Új hozzászólás Aktív témák

• #3056 schawo titán schawo #2906

  Új Válasz 2017-03-22 15:41:15 #3056

  Új hozzászólás

  Összes hozzászólása itt Válaszok az összes hozzászólására itt Válaszok erre a hozzászólásra

  Privát üzenet küldése

  

  schawo

  titán

  válasz schawo #2906 üzenetére

  Végül kiderült. A ScanPro által a teszthez használt 6800k hibás darab. Szóval nem a cache különbség az oka, hanem valami miatt folyamatosan throttlingolt. Kint vannak az új DAWBench Vi mérési eredmények. (Mert közben már a ScanPro is bekapcsolódott a hibakeresésbe.)

  Think I've got it and well spotted. Thanks for feeding it back.

  I've grabbed the benches again this morning and re-run the DAWBench Vi tests on the original kit and the was no change in the results seen.

  So I changed the 6800 for another chip from a new batch and it jumped in performance by about 10% and the CPU load percentages scale more evenly now. I've also re-run the Reaper standard Dawbench and it reflects the same pattern, which it does

  So, it looks like I had a duff i7 that was throttling performance under tight loads.

  As I've re-tested a fresh 6900K as fine yesterday, I wanted to make sure the 1700X got fair treatment too. I've pulled a fresh one of those today from a release batch (straight off off the shelf) and the results for the AMD remain unchanged as they did for the 6900K.

  So what does this mean overall? Not a whole lot, the 6800K makes some gains that obviously improves it's bang per buck value, but otherwise doesn't change the results charts running order to any real degree.

  I've edited the CPU load results, DPC chart and DB Vi chart to reflect the retesting. It isn't a big enough jump to change the placing of each chip in the running order however, but it does correct the scaling of the results that you pointed out.

Új hozzászólás Aktív témák