Now a lotofpeopletendtothinkthatifyoutakeanAMGGPUandyouparentwithan A M D.
C p youthatyou'regonnagetbetterperformance.
That's nottrue.
Thefactofthematteris, aslongasyourGPisnotbeingbottlenecksby a CPU, that's tooslow.
Weatherby, IntelorAndyyourGPisgonnaperformtothepotential, whichitcanperformwhen a lotofpeoplearetypicallyconfusingistheideaofgamedevelopmentandsubroutineswhereeitherAMGengineersareinvideo.
Engineerswilldump a much a bunchofmoneyintothedevelopmentof a gametoeleveragetheirparticulartechnologies, whetheritbesomethingspecifictoAndyorsomethingspecifictoIndia.
AndthenwhathappensisasthoseparticularGPis a recognizedbythegame.
Theyrun a wholedifferentsetofroutines, thenthecompetitors.
Sothat's whyyou'regonnaprobablysee a biggervarianceandperformancebetweenthegametitlethantheCPUitself.
Infact, youmightsee a lotmoreMDlogosongamesmovinginthefuturebecauseofnextgenconsolesandwiththeriseinpopularityoftheriseandseeyou.
Sometimesit's funjusttotake, like a 30 oughtsixandjustshootonename, andthentheycouldstillsometimesgetsomedataoffofit.
NowThisisonethat I havepersonallyhadpeoplesaytomeonthischannel, andthatiswhyyouputtingsuch a bigpowersupplyonthey'reputting a biggerpowersupplyonyourcomputer's gonnadamageitbecausethecomputercan't useallthatpowerorit's pushingmorepowerthanthepartscanuse.
There's a maximumdrawavailablebasedonthepowersupplythatyouusenow, thereasonwhy I usebiggerpowersuppliesandpeopleoftenthinkthat I need.
Forinstance, I'llrun 1000 wattpowersupplyon a singlegraphicscard.
NominalCPUsystemthat's notrunning a tonofharddrivesisbecauseofefficiency.
Kurt, It's ahprettycommonifyoulookattheefficiencycurveof a powersupplythatit's mostefficientbetween 50 and 70% utilizationofitsMaxpoweroutput.
Sotokeepthingsasefficientascoolaspossiblewithmypowersupplies, I tendtooversizethem.
Sometimesit's a wasteofmoney, but I'm nothurtinganything, Sothisnextoneisonethat's onlykindofbeenaroundforthelastcoupleofyearswhenitcomesto I knowthingsbecoming a littlebitmorenormalizedwithPCTechandthatisthatsolidstatedrivesdon't liveaslongasharddrives.
Well, Philactuallysaidthatone's plausible.
SothereasonwhyPhilsaysthatthisone's consideredplausibleisbecause, yes, as I justsaid, thereis a finiteamountofreadrightsavailabletoanSSTbeforeitstartstodegrade.
Thelifespanof a CPUinthemanufacturingprocessofsilicontodayhasbecomesogoodthat a CPUthatwouldlast 10 plusyearsatfactorysettings, whenoverclocktoitsmaximumperformanceandmaxstabilitymightshave 10 15% offofitslifespan, makingitnowineightand 1/2 to a nineyearlongprocessor, makingitlastlongenoughtodefinitelybenolongerrelevanttomoderntech, butnottothepointwhereyou'reprobablygoingtokillit.
Whattendstokillprocessorsistoomuchvoltage.
So I sayit's sortof 1/2 truthbecauseyouhavetobesmartaboutthewayyouoverclockit.
Let's faceit, mostpeopleupgradetheirsystemsthataroundthefiveyearmarkandifyoucouldhavehad a CPUrun 10 years, andnowit's runningeightatmaximumoverHKlockableperformanceYoureallydidn't changeanythinginthelongpictureofthelung, thebigpictureWhenitcomestothings, whatyoumightnoticeovertimeis a slightdegradationofthesilicon, whichmeanswhatisachievabletodaymightnotbeachievablein a yearortwo.
BothPhiland I haveexperiencedpecestowherewe'veoverclockedthemtothepointwhere, after a yearortwo, we'vehadtodropthemmaybe 100 megahertzandaddanother 50 millionvoltsofvoltagetogetstableonceagain.
Infact, I'm onlykilledoneprocessorinthelifespanofthischannelinsevenyears, andthatwasactually a nametheFXprocessor, whichwaskindofnotthemostoverclockingfriendlyintermsoflongtermuse.
Butthat's kindofbesidesthepoint.
Whatguys?
I hopeyou'veenjoyedtoday's videowherewejustkindofhadsomefuntalkingaboutsomePCadmits I amgonnabeheadingtoAmericanliterallythisweekendbythetimeyouguysarewatchingthis.