel here seems to look about the same as they was 4 years ago to me.

Goose made a very false statement regarding ALL Amd systems, and is telling all to beware. I am concerned with his claim, regardless of any issues Amd systems have, because what he has claimed is in fact BS. This is the kinda Intel Juice I am talking about among some, not all. This is what motivated me to reply in this thread and the other. I agree with you guys, I am sure you are seeing some higher numbers with your Intel rigs, depending on what they are, the specs, etc. etc. My point is that the FX-8350 I purchased runs this game to my expectations, and doesn't "gimp" me. If it did, do any of you really think with what I have said I wouldn't get online on the egg or head up to Chicago's Micro center on the west side, and pick up a new Intel CPU/mother board combo?

I don't know what is going on with the fraps log I submitted here, the 0's seem odd, and so do the dips into the 20's, and they only seem to happen when I first hit the F11 key. All I do know is that even with my ATI 7970 3gb and my FX-8350 at 4.5 ghz I am seeing smooth frame rates, within reason in this game, and not seeing dips on my fps counter in game lower then mid thirties unless I hit the F11 key to run the bench mark. I don't see tearing, any real stutters, or anything that anyone else isn't seeing, and from all reports and word we have all gotten, these issues are on Intel/nVidia rigs, as well as Amd/ATI rigs. Do the Intel rigs push higher fps, no doubt, I am sure they do.

But at the end of the day, my FX-8350 even close to stock clock and 7970 3gb provides me a good gaming experience in this game, and stomps any other game I have played on it with all settings to the max. What more could a gamer ask for? That really is the bottom line, unless you and others are chasing the highest Fps you can obtain in this game. All I can say is good luck with that, I have built computers ranging from $480 bucks all the way to $5000 for clients. In 6 months to a year, if all you care about is the highest scores on bench marks/synthetic tests scores, you will be shelling out another $800-2000 grand to stay at the cutting edge. I have found from my experience, its just not worth it to chase the hardware dragon. You will end up getting burned at some point, or go broke doing it.

About 4 years ago, I made the choice to buy hardware around a cycle or two behind the bleeding edge, this gives time for all the hype to wash out, and allows me to make sound choices when purchasing hardware for my gaming rigs. Plus, It save me a butte load of money on top of it all. ;)

Edited by Bill Lumbar, 21 December 2014 - 10:20 AM.


#42 LordKnightFandragon

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 7,239 posts

Posted 21 December 2014 - 10:11 AM

Wouch, so MWO is a $3K+ Hardware or poor FPS kinda game eh? Dang....High end people can get 45-60+.....

So when is PGI going to do optimization patches and other various performance tweaks to this game to kinda...help its performance on PCs that are not $3K+.

#43 Bill Lumbar

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 2,073 posts

Posted 21 December 2014 - 10:27 AM

View PostLordKnightFandragon, on 21 December 2014 - 10:11 AM, said:

Wouch, so MWO is a $3K+ Hardware or poor FPS kinda game eh? Dang....High end people can get 45-60+.....

So when is PGI going to do optimization patches and other various performance tweaks to this game to kinda...help its performance on PCs that are not $3K+.

The short answer is no, MWO is not a $3K gaming rig or poor frame rates kinda game. I would say it is a $800 gaming rig at least or experience a lower quality game, lower - mid settings, and just getting what most players would deem playable 25-30 fps in most games. $1200- 2000 grand should get you a system that can push 45-60 fps on this game. Anything higher in hardware and rigs pushing in the $3000 grand range is just gravy on top.

I still have some older AM2 AMD 6000x2 chips around, with older Micro atx boards with onboard Nvidia graphics laying around, I could always try to test them out with the 7970 or the 6870 or a 4870 and see what they can do in this game..... LOL, might be interesting. :lol:

Edited by Bill Lumbar, 21 December 2014 - 10:30 AM.


#44 xWiredx

    Member

  • PipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 1,805 posts

Posted 21 December 2014 - 11:22 AM

I completely decommissioned my Phenom II X3 720 BE when I built my Haswell-E system. My old 2600K took over its duties, so I can't test how it would have handled MWO anymore. I'm sure that going back in time to my X2 6400+, the architecture would be too weak to make this game actually playable. I think I still have a 9800GTX+ I could test with, but I dislike abysmal frame rates. Not even willing to test for the lulz.

