BenchMark Score low?

[Frankie]

I have just purchased a Gigabyte Geforce G1 Gaming graphics card and have done a benchMark test using FurMark. I have the resolution on FurMark set to 1024x576 and I'm getting a BenchMark of 3602. I then do the same test with my 970 overclocked  more and then get a BenchMark that's actually lower 3601? Can someone tell me why I'm not getting a benchmark over 4000 plus like many other people get using this card? Thanks!

Motherboard: Gigabyte X58a Ud3r
Cpu i7 930 overclocked 4.0 Ghz
Ram 12 GB
Psu Collermaster 700 Watt
Graphics card Gigabyte G1 Gaming 970.

[DaveLembke]

Curious as to if your games run better at this overclock or if there is any noticeable flaws while running overclock. As for an overclock does not always guarantee faster/better performance. Reason why i state this is because I once overclocked my CPU and GPU to the max and my frame rate in the games instead of holding steady was now jumping all over the place due to lag and surging. it wasnt until i backed off on the overclocks which were 12% initially down to about 10% that I found the sweet spot to where I had a performance gain and no problems. I also ran the benchmark at baseline wrote down this value, after 12% overclock and wrote down this value, and saw oddities in the results measured, but when gaming the flaws in these overclocks stood out like a sore thumb vs staring at Windows desktop. When I backed down to 10% I had consistent overclock scores showing the gain and games ran better and most importantly the system was running ok and not crashing.

I was using passmark 30-day trial edition for my benchmarks.

[patio]

Many many other factors can affect benchmarks...RAM speed, FSB, HDD read and write speeds, CPU, etc. etc...

[Frankie]

I just ran a system test and the score shows the HD having a score of 5.9. The rest of my scores are 7.9 so would this indicate that it's the hard drive? Would a SSD increase my score by alot more?

[patio]

Probably...

[Calum]

It certainly would help your system feel a lot more responsive but not sure it'll make any difference in Furmark as Furmark is just a GPU stress test.
I would perhaps give one of the Unigine benchmarks a try, Heaven or Valley, and compare using that as Furmark isn't a realistic benchmark or load test, it's mostly only good for testing maximum stress and heat generation.  That said, it's also possible as above that your overclock just isn't stable especially if you clocked the memory, modern cards use error correction meaning you might not see artifacts or crashes, just slower performance as the error correction kicks in.