Takze to nakonec zbylo na me 
Udelal jsem XPL built in testy a vysledky fps viz nize
Test 1 Test 2 Test 3 Test 4 Test 5 Test 54
i3-8350K 4.5GHz, TO on 81.75 73.04 54.71 32.51 23.05 31.1
i3-8350K 4.5GHz, TO off 84.04 71.92 55.03 30.72 21.94 29.47
i7-9700K 4.6GHz, TO on 103.71 93.87 74.32 42.41 31.73 39.62
i7-9700K 4.6GHz, TO off 92.25 82.32 64.52 38.06 29.06 35.66
i7-9700K 4.8GHz, TO on 110.12 98.67 77.81 43.98 32.82 41.73
i7-8086K 4.8GHz, TO on, TH off 108.39 97.78 75.38 43.24 32.38 40.67
i7-8086K 4.8GHz, TO off, TH off 92.84 83.78 66.11 39.31 29.84 37.22
i7-8086K 4.9GHz, TO on, TH off 109.2 96.61 77.58 43.78 32.9 41.27
i7-8086K 4.9GHz, TO off, TH off 95.07 86.91 66.59 39.88 30.9 37.43
Vse testovana na XPL 11.30b5
Nejdulezitejsi zjisteni:
- u vykonych CPU s vice jadry Vam pomuze zapnuti OpenGL Threaded Optimizations (v tabulce TO, samozrejme jen pokud mate Nvidia GPU).
- Hyperthreading jen topi, nic neprinese
- vuci puvodnimu 4core procesoru doslo k 20%!!! narustu fps, a to i pri temer shodne frekvenci (delal jsem i testy i na uplne stejne frekvenci a rozdil byl cca 17-18%)
- vybral jsem nakonec 9700K, bezi solidne na 4.8 v napetim 1.265 v XPL topi okolo 50-60C, v syntetickych testech okolo max 65C (torture test nedelam, nedavaji mi smysl)
- Doufam, ze Vulkan bude umet jeste lepe skalovat vykon / jadra, zatim se CPU flaka. |
|
2x 
