10 hours ago, martin-w said:

 

Trust me, Westman DOES know. One of the most experienced round here. And tests a plethora of CPU's and motherboards. I've been doing this stuff for many years and would pay attention to what Westman has to say every time.

 

But of course, if you are happy with your rig you'd be foolish to upgrade.

That confused me a little bit for a moment but I realised what you meant after re-reading! In essence, Westman is saying the 7700k is a good choice. Although I'll be sure to follow him now as he does sound like he knows his stuff!

On 3/5/2017 at 7:57 AM, warbirds said:

Sorry but I am going with the guys that know. I get great fps with my 5930k at standard clocks and keep well up with a buddy that has a 7700k.

Even if the 7700k would give me 5 fps more it is not worth the expense or time to switch and waiting for a better cpu than the 7700k is the proper choice. 

It's not a 5 FPS difference.  I have a 5820k @ 4.4GHz and it gets average FPS in the high 40s/low 50s in FSXMark.  7700k @ 68.  Also the minimum FPS is much higher on the 7700k.  In the 50s.  On the 5820k its right around 40.  

3 minutes ago, TechguyMaxC said:

It's not a 5 FPS difference.  I have a 5820k @ 4.4GHz and it gets average FPS in the high 40s/low 50s in FSXMark.  7700k @ 68.  

Thanks Max, seen the same had 5820.5930 have a 5960x as second all day PC nowdays.

Not run any FSXmark with KabyLake but done some with 5960X and SkyLake and ton of other Cpus like you.

I run a 5930k with 1080gtx. My buddy runs a 7700k with a 1080gtx. With identical P3D setup and aircraft he gets about 5 to 7 more fps than I do. I just do not think it is worth the cost and trouble of buying and installing a new motherboard, new ram, new 7700k cpu and then buying a new water cooler to fit it.

I am running my 5930k stock and if I overclock it I might even keep up with my buddies system. 

I am putting my money on a new 1080ti instead. 

2 minutes ago, warbirds said:

I run a 5930k with 1080gtx. My buddy runs a 7700k with a 1080gtx. With identical P3D setup and aircraft he gets about 5 to 7 more fps than I do. I just do not think it is worth the cost and trouble of buying and installing a new motherboard, new ram, new 7700k cpu and then buying a new water cooler to fit it. 

Then he's doing it wrong.  Also, you spent signficantly more on your setup than he did.  A 5930k was a $600+ CPU when new.  X99 motherboards and quad channel DDR4 are signficantly more expensive as well.  You probably spent $500+ over your friend and you get lower performance.  You could fund that "downgrade" with the difference in cost alone.  

Lol, I bought my setup used off Chicago Craig's list. I paid 1500 for it and it had three 980 cards, two ss drives, 5930k cpu, 32 gig fast ram and Corsair 1000 power supply and case and water cooler. I eventually sold the three 980's and got the 1080 but all else is the same. I used to build my own pc's and now find I can get what I want already built much cheaper by going for a used year old high end machine. 

My buddy is a pc and net tech and is not doing anything wrong and we both have been doing this for twenty years now. 

I think going with at 1080ti is my best solution at this time and later, when cpu's get a bit more powerful I will update mine. 

6 minutes ago, warbirds said:

Lol, I bought my setup used off Chicago Craig's list. I paid 1500 for it and it had three 980 cards, two ss drives, 5930k cpu, 32 gig fast ram and Corsair 1000 power supply and case and water cooler. I eventually sold the three 980's and got the 1080 but all else is the same. I used to build my own pc's and now find I can get what I want already built much cheaper by going for a used year old high end machine. 

My buddy is a pc and net tech and is not doing anything wrong and we both have been doing this for twenty years now. 

I think going with at 1080ti is my best solution at this time and later, when cpu's get a bit more powerful I will update mine. 

The fact remains, retail prices of the setups you mentioned are worlds apart.  If you want to gamble on used hardware that's your choice.  I would never recommend that personally, you never know what kind of abuse the hardware saw.  

Anyway, price aside, your assertion that your setups perform similarly is based on what, exactly?  Did you both run FSXMark tests?  Can you provide the frame rate data to back it up?  If you're comparing unequal flight sim configurations who is to say that your results are even remotely comparable?  

