3DMark11 and Hyperthreading = High Physics but Low Graphics Scores
I posted these strange results in an earlier thread from a user which answered a question i myself had raised in this forum. I thought a light needs to be shone here.
Basicly, having a Hyperthread enabled CPU can throw the balance right out with 3DMark11's bench test. Getting very high Physics scores whilst at the same time showing a decline in Graphics performance would show that indeed the i5's do infact perform as good if not better than i7 CPU's in lot of games given that these said games do not take advantage of HT.
NVIDIA GeForce GTX 690 video card benchmark result - Intel Core i7-3930K Processor,ASUSTeK COMPUTER INC. RAMPAGE IV EXTREME score: P23009 3DMarks
NVIDIA GeForce GTX 690 video card benchmark result - Intel Core i7-3930K Processor,ASUSTeK COMPUTER INC. RAMPAGE IV EXTREME score: P20995 3DMarks
So this also raises another question! If Nvidia card support Physics on a hardware level, then why is this still showing a bias towards hyperthreaded enabled processors?
3Dmark 11 runs physics on the cpu, not the gpu by design.
Edit: Also notice that your comparing results run on two different driver versions & different os's
Didn't see that!
Thanks for noticing that...The original post came from a user that was using these 2 posts as with and without HT. I'm wondering now how he or she achieved that (was there 2 PC's used for testing?). I'll have to go find the article again. But yes - it isnt a clean test if the drivers and OS doesn't match.
I may do a test with my laptop which has a HT processor and radeon
Watch this space for "MY" test. Anyone else like to test this theory, feel free to post on this thread.
|All times are GMT -7. The time now is 09:00 PM.|