Is this a legit way of comparing cpus?

[crazymonkeyzero]

I just found this neat chart recently and it apparently has most of the new cpu's ranked by their passmark scores. Is this a good way to compare their relative speeds/performance between chips or am I doing it wrong...lol Thanks for feedback and or comments.



http://www.cpubenchmark.net/high_end_cpus.html

 

[nonameo]

nooooooooooooooo



not another passmark!!!!

I don't care how much they swear to it on slickdeals, passmark is crap. use the anandtech bench and get a nice, detailed feel for yourself.

edit: they put the FX-8100 just below the 2500k. Like, within a negligible distance. And then the A8-5500 is wedged in between the two. WTF?!?!

 

[Puppies04]

It is a good way to compare the speeds/performance of different cpus on passmark. Other than that not really.

 

[palladium]

A good way to compare CPUs is to compare their performance (and other metrics, eg price, power etc) with the apps YOU want to run. If passmark is your main app, sure why not?

 

[Puppies04]

A good way to compare CPUs is to compare their performance (and other metrics, eg price, power etc) with the apps YOU want to run. If passmark is your main app,go stand in the corner and have a quiet word with youself

Fixed that for you

 

[ShintaiDK]

Passmark is a joke.

 

[Borealis7]

Passmark is my favorite game!

 

[nemesismk2]

Passmark is ok as a cpu test but i always check reviews on Anandtech as well.

 

[Magic Carpet]

I just found this neat chart recently and it apparently has most of the new cpu's ranked by their passmark scores. Is this a good way to compare their relative speeds/performance between chips or am I doing it wrong...

If you know how those benchmarks work, then, it can be used as a good source. However, if you come from nowhere and try to use it, you might see a lot of data, mixed up, in a bizarre way.

 

[T_Yamamoto]

Passmark is a joke.

Yes sir it is

 

[TakeNoPrisoners]

There isn't a single benchmark that can tell you exactly how good a CPU is. You have to compare performance in applications that you actually use.

Same thing with GPUs, 3Dmark is useless and should not be used to when shopping for a GPU. The only reason these benchmarks exist is for epeen.

 

[AsusGuy]

There isn't a single benchmark that can tell you exactly how good a CPU is. You have to compare performance in applications that you actually use.

Same thing with GPUs, 3Dmark is useless and should not be used to when shopping for a GPU. The only reason these benchmarks exist is for epeen.

Agreed, looking at the performance of a given CPU or GPU in the applications that you will be using is the best way to judge performance. I think the draw to things like Passmark is that people like to see a large grouping of CPU performance in respect to other similar CPUs. OP just remember to compare application performance for the given CPU and carefully observe testing methods and make sure the results are not BS.

 

[IntelEnthusiast]

I use the benchmarks on Anandtech.com. I think they have just about the best spread of benchs for a number of applications that will give you the best idea on your performance in those applications.

 

[Charles Kozierok]

I use Passmark for casual comparisons. No, it's not as detailed and accurate as those from a proper review site like AT. But it's a good simple first-pass way of comparing CPUs in the aggregate, and it's something ordinary people can understand.

I wouldn't necessarily buy a CPU with a 4,500 rating over one with a 4,400 rating. But when we're considering two options for a laptop for one of our kids and my wife wants to understand whether or not an extra $200 for processor Y over processor X is worth it, a simple summary rating is useful -- she's not going to read or understand a review.

 

[crazymonkeyzero]

Ok, thanks for the clarification guys. The reason I even tried using passmark was because it was the only chart I found that compared many server processors as well, such as the Xeon E5s and the Opterons, which I was inteterestest in. When I checked the Anandtech benches, I found only desktop chip comparisons, but I guess that applies to the 99%, so can't complain. Does anyone know a chart which compares server cpu performances, in applications such as floating-point?

 

[Ratman6161]

Ok, thanks for the clarification guys. The reason I even tried using passmark was because it was the only chart I found that compared many server processors as well, such as the Xeon E5s and the Opterons, which I was inteterestest in. When I checked the Anandtech benches, I found only desktop chip comparisons, but I guess that applies to the 99%, so can't complain. Does anyone know a chart which compares server cpu performances, in applications such as floating-point?

Probably the main reason you are finding mostly comparisons of desktop CPU's is that server stuff is much more difficult to compare. When you are talking about servers it's even more important to see benchmarks specifically testing the software and workloads you are interested in - and generic benchmarks are even more useless than they are with a desktop.

AnandTech has done some decent comparisons over the years. Try going to the main site and searching on "server benchmark"