Xeon E5-2660 V2 $124

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

LukeP

Member
Feb 12, 2017
183
21
18
44
for example: PassMark CPU Benchmarks - Single Thread Performance

exactly as i described. of course it depends on algorithm and memory access. but you wanted an example.

cinebench single threaded will show the same. the IPC improvements are not so great. the multi threading ability is a much greater architectural improvement however.

there are many choices for high ram, high core count, and in most commercial circumstnces thats good. but none for high ram-high single threading. thats really all im saying here.

i would like 512GB ram and 6Ghz cpus.
 
Last edited:

fossxplorer

Active Member
Mar 17, 2016
556
98
28
Oslo, Norway
While reading latest comments regarding performance, i just could resist running Geekbench on a Precision Tower 3420 we have here running Linux: Dell Inc. Precision Tower 3420 - Geekbench Browser
I'm amazed by the the single thread performance and the multithreaded performance in just below a MacPro5,1 with 2 x X5670 :)
Not bad at all for a 4C/8T CPU, although with pretty high freq(3.5/3.9Ghz)!
 

am4593

Active Member
Feb 20, 2017
150
36
28
44
Yeah thoes Mac Pro 3,1-5,1 systems can produce great cinebench and other benchmarking scores but their electrical use and heat generation make me want to cry.
 

BLinux

cat lover server enthusiast
Jul 7, 2016
2,672
1,081
113
artofserver.com
Yeah thoes Mac Pro 3,1-5,1 systems can produce great cinebench and other benchmarking scores but their electrical use and heat generation make me want to cry.
a lot of this perception might be relative so it's hard to say, but I recently compared a westmere dual CPU system to a sandybridge dual CPU system that was very similar and the idle power consumption difference was only about 20-30W. that might be a huge difference to some, but that's where i think it is relative. i'm sure the gap is a bit bigger with IvyBridge, but say you double that to 60W difference... that's 525.6Kw-H per year, or about $105/yr more @ $0.20/Kw-H. Does it justify the cost of buying two v2 CPU? I haven't been able to convince myself of that yet.

I did however, find myself a rare deal on some E5-2660 (v1) for $25 each, and that was about the same or less than any Westmere I was looking at so I went ahead with it. with with E5-2660v2 at around $130-150, it doesn't make sense yet for me.
 

Evan

Well-Known Member
Jan 6, 2016
3,346
598
113
i think the biggest power savings between the platforms comes from the PCH
And seems that's getting worse with the new scalable Xeon's :-/
Not really sure till we see people build some systems buy does not look like (and they are both 14nm so to be expected) that much difference from e5 v4
 

realtomatoes

Active Member
Oct 3, 2016
251
32
28
44
Edit: If you really want to compare then you can use the E5-1680 V3 or the E5-1660 V3 which can not only be overclocked but are good 8c/16t to compare to the 2687W V2. Once overclocked any single threaded clock speed advantage the 2687W V2 had would be easily overcome at for less than what a 2687W V2 costs.
TIL you can overclock a xeon.
i've been using tons of these servers, never thought i'd think about overclocking one. i wonder how hot these things get... time to google more stuff...
 

T_Minus

Build. Break. Fix. Repeat
Feb 15, 2015
7,640
2,057
113
And seems that's getting worse with the new scalable Xeon's :-/
Not really sure till we see people build some systems buy does not look like (and they are both 14nm so to be expected) that much difference from e5 v4
As long as watt / core or /ghz doesn't get worse I don't think anything else is relevant in terms of power consumption for cpu or motherboard that is unless we're talking >20w by motherboard alone or cpu at idle.

Just me? Not sure...
 

realtomatoes

Active Member
Oct 3, 2016
251
32
28
44
ah, E5-16XXs. gotcha.
yeah, not fond of base clock OC even on my gaming rig. definitely wouldn't want that on my server too.
also i got an E5-26XX, so that's a no go.