#45 Bill Lumbar

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 2,073 posts

Posted 21 December 2014 - 11:54 AM

View PostxWiredx, on 21 December 2014 - 11:22 AM, said:

I completely decommissioned my Phenom II X3 720 BE when I built my Haswell-E system. My old 2600K took over its duties, so I can't test how it would have handled MWO anymore. I'm sure that going back in time to my X2 6400+, the architecture would be too weak to make this game actually playable. I think I still have a 9800GTX+ I could test with, but I dislike abysmal frame rates. Not even willing to test for the lulz.

I never got into the Phenom II X3 unless it was for a client wanting to unlock the 4th core for shits and giggles to get extra for less. I m not talking about breaking out the X2 6000 I have and throwing it with a older card, I think I have a 9800 also in a hackintosh build. My first build back when I started was a X2 6400 black and I loved that processor. It is still in service in my ex's fathers build that I sold him when I upgraded, and still chugging along just fine.

I am talking about maybe setting up a X2 6000 with the older Nvidia Geforce 4(I think) mother board with all its known issues and setting up a water cooling loop on the X2 6000 with a 7970 or 6870 card, drop 4-8 gb of 1066 ram in her, maybe with a SSD and OC the X2 6000 to all she can give and see just what this game can do to it. Could be interesting guys... to see what the fps are, and what results it would produce vs. newer CPU's.


I just cranked my FX- 8350 back up, but bumped up the V.core this time from 1.33 to 1.47 and bumped the multi up to 24 and am sitting on a 4.8 OC, no change to my 7970 3b stock clock. I have several web pages open, running MWO in test grounds, it is still open and running, and I am at 26C on my CPU, haven't seen it hit above 32C. Here is a fraps bench from the test grounds on Mining.

2014-12-21 16:01:32 - MWOClient
Frames: 17966 - Time: 180000ms - Avg: 99.811 - Min: 78 - Max: 118

I will run some in normal matches and in CW. I will also push my CPU back up to 5 ghz and beyond and see what difference it makes. Next will also be to bump up my 7970 and see what that gets me.

Wish me luck :D

These are from Crimson Strait, pretty heavy LRM fest, I managed to take high match score with 118 and 4 kills, 8 assist, and around 685 damage in my BJ-X1. This is with a 4.8 ghz over clock once again and stock clock on my 7970 gb.

2014-12-21 16:31:22 - MWOClient
Frames: 11226 - Time: 180000ms - Avg: 62.367 - Min: 36 - Max: 111

2014-12-21 16:35:44 - MWOClient
Frames: 11365 - Time: 180000ms - Avg: 63.139 - Min: 35 - Max: 117

Here is two from same match on Mining in normal game mode. I bumped up the 7970 3b to 1050 mhz and 1475 mhz, pretty small OC for what the 7970 is capable of doing. FX-8350 still at 4.8 Ghz. Once again I took high match score at 121, 3 kills, 6 assist and 747 damage in my BJ-X1.

2014-12-21 17:11:54 - MWOClient
Frames: 12051 - Time: 180000ms - Avg: 66.950 - Min: 31 - Max: 103

2014-12-21 17:15:35 - MWOClient
Frames: 12617 - Time: 174409ms - Avg: 72.341 - Min: 47 - Max: 109

These results are rather interesting and concerning. Match was on forest alpine, and a team mate ran into my mech and it happened several times as we went out of the spawn zone. Game play seemed a bit funky and maybe out of sync? Lots of LRMS and we got beat this one, still took top score.

2014-12-21 17:18:32 - MWOClient
Frames: 4847 - Time: 180000ms - Avg: 26.928 - Min: 0 - Max: 103 Forest Alpine beginning of match

2014-12-21 18:44:07 - MWOClient
Frames: 8630 - Time: 180000ms - Avg: 47.944 - Min: 24 - Max: 82 Forest Alpine mid way through match

This match was on the moon Lots of LRMS. Game play was smooth, and was able to take top score yet again.

2014-12-21 19:00:18 - MWOClient
Frames: 8732 - Time: 180000ms - Avg: 48.511 - Min: 29 - Max: 87

This one was in River city night, had night vision on the entire time and LRMS was very heavy. They base rushed us from the get go and fighting was thick and around the base and buildings. Ran into a few mechs on this one, and the common warping issues. Very intense!

2014-12-21 19:42:24 - MWOClient
Frames: 7212 - Time: 180000ms - Avg: 40.067 - Min: 18 - Max: 64

Edited by Bill Lumbar, 21 December 2014 - 04:51 PM.


