TycStyle

Thanks for that. I considered it of course, but until today did not act due to the YouTube vid by Optimum Tech on Rtx 3080 + Overclocked Intel cpu (with multiple lower watt psu's) was no issue for my specific one, but how to be sure....

Hi all, This week I swapped my 3700x for a 5800x as well as a recent upgrade from a 5700xt to the Evga RTX 3080 Xc3 Ultra. All of this is stuffed in the Lian Li TU150 (I don't go to LAN parties, just found it was a cool case without all pre-built stuff inside) and installed on the Gigabyte I Aorus X570 Pro Wifi. Combined with CL16 G.Skill Trident Z Neo (2x8) 3600, powered by the Corsair Sf600 80+ platinum. The cpu cooler is the Scythe Mugen 5 PCGH edition, which receives it's push from a Kaze flex 1200 rpm fan (for max. Silent cooling, replacing the 800rpm ones out of the box). The pull/exhaust and the case intake, are the Noctua NF-a12x25. M.2 nvme pcie 4.0 1TB Gigabyte Aorus + Samsung 970 Evo Plus. No OC on GPU or CPU only XMP profile activated. So here's the deal; 3DMark Timeslot (not the holy grail, but at least gives something proper to analyze), gives a result of just over 16000, which is a bit under average, but doesn't feel that bad. But the temp of the gpu goes above 80 (84/85), cpu stays nicely cool around 60. Because the latest chipsets were missing the Ryzen powerplants, plus in my old setup I tweaked the fans so the curve was much smoother, I made some similar adjustments. But long story short; when I turn up the case fans (f.e. simply to performance or Super speed) , the performance of the benchmarks crashes into the mud, like into 13000 range. So at this point, the BEST result is achieved by the quiet setting on the fans. Btw, the front intake and exhaust are split to the fan header, the intake fan is on its own on the CPU fan header. Could this be caused by some kind of turbulence? Or do you guys have ANY idea what is causing the performance hit? The new combo is a bit hotter obviously, and the sff cases suffer from flow, but does this sound logical? Or any other pointers? This is really tweaking that makes you go mad, just want some nice balanced performance/thermals/noise setup but this doesn't make any sense to me... Much appreciated,

Check thanks! Will have a peak at these tips. It's 1 dimm and looking for a second one for testing purposes. I was checking usage during running the game and cpu seems to stay within 90%, but still at a certain point I see stream quality just dropping into pixel wonderland, which seems to indicate it can't handle/achieve full data transfer? Do you know what the system is doing when running those speedtests making it jump to 100%? Because that behavior is occuring 100% of the time?

hi all, I setup a HTPC a couple of years ago, which I hadn't used in a while. So I thought to give it a go with Geforce Now, as it should be able to handle requirements. So, I installed Windows 10 without any issues, upgraded mobo bios and drivers to latest available.GeForce Now installed without issues again incl. rainbow six siege. During the game, at certain moments, the connection speed would drop significantly, so I went ahead and performed a speedtest on the regular sites (speedtest.net, fast.com , dslreport, big file download etc.)On there, the machine would not achieve more than a absolute max. of 100-140mbps (wired with cat. 5e) out of the max 250, which is achieved with other machines.After a lot of testing, ruling out etc. I noticed the CPU would max at 100% when starting these tests, but only when running the download, not the upload. I have also performed a dos based speedtest, which actually reaches max performance (250mbps), with NO impact on the CPU.Now, what I'm struggling with is to understand why the CPU usage spikes when executing that webbased speedtests and what the further implications might be when running GeForce Now.The CPU usage totals in Task Manager also do not match the sum of the parts, which is very frustrating and stange (yes the list is sorted by usage). So when writing this, the CPU usage keeps jumping up and down (actually when typing...). And in "idle" it sits between 60-70% without any heavy stuff running. But after a (long) while goes into "normal" mode below 10%, except for the increase when typing(???!!). But when starting the speedtest, it's peaktime all over again. Any ideas on this behavior? The CPU performed very poorly in benchmarks (1552!? vs. 2317 avg.), but does not fail testing and the whole system in general boots like it should. But something strange is messing with the CPU for sure... Could it be related to the fact it's an APU? All help is welcome. Thanks in advance. TycStyle Setup: Board: Gigabyte GA-F2A88XM-D3H CPU: AMD A serie A6-6400K - 3.9 GHz - 2 cores - 2 threads - 1 MB cache - Socket FM2 - Box Cooler: Scythe Shuriken Rev.B SCSK-1100 Mem: Crucial - 4 GB - DIMM 240-pins - DDR3 - 1600 MHz / PC3-12800 - CL11 - 1.5 V HD: Samsung 840 EVO Basic - Solid state drive - 120 GB PSU: Be quiet! System Power 7 300W