Will V5 Performance be still single thread dependent?

[pmb 5,137]

3 hours ago, Egbert Drenth said:

What I meant was is that XP11 (or pick another sim) has been build (more or less) from scratch with the latest technologies and also can brought to its knees.

I tend to disagree here. Although it's hard to find X-Plane's very roots, it's as old as the MSFS/ESP series and both have been gradually amended and improved several times over the years. Even AeroflyFS2, which certainly has a more up-to-date engine, is based on AeroflyFS1 (2012) which again is based on the AeroflyRC engine starting back in 1998.

Kind regards, Michael

[Mace 1,902]

On 1/15/2019 at 6:37 AM, AnkH said:

All I can say that I have HT active in my BIOS for my 8700K running at 5.0GHz yet I have the best result in P3D with using an affinity mask of 1365, basically limiting the simulator to 6 cores. Others turn off HT completely and see good results. Means: in any case already now the 9700K would be the better option compared to an 8700K or 8086K (marketing bull* of Intel, the 8086K has NO advantage over an 8700K), as you have 8 "real" cores straight from the beginning.

With our cpu, I wonder what the benefit of 1365 is?  It seems to allocate every other core to P3D, but why is that advantageous?

Incidentally, I tried this setup last night, and it was nice.  I'm not sure it was any nicer than HT=off and no AM, but I'll do further testing as to CPU load etc.

[virtualstuff 751]

On 1/16/2019 at 5:18 PM, Luke said:

You'd be amazed at the leaps and bounds in terms of processing power since 1995. There's a good chance that a high-end iPhone or two can crunch more numbers than those units.

In 1995 I had a 33Mhz 80486DX at around 25 MIPS. Nowadays we can get numbers approaching 1,000,000 on higher-core-count processors.

Cheers!

Well that was a big step up at the time from that 80286 cpu

[Guest]

All I am interested in is the whether there are any leaps and bounds since the i7 4770k!

I am thinking not leaps and bounds but on a few slight improvements. It just might be the case that going form an i7 4770K at 4.8ghz to an i7 9700k at 5.2ghz might not be worth the money and the 5.2ghz is assuming that I win the silicon lottery.

I'll have to do some more testing with VR before I proceed.

[eslader 1,779]

If you already have the 4.8ghz chip, then I don't think it's worth it to go to the 5.2. But if you're upgrading from something old and trying to decide between the two, then if you can swing it get the 5.2 -- not only does p3d *really* like clockspeed, but it's also just a generally good practice to get the best you can get when you build so that you don't have to build them as often.

 

[AnkH 2,233]

14 hours ago, Mace said:

With our cpu, I wonder what the benefit of 1365 is? [...]

Incidentally, I tried this setup last night, and it was nice.  I'm not sure it was any nicer than HT=off and no AM, but I'll do further testing as to CPU load etc.

Basically it is exactly the same as HT=off for a six core. All it does is limiting P3D on all "real" cores of the CPU and not the "HT cores". I use this kind of AM (0101... etc) already since the first days when I was still on FSX with a i7-3770K (there, AM=85 was used) and it always provided me the best results.

The easy reason why I prefer this AM over HT=off is the fact that I do not want to resign from using HT for all other applications and games I use on my rig. And switching HT on and off each time I want to use P3D or something else is waaaaay too cumbersome...