#46 Goose

    Member

  • PipPipPipPipPipPipPipPipPip
  • Civil Servant
  • Civil Servant
  • 3,463 posts
  • Twitch: Link
  • LocationThat flattop, up the well, overhead

Posted 21 December 2014 - 04:15 PM

How do you play this thread, again?

Acid Bath Caustic in a Locust:
Posted Image

Seems underwhelming, right?
Posted Image

That looks more like it!

The Machine:
Posted Image
http://valid.x86.fr/zemaun

Teh Settings:
Posted Image

I guess I'm to the point of turning down the Object Details to see if I can unload ca_thread0Affinity. :-(

Edited by Goose, 22 December 2014 - 11:25 AM.


#47 Catamount

    Member

  • PipPipPipPipPipPipPipPipPip
  • LIEUTENANT, JUNIOR GRADE
  • 3,305 posts
  • LocationBoone, NC

Posted 22 December 2014 - 08:23 AM

View PostBill Lumbar, on 21 December 2014 - 10:08 AM, said:

Goose made a very false statement regarding ALL Amd systems, and is telling all to beware. I am concerned with his claim, regardless of any issues Amd systems have, because what he has claimed is in fact BS.


Goose's claim was not incorrect; you just misunderstood it. Goose is not saying that if you buy a "top notch" AMD system and leave it at stock clocks that you can't subjectively lower your expectations enough to enjoy any game with it. He puts that into disclaimers for builds he puts up alongside Intel, so his statement is simply that to perform in line with the expectations for an Intel system, ie the expectation of most gamers, an AMD system will have to be overclocked. Yes, if 20fps stutters are okay to you, and one can only imagine in horror how much lower that would be without your 4.5OC (15?!), then AMD is okay. Most of us do not consider that okay. Most gamers do not consider that okay.

His disclaimer is entirely honest. In fact, I think it's a case of excessive false balance to even include AMD builds into that thread, so be happy that he even goes that far.

Quote

I don't know what is going on with the fraps log I submitted here, the 0's seem odd, and so do the dips into the 20's, and they only seem to happen when I first hit the F11 key. All I do know is that even with my ATI 7970 3gb and my FX-8350 at 4.5 ghz I am seeing smooth frame rates, within reason in this game, and not seeing dips on my fps counter in game lower then mid thirties unless I hit the F11 key to run the bench mark. I don't see tearing, any real stutters, or anything that anyone else isn't seeing, and from all reports and word we have all gotten, these issues are on Intel/nVidia rigs, as well as Amd/ATI rigs. Do the Intel rigs push higher fps, no doubt, I am sure they do.


As I already explained, the 0s are because you ran the test too long and it ran into a loading screen. This suggests you test differently from what I would recommend. I start when my mech is powered up and end when the match ends because that's when all the system-killing action is going on. If you're accidentally creeping 0s in there then you must be trying to include the match start and end-score screens; do not do that (not performance relevant, skews the numbers)

As for your framerates, the 20-25 minimums are not "oh hey, suddenly I'm running at 20fps for ten seconds" type instances. They're stutters, like microstutter in a dual-GPU system, and hence you'll percieve them as stutters, not observably long periods of low fps. Most of us do not consider stuttering acceptable, which is why SLI and Crossfire have been largely looked upon with disdain in serious gaming circles for a long time, with serious setups usually including 3 GPUs instead of 2 (for some reason MS doesn't occur with 3, and I don't know why but it's a widely published fact). Now, of course, dual-GPU stutter has been resolved, but the point is serious gamers don't typically tolerate stutter if they can help it.

Quote

But at the end of the day, my FX-8350 even close to stock clock and 7970 3gb provides me a good gaming experience in this game, and stomps any other game I have played on it with all settings to the max. What more could a gamer ask for?


You could ask for a better gaming experience, like what an i5 will give you for the same price, moreso if you're willing to pay the incredibly modest difference between an 8350 and a 3570k or 4670k (both have been concurrent with Piledriver). You could ask for a gaming experience with higher average FPS, less stuttering, and less time spent in the CPU-dragging gutter (I get considerably higher minimums than you, but I get WAY higher averages than you, so I spend less time away from my minimums; just did some fraps runs yesterday, will post if you want)
I'm truly glad you get along well with your AMD system. Don't get me wrong, it's not a bad system. It is still a high-end gaming rig, and that 7970 especially is still doing serious work. I use a 7970 myself (Sapphire Dual-X). It's just that you will get a better rig, objectively better, in every way, with Intel, for about the same money. At least, that holds for gaming. There are a few software titles FX chips really like, and if you run those a lot, fine, get FX for a purpose-built system there, but none of those titles are games.

