Jump to content
  • entry
    1
  • comments
    3
  • views
    660

Xeon E5 Family which is Faster?

enanobo

573 views

in the Xeon E5 V3 Family there are a lot of diferent processors and if i do the staight matts iv got conffuse:
For Ex. the E5-2690 V3 12 Core - 2-6GHz (12 x 2.6 = 31.2)
and the E5-2687W V3 10 Core - 3.1GHz (10 x 3.1 = 31)

 

in Straight speed, this two has very similar Result: (31.2 vs 31)

 

but their have a $500 Dollar difference

 

can any body tell me if this Straight matts apply to the speed in the sistem performance and in particular for Rendering?

3 Comments

You mean 'maths'? other than core speed which is all MegaHertz & GigaHertz it doesn't totally because you're not factoring how many cpu cores are running, temperatures in a realistic scenario, turbo-boost (this can significantly increase performance in a single core work load & multiple core workload.), socket revision/micro architecture etc.

 

Regarding the processors in this comparison, Xeon® Processor E5-2690 v3 | Xeon® Processor E5-2687W v3 here are the links to the official Intel Ark of both E5 family Xeon's

 

Single threaded workload : Both are more or less definitively even. Both have a maximum core speed of 3.5Ghz with that said not all cores will be running at this speed due to thermal limits, voltage & cooling.

 

Multi-threaded workload : My opinion between these two xeon processors is hard to put a complete winner since I cannot personally test either in person but I would state that they're within a margin of error since both are similar in-terms of high core count with decently high...(ish) clock speed (also factoring in turbo boost). Having said that I would believe the 2687 v3 would score a little bit lower in a synthetic/realistic benchmark such as Cinebench (widely regarded to show performance fairly accurately using a very large community database)

 

10 Cores 20 Threads < 2687 v3 | 3.1 Ghz - 3,5 Ghz

12 Cores 24 Threads < 2690 v3 | 2.6 Ghz - 3.5 Ghz

(Both cpu's are supported on the LGA 2011-3 socket up to dual socket configurations in a workstation/server scenario)

 

Now looking at these two factors one would believe that the 2687 v3 would have the edge due to having a closer in-terms of clock speed threshold a high & more stable speed running 10c 20threads however it would fall slightly short of the 12c 24 thread beast of a cpu which would more or less beat it due to more cores/thread performing on par if not slightly lower.

 

You should @enanobo ask this question in the CPU category in the forum, not the Blogging section since not many users tend to be given a tonne of answers & user feedback.

I'm not 100% right but I believe in different scenarios each cpu has a place regarding value, performance per dollar & the use-case scenario

 

Here are a couple great & intuitive videos to watch created by Linus Media Group & other sources like Tom Logan from OC3D TV. If you have time watch them all in your own time. It's all quite knowledgeable & intriguing reviews & first looks. Watch it in whatever order you'd like but I feel if you watch it in the way I've ordered them it's a more understandable.

https://www.youtube.com/watch?v=khDsbxa5_G0

 

Regarding if All megahertz are the  same

 

This video is fairly old but still funny to watch

Link to comment
Link to post

Fewer faster cores are better for gaming. More slower cores are better for stuff like video editing, mass data compiling and server stuff. But with a 500 dollar difference get whichever is cheaper, as you won't see a huge difference.

Link to comment
Link to post
×