This video tells it all,,keep in mind I play other games as well, not only P3D but shooters and adventure games.

 

Will there be an 'E' version of KL?  6-8 cores?

1 hour ago, warbirds said:

This video tells it all,,keep in mind I play other games as well, not only P3D but shooters and adventure games.

 

One person's opinion doesn't cover every scenario.  I play games too.  At 2560x1440 and 144Hz.  Which means I absolutely need the extra clock speed of Kaby Lake to get close to those frame rates.  All the recent Ryzen reviews show most games still run at higher frame rates on a highly-clocked 4 core chip rather than a lower-clocked 6-10 core chip.

39 minutes ago, Noel said:

Will there be an 'E' version of KL?  6-8 cores?

Not exactly.  Rumor is there will be a 4-core HEDT Kaby Lake chip called Kaby Lake X that has a higher TDP than Kaby Lake S (i.e. 7700k and lower models).  There will be 6+ core Skylake chips also similarly called Skylake X.  

4 minutes ago, TechguyMaxC said:

One person's opinion doesn't cover every scenario.  I play games too.  At 2560x1440 and 144Hz.  Which means I absolutely need the extra clock speed of Kaby Lake to get close to those frame rates.  All the recent Ryzen reviews show most games still run at higher frame rates on a highly-clocked 4 core chip rather than a lower-clocked 6-10 core chip.

Not exactly.  Rumor is there will be a 4-core HEDT Kaby Lake chip called Kaby Lake X that has a higher TDP than Kaby Lake S (i.e. 7700k and lower models).  There will be 6+ core Skylake chips also similarly called Skylake X.  

I'm convinced 6+ cores is the better option for improving smoothness w/ V3.x P3D, but that could change, or not, w/ V4.  Perhaps being able to pre-load the entire scenery path for a flight plan will obviate the need for terrain texture loading in realtime.  My 5 y/o SB-E at 4.422Ghz has served me so well now and it seems to still be doing quite well though I made a tiny bump in core voltage to remain stable at that clockspeed.   I'll keep it until it's truly died at which point hopeful V4 will be here and we will know what it needs.  At that point it will be top nV card and top processor, and another 32Gb of ram--I hope V4 can use at least that much!

11 minutes ago, Noel said:

I'm convinced 6+ cores is the better option for improving smoothness w/ V3.x P3D, but that could change, or not, w/ V4.  Perhaps being able to pre-load the entire scenery path for a flight plan will obviate the need for terrain texture loading in realtime.  My 5 y/o SB-E at 4.422Ghz has served me so well now and it seems to still be doing quite well though I made a tiny bump in core voltage to remain stable at that clockspeed.   I'll keep it until it's truly died at which point hopeful V4 will be here and we will know what it needs.  At that point it will be top nV card and top processor, and another 32Gb of ram--I hope V4 can use at least that much!

You can convince yourself of whatever you like.  Go read the FSXMark results thread.  ESP is not a well-threaded engine.  It has 1 heavyweight thread and the rest are just worker threads for texture pre-loading.  Clockspeed * IPC is king.  

16 games tested @ 1080p & 1440p with a Titan X Pascal:

http://www.techspot.com/review/1348-amd-ryzen-gaming-performance/page2.html

 

7700k is the fastest chip in 14 out of 16 games.  Only Mafia III and Watch Dogs 2 play better on 8-core chips.  

13 minutes ago, TechguyMaxC said:

You can convince yourself of whatever you like.  Go read the FSXMark results thread.  ESP is not a well-threaded engine.  It has 1 heavyweight thread and the rest are just worker threads for texture pre-loading.  Clockspeed * IPC is king.  

How well does FSXM document freedom from macro/micro stutters?  If it doesn't document this it's only getting 50% of the story.   We've all seen scenarios w/ more than ample frame rates yet w/o complete freedom from stutters of any kind, which is why I stated in terms of smoothness I'm convinced more cores is superior.   As well, I will be waiting to see what's happened in V4 P3D w/ regard to threading optimization.  Currently, when living w/in what my old SB-E chip can handle I experience complete freedom from even the slightest hint of a micro-stutter, hence I remain convinced.

You can use FRAPS to log frame time.  I don't get micro stutter on my 7700k.  Single-threaded performance is still the number one determinant of overall performance in any ESP-based application.