I would never tell you to ditch your system and go buy Intel right now, because your system is still pretty good. But it will be nothing short of total lunacy to suggest that a new buyer today purchase AMD Piledriver for a new gaming rig. Again, I'm not even sure why Goose includes them. I can only guess it's for those people who will buy AMD no matter what, and he's trying to suggest the best system then can get if they go that route, but honestly his disclaimer for AMD should say "Warning: will be objectively inferior to Intel systems in the same price range". If anything, his disclaimer understates the difference and shows a pro-AMD bias, and I say that with more than a little pro-AMD bias of my own.

Quote

About 4 years ago, I made the choice to buy hardware around a cycle or two behind the bleeding edge, this gives time for all the hype to wash out, and allows me to make sound choices when purchasing hardware for my gaming rigs. Plus, It save me a butte load of money on top of it all. ;)


But it's not about timing. Sandy Bridge is older than Piledriver. Ivy Bridge was released basically concurrently with Piledriver (2012), and although Haswell is newer, there's no notable price advantage in getting Piledriver today if someone is building a rig today. When you consider the grotesque power difference, and thermodynamics says what comes in must go out, and so consider the difference in cooling and PSU requirement for good OCing between the two, the price gap between an 8350 and 4670k is minimal at most. Even if you don't take those considerations into account, which is simply wrong, a 4670k is still insignificantly more expensive as a percentage of a total system build, and it will make the total system more effective.

Edited by Catamount, 22 December 2014 - 08:30 AM.


#48 Bill Lumbar

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 2,073 posts

Posted 22 December 2014 - 10:35 AM

View PostCatamount, on 22 December 2014 - 08:23 AM, said:


Goose's claim was not incorrect; you just misunderstood it. Goose is not saying that if you buy a "top notch" AMD system and leave it at stock clocks that you can't subjectively lower your expectations enough to enjoy any game with it. He puts that into disclaimers for builds he puts up alongside Intel, so his statement is simply that to perform in line with the expectations for an Intel system, ie the expectation of most gamers, an AMD system will have to be overclocked. Yes, if 20fps stutters are okay to you, and one can only imagine in horror how much lower that would be without your 4.5OC (15?!), then AMD is okay. Most of us do not consider that okay. Most gamers do not consider that okay.

His disclaimer is entirely honest. In fact, I think it's a case of excessive false balance to even include AMD builds into that thread, so be happy that he even goes that far.



As I already explained, the 0s are because you ran the test too long and it ran into a loading screen. This suggests you test differently from what I would recommend. I start when my mech is powered up and end when the match ends because that's when all the system-killing action is going on. If you're accidentally creeping 0s in there then you must be trying to include the match start and end-score screens; do not do that (not performance relevant, skews the numbers)

As for your framerates, the 20-25 minimums are not "oh hey, suddenly I'm running at 20fps for ten seconds" type instances. They're stutters, like microstutter in a dual-GPU system, and hence you'll percieve them as stutters, not observably long periods of low fps. Most of us do not consider stuttering acceptable, which is why SLI and Crossfire have been largely looked upon with disdain in serious gaming circles for a long time, with serious setups usually including 3 GPUs instead of 2 (for some reason MS doesn't occur with 3, and I don't know why but it's a widely published fact). Now, of course, dual-GPU stutter has been resolved, but the point is serious gamers don't typically tolerate stutter if they can help it.



You could ask for a better gaming experience, like what an i5 will give you for the same price, moreso if you're willing to pay the incredibly modest difference between an 8350 and a 3570k or 4670k (both have been concurrent with Piledriver). You could ask for a gaming experience with higher average FPS, less stuttering, and less time spent in the CPU-dragging gutter (I get considerably higher minimums than you, but I get WAY higher averages than you, so I spend less time away from my minimums; just did some fraps runs yesterday, will post if you want)
I'm truly glad you get along well with your AMD system. Don't get me wrong, it's not a bad system. It is still a high-end gaming rig, and that 7970 especially is still doing serious work. I use a 7970 myself (Sapphire Dual-X). It's just that you will get a better rig, objectively better, in every way, with Intel, for about the same money. At least, that holds for gaming. There are a few software titles FX chips really like, and if you run those a lot, fine, get FX for a purpose-built system there, but none of those titles are games.

