HumbleChief

Thanks Perry. I should check and purge my CAD blocks some day, good suggestion.

Interesting and I tend to agree with your premise and even though those number don't seem correct as the both faster i7960 and i7970 show 1388 and 1391 respectively (couldn't find the 920 or the 5690) on this chart, I don't think the numbers invalidate your point regarding singe threaded operations. Here's another tidbit - Chief recommends more cores the better - why? Doesn't negate your theory, which again I think has merit, suggest that faster single threaded chips might be best? Or of course as suggested the fastest chip with the most cores, in that order of preference? https://www.cpubenchmark.net/singleThread.html BTW I am now running X5690's and they absolutely rip through RayTraces which should be expected. Too bad I don't RT much anymore.

Sounds great Lance. Curious about what drive you chose. Standard SSD or one of the newer m.2's?

Yes Graham, your analysis is correct and of course a new 7700K would be today's best choice but it still doesn't answer the question posed in the OP. I have an older i7 and the dual Xeons and they BOTH perform the same in certain operations. The CPU core speed and the number of cores has no effect. This is demonstrable and I just find that weird. There's something else going on behind the scenes where CPU speed has no effect.and I still wonder why. Is it simply a disk writing operation that can't be bypassed no matter how fast the processor? I am actually very happy with my system still but am curious about how Chief behaves and why out of simple curiosity.

Red slot - is that good or bad? And if it's good how is it good and what operation(s) would it effect? Do you have a good guess? Or where might we find out how things are effected and by what?

...and why doesn't Chief at least throw us bone? Disk operations affect this; CPU operations affect this: GPU operations affect this. If you want faster 'this' then do 'this'. There is obviously a LOT more going on than just CPU for RayTracing and GPU for 3D Rendering as is the common wisdom here.

Same exact setting(s) here.

BUT then you'd only have a limited number of undo's -correct? Us boneheads need our undo's but I'll try with fewer.

This is so confusing for anyone interested in getting good performance from their computer and it's a complete mystery with absolutely no help, hints or advice from the good people at Chief. I posted a plan that takes 5 seconds for any 3D move of the roof planes with my dual Xeon system. Not the fastest but pretty good. I also have an older i7 920, one of the original i7's and the operation is EXACTLY the same on that older, slower computer. I am foolish enough to think that a faster CPU and GPU processor might speed up the operation but both hardware items seem to have virtually no effect on the 3D redraw speed. Instead there are magical settings one must locate, set, reset, experiment, test, retest and MAYBE those setting will speed the operation but throwing more hardware speed at the problem will not. Sorry for the rant but this just seems a bit crazy. Is it just plain dumb to think that more CPU GPU power will speed Chief up? Why doesn't it? Where's the bottle neck? Is there any way to know without investing thousands only to find that the new system did nothing to boost performance? How do we boost performance Chief? Is there simply a wall that Chief hits and once there - too bad too sad - deal with the slowness? Hopefully some of the recent threads can serve as a notice to those who are looking to speed up their systems. First try and identify why the machine is slowing down - not easy but worth the effort. Second try and determine if a faster machine will help - even harder without investing in the faster hardware. So just beware about the false notion that more CPU power will equate to faster Chief performance. Sometimes it might, other times it definitely will not. https://www.dropbox.com/s/gx11twxukxd5zov/NICK CHEN PROPOSED 18.plan?dl=0

Same here

Did not work out Perry. You were correct, the older tech just really can't keep up no matter the muscle installed. System is still really quick and might be a bit 'snappier' feeling but nothing dramatic.

Installed new Xeons and noticed no real change. Did some bonehead benchmarks and no change in the numbers or perceived speed. OC'd to 3.9 Ghz and it ran too hot (100C during RayTracing). Back to stock settings and am done upgrading this hardware. Hope that helps someone in the future.

I have a suspicion that is true but no hard (drive?) data to back it up.

Yeah, couldn't agree more Perry and have explained my, I think, unique situation in a few threads. It will cost me around $2000 or perhaps closer to $4000 if I buy retail to upgrade to a newer system that will get me just a little bit better performance as my system is still pretty fast. Not smokin' fast any more but adequate. So I have instead chosen to invest $300 to see if this machine will last another 2 - 4 years. Could very well be a fools errand as you are suggesting but I am willing to find out. I'll post any results as objectively as I can.

Xeons just arrived. There's NO WAY I'll be able to wait for the weekend. They'll be in tonight just need to get some good thermal paste.

My new Xeons should be here tomorrow and I'll get them installed this weekend. A system with those processors was once deemed one of the fastest desktops in the world (about 5 - 7 years ago which I know is a LONG time in computer years) but I do not expect them to blow today's fast i7's out of the water, nor do I expect them to provide a definitive answer to the many questions in this thread, nor do I expect my sloppy benchmarks to be accurate, but I do expect to provide perhaps another clue to the behavior of Chief with a system with simply, all else being equal, faster processors. Looking forward to it with expectations on the low side.

Yeah, where do we look Chief? What do we change when our plans start to bog down? If it's not hardware then where/what could it be? And is there simply no solution?

Interesting...

Got it Perry thanks

Got it video 326 if anyone is interested.

Instead of lapping like a dropped beam would?

CPU, CPU, CPU - sad but a reality with today's as well as yesterday's computers.

Here's hoping but as always I am, at best, cautiously optimistic.

There is but they won't. I don't blame them as there's absolutely no ROI to basically re-write huge portions of the code and why? So we don't have to buy faster computers to operate ever more complex software? Not sure what world that is but it's not the one I live in. Besides they already, theoretically, made undoes faster but they are no faster on my machine than they were before. Not being negative, but realistic. Where's the motivation for Chief to invest many, many resources to make their program faster? If there's a quick fix to sloppy code then maybe yes but overall speed boosts? I dunno. Oh wait X9 is faster than X8. Is it really? Is the speed difference perceivable? Anyone done any actual every day testing? Our best hope is that Intel jumps back on the HPDT (high performance desk top) and gets us some rally fast chips for a reasonable cost. Only one man's opinion.

Got this one https://www.amazon.com/gp/product/B01GAI6478/ref=oh_aui_detailpage_o01_s00?ie=UTF8&psc=1