You will need a PSU with an 8pin CPU power connector (or 4+4, which equals 8 anyway). Are you sure an Intel 10core is the correct choice for you? It may well be but hopefully you will make an educated choice. Wait for the reviews!

dude i'm running a gamer, ofcourse i have 4+4 connector with be quiets psu installed ;)

though in all earnest i might switch the 2600 for the 3700x and see if that is that little extra i was looking for

Calm down! You might have had a model like mine that does not have more than a 4pin CPU power cable.

A possibility is also waiting for September and new AMD CPUs.

If you could try a program (game or other) that is problematic for you on a computer with something like the 3700X that might be good but with the current world situation, I guess that is not an option right now. A ten minutes way to see if that would solve your problem.

Infortunatelly, this is not true and have not been for a while now. A good exemple is the i9-9900k that have a TDP of 95W but on unlocked state (not overclocked), it pulls ~175W on full load.

I could rant a lot about this but will keep it simple. Manufacturers usually find ways to give "better than competitors" paper specs... TDP has been just one in a very long list.
TDP is not a standardized value, is very convoluted to calculate and is meaningless. For exemple, Intel calculates it's TDP based on "base frequency", wich none of the current CPU's follow. The very popular i5 9400f has a base speed of only 2.9GHz but it "boosts" all core to 3.8/3.9 depending on the load.

Edit: here is a nice link about the subject.

Same bollocks as the node names that are quite a bit bigger in reality than the 16/14/12/10/7/5 nm they advertise.

It only does so if your system is not properly setup. Either some setting on motherboard, a not optimized Overclock or the cooler is not correctly mounted. That CPU should be ok even on the stock cooler.

Where are you seing those 55 degrees?

That drop from 3.9 to 3.8 is normal and is nor considered "Throttling", is just the way the boost algorithm work.

ah oke, so we had a misunderstanding, at least i used the wrong terminology....

I thought trottling would be described as the auto drop in performance to preserve the cpu's/gpu's health and with AMD's precision boost doin just that... ya know ;)

It is kind of true, they are protecting the best cores so when they get a bit too hot they move to other cores that probably are not quite as fast. At least this is what I think but not 100 on this one. Not all cores are created equal.

Argh and now i still don't know what to do...

stick with ryzen try-out maybe the 3700x or the 3800x to see how they perform to get a general good look on AMD or pick intel again and feel the luxury i always had, not to mention distancing myself from some parts of the amd user community i'm somehow coupled with...

maybe try the 9900k

And going back on the powerdraw discussion

i belief this article sums it up pretty well

during gaming the cpu ( 9900k ) is more consumptive then others but....... we only talk about a marginal difference

Literal quote for the ones to lazy to run through the article

" during gaming we never passed the 100Watts

https://www.tomshardware.com/reviews/intel-core-i9-9900k-9th-gen-cpu,5847-11.html

I wish when those testers tested processors with games that they would test computing heavy games like X4:Foundations. Turn based games can also draw a lot when they calculate the turns.

lol not easily satisfied are you now?

the heaviest game, cpu wise, on my system is NFSU payback go figure

oh wait no i am lying...

it is that stardock debacle called ashes of singularity

NFSU does 50% loads straight and ashes manages to go between 60 and 80% load of my ryzen 5 2600 cpu

Following post 689 , this is probably not the place since it's not gaming related but may trigger someone interest...

After playing a little with the Ryzen system, found something interesting, maybe because I wasn't aware and never seen anything about it on the internet. The power limit set by the motherboard does include the RAM power as well. We all know Ryzen love more and faster memory but not in all case scenario.

What this means? On a freshly installed Windows, with nothing more than 1 stick of RAM, a gt710 and a sata SSD attached, the idle power consumption is ~35W with the RAM set to 2133MT/s. Setting the RAM to 3000MT/s, the idle power jumps to ~43W.
This means the jump in frequency needed~ 8W, wich actually is quite big in % and is one of the reasons OEM only fit 1 stick of slower memory on many laptops.

For reference, the stock R5 1600AF (motherboard defaults to 2133RAM) does 1200 cinebench r15 points @3.6 GHZ and 119W draw from the wall. On CpuZ benchmark hits 3350 @3.6.
With 4.1GHz overclock and 3000RAM, the cinebench score is 1378 with a power draw of 177W. (CpuZ 3800 points).

Now, with the motherboard power limit (PPT and 0.1v undervolting) set to 30W things get interesting.
Setting the memory to 3000MT/s the cinebench power draw is a measly 65W while scoring 766 @2.25GHz and CpuZ did 2450 @2.65GHz..
Since the memory also is included on the motherboard power calculation, setting it at 2133 allow the CPU to increase the frequency to 2.8GHz on Cinebench (963 points) and 3.05GHz on Cpuz (2450 points) at the same 65W power draw

The single core scores and speed are unafected since is still inside the 30W power envelope set by the motherboard.
I'm yet to found a way to increase the single core with overclock, at the same time limit max power on the BIOS. Should be possible, though.
Wasn't able to test 2 sticks of RAM since I borrow some of my parts to a friend, at the moment.

Someone with a Ryzen laptop may eventually be able to dig into and find some more performance on the CPU...