I would never tell you to ditch your system and go buy Intel right now, because your system is still pretty good. But it will be nothing short of total lunacy to suggest that a new buyer today purchase AMD Piledriver for a new gaming rig. Again, I'm not even sure why Goose includes them. I can only guess it's for those people who will buy AMD no matter what, and he's trying to suggest the best system then can get if they go that route, but honestly his disclaimer for AMD should say "Warning: will be objectively inferior to Intel systems in the same price range". If anything, his disclaimer understates the difference and shows a pro-AMD bias, and I say that with more than a little pro-AMD bias of my own.


But it's not about timing. Sandy Bridge is older than Piledriver. Ivy Bridge was released basically concurrently with Piledriver (2012), and although Haswell is newer, there's no notable price advantage in getting Piledriver today if someone is building a rig today. When you consider the grotesque power difference, and thermodynamics says what comes in must go out, and so consider the difference in cooling and PSU requirement for good OCing between the two, the price gap between an 8350 and 4670k is minimal at most. Even if you don't take those considerations into account, which is simply wrong, a 4670k is still insignificantly more expensive as a percentage of a total system build, and it will make the total system more effective.

The only thing I am gonna shoot down and comment on with this atm is the fact that Goose did in fact claim ALL AMD SYSTEMS will need to have Config files modded and OC'ed like crazy. This is false... and is not a fair statement at all. I did not misunderstand what he claimed, and it is false. Thank you for showing your bias even more by saying,

" In fact, I think it's a case of excessive false balance to even include AMD builds into that thread, so be happy that he even goes that far."

I believe the thread topic was not Intel only builds for $500-1000 + or - 100 ;)

Ok, I have now read your entire post. And I agree with you on the majority of what you have said, other then what Goose claimed. If I was to spec out a new system today, I would with out a doubt go with a newer CPU from Intel over what Amd has out atm. Until Amd release's something different then they have out now, since I already have their "top of the line" CPU, I wouldn't build a new rig ground up on their FX Platform. However, there are some gamers I am sure that have AM3+ boards out there right now, that could depending on their mother board drop a FX-8350 in, and have a very nice upgrade for around $160 bucks. IF they are not in the market for a new rig or can't afford one right now, this is a legit option for them.

As to my method of testing with fraps, I wait until the match starts, after I am powered up, and start walking, I test for 180's, and the only time I notice my frames drop down is when I hit the F11 key, then they shoot back up. I do not see any stutters at all while doing this, and even when I don't do the bench, and I can keep a eye on the fps counter, I never see it dip below mid 30's in normal matches. In CW drops, I have seen it dip as low as 25 and not for 10 seconds..... just a very brief second, and shoots right back up. I also take a run from mid match, once the fighting really gets heavy. I try to stay away from benching on the loading screens or the end of match screes. IF I don't bench using the F11 key and just keep an eye on the fps counter, I see ranges in normal matches from 35-120's and no stutters. In CW I have seen ranges from 25 low on the counter to highs of 120's also, but in both modes I hover around 45-75 on avg through out the match, once again, no real stutters seen from low fps rates.

Once again, I am not seeing any micro stutters, tearing, or while turning my torso left to right any slow downs at all. If I zoom in and out, or turn on night or thermal I do not notice hits or slow downs. The only time I notice things get wonky is when making contact with another mech, and then at times I will warp, and game play seems to have to "adjust" to the warp from hitting another mech. I do not believe this has anything to do with my rig.... this is a issue with MWO, lack of a polished game. Would you not agree on this?

I am not bragging or trying to be cocky, but last night out of around 15 played in my BJ-X1 in solo cue, I was able to take top match score, hitting as high as 140's and getting 3-5 kills, 6-10 assists, and steady 600-900 damage in 11 out of 15 games. How does taking top score 11 out of 15 matches happen if I am running a gimped system as Goose has claimed, how could I put out steady match results if my rig was stuttering and dipping so low effecting my game play experience?


I have also seen post, and on player in the last town hall meeting in fact that is running a very nice Intel I7 or I5 rig, he claimed he is seeing dips into the teens as well with his Intel system, much like the ones you guys are on here claiming has a smoother feel and better gaming experience then playing on a Amd system.... How do you explain this. He is not the only one that is experiencing this for sure.... explain please.

