Now, yes, I knowtherearedifferencesinperformancebetweensyntheticbenchmarkandtheactualgameplay, butwe'relookingforherearethedeltasofthedifferencesbetweenthedifferentcanfix.
Butweneeded a realisticbacktobacktesttoallowustoatleastfeeltomeasurethosedifferencesWeretheWarcraftactuallywereinorderMarbecausethemajorcitiesairwherethebiggestslowdownsareintermsofplayermodelsandthingsthattheCPIwashavingtocalculatehastocommunicatewiththeservertowheretheseplayersareanddrawitandallthatsortofstuff.
Ourbudgetbuildsthat I'vedoneinthepast, andthereasonwhywe'redoingthisvideoisbecause I alwaysrecommend, ifyou'reon a budget, gowithtwosticksoffourgigabytes.
I saidthatineverysinglereasonthatwe'veeverpurchasedmemoryforourbudgetbuilds, andeverysingletimeplentyofpeoplewillstepupandsay, J thatstupidgeteightgigabytessinglestick.
Sogoingforhighframerateandthenmakingthememorychangeswillsortofexaggerate a littlebitsomeofthememorystuffthatwe'retryingtodemonstratehere, butrealistically, a 28 e t.
I for a 10.
80 p isnotgonnabe a verysmartmove.
Wefindthatobviously, ifyou'rerunning a tendency, I doatleast 14.
40 p orhigherbutmovingonourbaselinetestfor 16 gigs.
Dualchannelagain.
All 26.
66 memory 134 F.
P.
S.
Soit's kindoffunnythatitwassoclosetotheWorldofWarcraft, but 16 gigabytessinglechanneldroppedallthewaydownto 110 f.
P s.
Sothatis a 24 f.
P s loss.
That's a prettybigpercentagelossbyjustliterallyhavingonestickmovedintotheRamthewrongRamslot.
Soeightgigabytesdualwentfrom 1 34 to 1 33 Soonceagainthecapacityof 16 gigabytes, which a lotofpeoplearesaying 16 gigsorgohomewiththesebuilds, youcankindofseeit's notholdingwaterhereatleastwiththistitle, becauseone f psdifferenceandthat's stillmarginoferror.
Weranallthesetestsfourtimesandthenaverageofeightgigabytedueling 1 33 eightgigabytes, Single 108 So 108 foreightgigs, singleversus 1 10 for 16 gigsingleto F PSdifference.
Buttwicethecapacity.
Yeah, youcanseerighttherethatobviously I wouldfavor a lowercapacitydualchannelsetupforthistitlethananyofthehighercapacity, singlechannelstuff.
I mean, I don't think I wasgonnarun a 16 gigsinglestick.
Infact, I wasrunning a bookthatbeattheBIOS.
Butanyway, I digressforgigabytesinglegaveus 101 F.
Butstill, it's still a beautifulgamethatholdsupfouryearsafterthefact.
Sonowthequestionattheendofthedayis J, whichcanthink, Should I berunningthen?
Well, I thinktheobviousansweristhebiggestcapacityyoucangetinatleastdualchannelNow, withthreadRipperand X 2 99 onIntel, you'vegotQuadChannelthere, whichmeansyouwantatleastonesticktooccupyallfourchannels.
So a futurepeacewe'regoingtodoisisanythingmorethandualchannelevennecessaryforgaming?
I thinkthespoileralertthereisprobablynot.
Butyou'd neverknow.
We'vedonetestbefore.
We'vebeencompletelywrongandsurprisedbythat.
Nowit's gonnacomedownentirelytothetitlethatyouareplaying, andyoucanusuallytellhowdemandingthattitleisonCPUandmemorybytheamountofmemorythatitsortofhas a mincepeckversusrecommended.
I thinktheresultsherearestillsurprisingtomeonlybecause I thinktherewas a biggergaptherethan I wasexpecting.
I knewtherewasgonnabe a difference.
I justdidn't knowbyhowmuch.
Sowecameintoworktoday.
I said, I wanttodosomeramtesting.
Andthenduringthistest, westartedplayingaroundwithRAMspeeds, andwenoticedthattherewasquite a bitof a differencein, likeWorldofWarcraftandevenfarcryfivewhenwestartedslowingdownthememory.
Sowhatwe'regonnadoinournextpartofthisparticularramtestingiswe'regoingtotakeonourIntelplatformhere, andmaybe I amtheoneaswell.
Andsoifyouguyshaveanytitlesthatyouknowofthatareveryramdependentandram, speedandcapacitycanmake a majormeasurabledifferenceor a tangibledifference, thenpleasecommentdownbelow.
OneOtherpoint I think I wanttomake, tooisifyou'renotrunning a 28 e t.
I youprobablywould'vesawevenlessof a ahdifferencehere.
Butyoustillwouldhaveseenthedifferencebecausethey'restillbecomes a pointatwhichyou'reexceedingthespeedoftheCPUcanactuallyhandletheamountofdata, theGPSsendingitaswellasthememoriesabletolettheCPUaccessitwherethatpointis, I don't know.