Low Geekbench score.

[fenderbass146]

So I have a Mac Pro 1,1 I just obtained. Put maybe $50 worth of work into it in hopes of making it my home desktop/server to compliment my MBPr.

So essentially it was 100% stock when I got it and I was originally getting somewhere in the ~5000 on Geekbench. I just did the upgrade to 2xQuad Xeons and according to the other people who have done this upgrade it nearly doubles there scores into the ~9000 range but Im only getting something in the ~6000.

What is going on? I am using the free Geekbench if that makes a different. I also have 20 GB of ram in the MacPro and I am running El Cap.

 

[h9826790]

You didn't mention the critical info. e.g. CPU name, GB version, etc

 

[fenderbass146]

You didn't mention the critical info. e.g. CPU name, GB version, etc

Xeon x5355 quad. I had the stock 2.66 dual core originally. I'm using the current free 64-bit geek bench. Not sure of the version.

 

[jerwin]

Is THIS your Mac?

Mac Pro1,1 X5355
1495 single Core. 5588 Multicore

Looks like they renormalized the benchmarks from Geekbench 3 to Geekbench 4

 

[fenderbass146]

Is THIS your Mac?

Mac Pro1,1 X5355
1495 single Core. 5588 Multicore

Looks like they renormalized the benchmarks from Geekbench 3 to Geekbench 4

Yea that's it. Is it only recognizing one chip? The multicore should roughly the single core x8 right. Like I said the the people on YouTube that did this same upgrade were getting 9000+. If it's just something wth the way they give the scores that has changed im not worried about it but I was concerned something isn't working correctly.

 

[jerwin]

I don't have time to do proper benchmarks-- shutting down everything else, etc, etc, but it's my recollection that the multicore tests don't scale perfectly. Look through the individual tests, and calculate the scaling factor for each one (using throughput, not score). If you always get a scaling factor of less than 4, something is probably wrong.

 

[fenderbass146]

I don't have time to do proper benchmarks-- shutting down everything else, etc, etc, but it's my recollection that the multicore tests don't scale perfectly. Look through the individual tests, and calculate the scaling factor for each one (using throughput, not score). If you always get a scaling factor of less than 4, something is probably wrong.

What explains the higher scores from others who have done this upgrade. Different geek bench version?

 

[h9826790]

What explains the higher scores from others who have done this upgrade. Different geek bench version?

You better try GB3, I doubt how many 1,1's youtube video's result are based on GB4. And you can't compare different version's GB result.

 

[fenderbass146]

You better try GB3, I doubt how many 1,1's youtube video's result are based on GB4. And you can't compare different version's GB result.

Good to know. I'll see what happens.

 

[fenderbass146]

You better try GB3, I doubt how many 1,1's youtube video's result are based on GB4. And you can't compare different version's GB result.

You were right. I ran GeekBench 3 and it was 9000+