Edited by Bill Lumbar, 22 December 2014 - 11:10 AM.


#49 Oderint dum Metuant

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 4,758 posts
  • LocationUnited Kingdom

Posted 22 December 2014 - 10:46 AM

View PostBill Lumbar, on 22 December 2014 - 10:35 AM, said:

The only thing I am gonna shoot down and comment on with this atm is the fact that Goose did in fact claim ALL AMD SYSTEMS will need to have Config files modded and OC'ed like crazy. This is false... and is not a fair statement at all. I did not misunderstand what he claimed, and it is false. Thank you for showing your bias even more by saying,

" In fact, I think it's a case of excessive false balance to even include AMD builds into that thread, so be happy that he even goes that far."

I believe the thread topic was not Intel only builds for $500-1000 + or - 100 ;)



Except its not false.

For christ sake, an AMD chip posts worse Minimum framerates than an Intel chip 3/4 it's weight. That's just how it is, because of how AMD developed its chips and PGI's code for the game.

AMD's poor performance in this game is well documented, in 100's of threads in this forum!

#50 Bill Lumbar

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 2,073 posts

Posted 22 December 2014 - 11:16 AM

View PostDV McKenna, on 22 December 2014 - 10:46 AM, said:



Except its not false.

For christ sake, an AMD chip posts worse Minimum framerates than an Intel chip 3/4 it's weight. That's just how it is, because of how AMD developed its chips and PGI's code for the game.

AMD's poor performance in this game is well documented, in 100's of threads in this forum!

I have already established that what Goose has claimed does not apply to ALL Amd systems, I have not touched any of my Config files, this is a fact. This makes what Goose claimed false right out the gate. I have tested my system on a lower clock, so no, my system doesn't need to be OC like crazy, game play was still smooth and very playable. While what Goose has claimed could very well be true for other Amd CPU's, it does not apply to ALL Amd systems.

Once again,

Why in the town all meeting was their a player, with a $1500-2000 dollar Intel system claiming he had severe issues with fps dropping down into the low teens, even single digits at times? There have been others in these very threads, with very respectable Intel rig, same issues.... please explain how this could happen, and why?

Edited by Bill Lumbar, 22 December 2014 - 11:17 AM.


#51 Oderint dum Metuant

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 4,758 posts
  • LocationUnited Kingdom

Posted 22 December 2014 - 11:39 AM

View PostBill Lumbar, on 22 December 2014 - 11:16 AM, said:

I have already established that what Goose has claimed does not apply to ALL Amd systems, I have not touched any of my Config files, this is a fact. This makes what Goose claimed false right out the gate. I have tested my system on a lower clock, so no, my system doesn't need to be OC like crazy, game play was still smooth and very playable. While what Goose has claimed could very well be true for other Amd CPU's, it does not apply to ALL Amd systems.

Once again,

Why in the town all meeting was their a player, with a $1500-2000 dollar Intel system claiming he had severe issues with fps dropping down into the low teens, even single digits at times? There have been others in these very threads, with very respectable Intel rig, same issues.... please explain how this could happen, and why?


Take a look at your minimum frame rates, and everything goose says is entirely correct.

In the town hall i didnt listen to, was this player talking about CW maps? that are totally un optimized and rushed out to keep the community happy and not miss a deadline or regular drops?

Because in regular drops Intels rock this game with higher minimum FPS overclocking K series chips nets a significant gain as well, in CW everyone is in the same boat.


Catamount is one of the staunchest AMD nuts you'll ever find, he used to spend days arguing for FX series chips.
And he still is an AMD nutjob. But even he admits, in real world MWO they just don't compete.

Edited by DV McKenna, 22 December 2014 - 11:44 AM.


#52 SixstringSamurai

    Member

  • PipPipPipPipPipPipPip
  • The 1 Percent
  • 930 posts
  • Twitch: Link
  • LocationYou Guys Are So Bad I'm Moving To The Moon

Posted 22 December 2014 - 12:09 PM

i5 4690k Devil's Canyon at stock 3.5
EVGA Geforce GTX 660 SC+
Gskill Sniper 9-9-9-24 1600 (2x4GB)
MSI Z97-G45 Gaming
Vizio 45" television as my monitor through HDMI

Pulling about 60-180 fps on average with occasional brief drops to 47 fps that last usually no more than a few seconds.

Edited by Sicarius Miyamoto Winters, 22 December 2014 - 12:11 PM.


