Is my FX 6300 getting better with age?

[mjslakeridge]

Back in 2013 I ran the Passmark Performance test on my FX 6300 and got very disappointing results. My results were in the range of 3,013 to 3,623 (depending on whether I overclocked or not). At that time the average per Passmark for that CPU was 6,541. (I kept my results documented in a Word Document, which I found today).

Today I downloaded the latest version of Passmark (version 9.0), and decided to run the tests again. My CPU score is now 6,233 ! Same crappy motherboard (Gigabyte GA-78 LMT S2P).

I am at stock frequency now. Here are the changes made to my computer since the first of the year:

Replaced my 2 x 4GB of DDR 3 1333mhz ram with 2 x 4GB orDDR3 1333mhz GSkill Ripjaws Gaming Series RAM.

Installed a Seasonic G series 450 Watt PSU. Old PSU was a Thermal Master 500 watt with 20 Amps on the +12 volt rail (if you can believe the info on the label).

Installed a Nvidea GeForce GT 710 GPU (Don't laugh, I am not a gamer and got it new from MicroCenter for $20 after rebate). Was using the GPU on the motherboard before.

Installed a Zalman aftermarket CPU cooler that I had laying around. Before I was using stock AMD cooler, but heat wasn't an issue, just wanted a quieter cooler.

Went from Windows 8 Professional 32 bit to Windows 8.1 Professional 64 bit.

This new information won't change how I use my computer, but was just curious how my CPU score has doubled at stock frequency.

 

[gasaraki]

No. Probably the scoring changed sometime in between the 2 tests.

 

[WildCard999]

The only noticeable difference I see is that you went from 32bit to 64bit OS which allowed you to use 8gb of RAM instead of 4, not sure if this would of increased your score.

 

[0ldsch00l]

I wouldnt really recommend passmark for any kind of performance bench. Cinebench would be better for CPU. All passmark does is create an avg number of performance on ALL parts. Bench each part individually is better

 

[mjslakeridge]

No. Probably the scoring changed sometime in between the 2 tests.

I don't think Passmark changed the scoring methodology because the average on their site is still around 6,300.

 

[mjslakeridge]

The only noticeable difference I see is that you went from 32bit to 64bit OS which allowed you to use 8gb of RAM instead of 4, not sure if this would of increased your score.

I did go to 64 bit for that very reason (RAM). I also tested this CPU back in 2013 on a Windows XP Home 32 bit OS and got about the same results as the Windows 8 Pro 32 bit OS.

 

[mjslakeridge]

I wouldnt really recommend passmark for any kind of performance bench. Cinebench would be better for CPU. All passmark does is create an avg number of performance on ALL parts. Bench each part individually is better

Passmark does multiple tests on each part (CPU, RAM, GPU, Storage Drives) and reports the results in detail. It does give an overall score, but that is not what I am looking at.

I may give Cinebench a try.

 

[Slumy__57]

I highly doubt that your CPU is getting better over time, PC hardware doesn't tend to age well. If I had to guess, I would say that either something was going on with your system when you tested the first time, (virus or something of the like) or the test just didn't work correctly.

 

[0ldsch00l]

I highly doubt that your CPU is getting better over time, PC hardware doesn't tend to age well. If I had to guess, I would say that either something was going on with your system when you tested the first time, (virus or something of the like) or the test just didn't work correctly.

Actually age helps with OCing some say its placebo others say nay. But electrical transistors when "burned in" years ago everyone was raving how they got better OCs after leaving say 48 hours burn in on CPU stress test and then did the OC

 

[mjslakeridge]

I highly doubt that your CPU is getting better over time, PC hardware doesn't tend to age well. If I had to guess, I would say that either something was going on with your system when you tested the first time, (virus or something of the like) or the test just didn't work correctly.

I just put that as my title. I know performance wouldn't double over time. Just wanted to get some input on the variable(s) that caused the improvement. Don't think it was a virus, because I had also installed a Windows XP OS at the time (on a different HDD), and got similar results as with the Windows 8 32 bit OS.

 

[0ldsch00l]

Sometimes OSes get updates where HW performs better simple as that, its all software. The HW has its performance from factory, software determines its full potential or not

 

[mjslakeridge]

Sometimes OSes get updates where HW performs better simple as that, its all software. The HW has its performance from factory, software determines its full potential or not

That's what I think too. Perhaps the current 64 bit OS I am using. Next time I clean out my computer I may remove one of the RAM sticks and run the test, but I don't think that will have too much of an impact. When I ran the test today, I didn't even close out of several apps I had running, along with 4-5 tabs in Chrome, so I could probably get a better CPU score if I closed everything first.

 

[0ldsch00l]

Sometimes OSes get updates where HW performs better simple as that, its all software. The HW has its performance from factory, software determines its full potential or not

That's what I think too. Perhaps the current 64 bit OS I am using. Next time I clean out my computer I may remove one of the RAM sticks and run the test, but I don't think that will have too much of an impact. When I ran the test today, I didn't even close out of several apps I had running, along with 4-5 tabs in Chrome, so I could probably get a better CPU score if I closed everything first.

Of course benchmarking should be done in optimal conditions

 

[mohitakundi]

Well, sometimes we do get weird bench results.....
My i3 used to always get ~850 single core (CPU-Z) but once got 243. I was shocked ! 2 Days later, The latest version again shows ~850.
Turns out that version of CPU-Z was responsible for the poor result