#53 Bill Lumbar

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 2,073 posts

Posted 22 December 2014 - 12:38 PM

View PostDV McKenna, on 22 December 2014 - 11:39 AM, said:


Take a look at your minimum frame rates, and everything goose says is entirely correct.

In the town hall i didnt listen to, was this player talking about CW maps? that are totally un optimized and rushed out to keep the community happy and not miss a deadline or regular drops?

Because in regular drops Intels rock this game with higher minimum FPS overclocking K series chips nets a significant gain as well, in CW everyone is in the same boat.


Catamount is one of the staunchest AMD nuts you'll ever find, he used to spend days arguing for FX series chips.
And he still is an AMD nutjob. But even he admits, in real world MWO they just don't compete.



I am not disputing that Many Intel rigs will put out higher fps then Amd rigs, this is obvious and a fact. However, I do not base my game play experience of a little counter in the corner of my screen, and get bent if I see it drop to 25 fps as long as my game play is smooth, no stutters, tearing, or issues that gimp my game play. So at the end of the day, the numbers don't mean much to me, as long as my rig gives me a good game play experience in any given game. I am very picky, but I also stopped chasing the best numbers I could get with given hardware several years ago if I find hardware that can run a game that I like well enough. Would I like to get the highest fps I can in a given game?...sure I would as would anyone. If my rig handles games at 40-70 fps per avg. I am not gonna go buy or chase the next best thing to get more fps, IMO, its a waste of time and money. Your opinions might vary from mine, and if that is what you like to do with your time and money, that is just fine by me. ;)

I do not agree with what Goose claimed, because I have a Amd rig that doesn't require the tweaks he has claimed to have a very decent gaming experience in this game. What he has claimed given what I have shown with my rig is simply false. Not all Amd rigs need to be tweaked to not be gimped while playing this game, this is a fact.

I believe the player that asked Russ the question, with many chatters chiming in on the chat channel that also had Intel rigs, claimed mostly the really bad fps drops in the teens and even single digits was in CW drops, however, several also chimed in that it was in normal drops as well, but only into the low to mid 20's. They all agreed on one thing though, it was BS to have this kind of drops on any rig that is of higher end on the hardware scale. I also agree with them on this issue, it is pretty messed up.

Edited by Bill Lumbar, 22 December 2014 - 12:42 PM.


#54 xWiredx

    Member

  • PipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 1,805 posts

Posted 22 December 2014 - 12:45 PM

If 25fps is smooth to your eyes, well, I'm sorry about your eyes. We can fix them, we have the technology. I started noticing really bad stutters in CW before they started clearing out the dead mechs and making the smoke dissipate. Checked out the old F9 button, and sure enough I was hitting low 40s/high 30s. I have now idea how 25 is fine for you but 40 is abysmal to me.

#55 Bill Lumbar

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 2,073 posts

Posted 22 December 2014 - 12:56 PM

View PostxWiredx, on 22 December 2014 - 12:45 PM, said:

If 25fps is smooth to your eyes, well, I'm sorry about your eyes. We can fix them, we have the technology. I started noticing really bad stutters in CW before they started clearing out the dead mechs and making the smoke dissipate. Checked out the old F9 button, and sure enough I was hitting low 40s/high 30s. I have now idea how 25 is fine for you but 40 is abysmal to me.

How are you coming to the conclusion that I am getting 25 fps avg? If my rig dips to 25 for a split second or two, and pushes back up to 45-80 and mostly stays in the 40's-70 range and I have smooth game play, what difference does it make? To you it might, and that is fine, to me it makes no difference when I have a rig that puts out fairly smooth game play with out having to tweak it as Goose has falsely claimed I would have to.

Seriously, that would be like me claiming that your rig, (the deathmachine) which hands down is a very nice setup, is only getting 34 fps because it dipped down to that for a split second, and telling you that your system is gimped because you are only getting 34 fps. Its laughable at best.

Smooth game play in my mind has nothing to do with a little counter in the corner of my screen, however, if we are going to use that as a measure of smooth game play, I would like to see no less then 40-80 fps avg. and not see stuttering, slow downs when I zoom, switch vision modes, or while torso twisting, or taking fire from mgs, LRM's of other effects in this game.

That is how I define my gaming experience in a given game, if I see any of these slow downs, or stutters in a given game, its time to start looking for new hardware, or figure out ways to max out my current hardware if I safely can. Its really that simple, and I haven't had to go with Intel to do this yet, which makes Goose's statement completely false as he implied and stated, "ALL" Amd rigs are gimped right out the gate in regards to having a good experience in playing this game with out major tweaking. This is simply false.

Edited by Bill Lumbar, 22 December 2014 - 01:11 PM.


#56 xWiredx

    Member

  • PipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 1,805 posts

Posted 22 December 2014 - 01:41 PM

I never said average. Did I say average? I didn't say average. You, sir, love to put words into peoples mouths and make assumptions. The -real- hardware guys in this forum don't do that, and that's why we are calling you with all of the weird misinformation you've been spouting. We like numbers, shocking right? That geeks like numbers? Yeah, we like the hard evidence. We don't like the assumptions and opinions.

When you show, and say, that you have quick dips in the 20s and you don't "get bent" by those stutters (it actually sounds kind of like you don't even notice them), that's what I was referring to. Any dip that far down I would definitely notice, I used to have them sometimes on my old system when I cranked everything, and it drove me batty (hence the new system).

So what have we learned here so far? To recap:

Intel is faster
AMD is slower
MWO likes faster

From these very obvious, basic conclusions, we can surmise: Intel+MWO is better than AMD+MWO.

Dips in the 20s is not similar to dips in the 40s. The averages you have are nowhere near the averages I have. You do not have the same or even a remotely similar experience statistically-speaking in MWO. 65 is not similar to 85, and 48 is WAY off.

We will have no more misinformation. No thank you. Just stop until you are no longer in denial about your bulldozer/piledriver performance.

#57 Oderint dum Metuant

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 4,758 posts
  • LocationUnited Kingdom

Posted 22 December 2014 - 02:09 PM

I would just like to point out, in non CW drops so removing the buggy mess those are.

My Dual Core haswell refresh budget chip. achieves better minimum frames at 1080P medium settings.

Now that's at medium, consider your settings, and the fact you have 6 more "cores".

#58 Catamount

    Member

  • PipPipPipPipPipPipPipPipPip
  • LIEUTENANT, JUNIOR GRADE
  • 3,305 posts
  • LocationBoone, NC

Posted 22 December 2014 - 02:37 PM

View PostBill Lumbar, on 22 December 2014 - 11:16 AM, said:

I have already established that what Goose has claimed does not apply to ALL Amd systems, I have not touched any of my Config files, this is a fact.


and you get Abysmal framerates rife with low averages and stutter. I'd say that validates Goose well enough.

Just because you accept sub-par performance and have adapted to dealing with it doesn't make it not sub-par, and you have an overclocked system. To get performance that isn't sub par would additionally require messing around with the game a lot more than you do (dropping particles might not hurt either, though that's appearing to be OS dependent).

View PostDV McKenna, on 22 December 2014 - 11:39 AM, said:

Catamount is one of the staunchest AMD nuts you'll ever find, he used to spend days arguing for FX series chips.
And he still is an AMD nutjob.


Posted Image

Edited by Catamount, 22 December 2014 - 02:38 PM.


#59 xWiredx

    Member

  • PipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 1,805 posts

Posted 22 December 2014 - 02:55 PM

View PostDV McKenna, on 22 December 2014 - 02:09 PM, said:

I would just like to point out, in non CW drops so removing the buggy mess those are.

My Dual Core haswell refresh budget chip. achieves better minimum frames at 1080P medium settings.

Now that's at medium, consider your settings, and the fact you have 6 more "cores".


I'm sure if I turned down particles to medium I could keep my head above 60 at all times. I'm not sure I'm willing to sacrifice eye candy, though. I'll test it for a couple matches. I don't have a ton of time on my hands at the moment, but I think it might be worthwhile to test each setting individually by turning it down and seeing how much of a boost I get. Admittedly, I will stick with all very high because it is very rare that I see anything below 60. The majority of my fps, like 95% or more, is above 60.

Of course, testing will have to be with the 32-bit client since the 64-bit one crashes or won't run at all when I fire up FRAPS. I have had zero success running them at the same time.

#60 Catamount

    Member

  • PipPipPipPipPipPipPipPipPip
  • LIEUTENANT, JUNIOR GRADE
  • 3,305 posts
  • LocationBoone, NC

Posted 22 December 2014 - 03:17 PM

Windows TP doesn't have the particle problem, just saying :)

I'll have to test FRAPS/64-bit myself here later





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users