Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


Xeon E5-1620v2 vs W3520
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

Xeon E5-1620v2 vs W3520

Which server is better for VoIP? (VoIP Calling eats whole cpu so refer me the best one)

Link: https://www.soyoustart.com/us/offers/173sys10.xml (E5)
Link: https://www.soyoustart.com/us/offers/142sys4.xml (W3)

Comments

  • What sort of VoIP calling do you have eating a whole CPU? That sounds very wrong, and probably should be fixed before comparing any CPUs. For reference, X5690s can handle 100+ simultaneous ulaw channels without using much CPU at all, so you must be causing some serious load or your software is misconfigured or bad

  • the E5 it's single thread is nearly double that of the W35

  • @scaveney said:
    What sort of VoIP calling do you have eating a whole CPU? That sounds very wrong, and probably should be fixed before comparing any CPUs. For reference, X5690s can handle 100+ simultaneous ulaw channels without using much CPU at all, so you must be causing some serious load or your software is misconfigured or bad

    I have 1000+ simultaneous g729 and uLaw/aLaw channels.

    Thanked by 1scaveney
  • RayhanRayhan Member
    edited August 2017

    I was used their E5-1620 v2 some days for testing something.

    The performance was really great. (Proxmox+ZFS caching)

  • AthylAthyl Member
    edited August 2017

    @SPUZE said:
    I was used their E5-1620 v2 some days for testing something.

    The performance was really great. (Proxmox+ZFS caching)

    E5:

    [root@xXxXxXxXx~]# wget freevps.us/downloads/bench.sh -O - -o /dev/null|bash

    Benchmark started on Tue Aug 15 02:28:28 CEST 2017

    Full benchmark log: /root/bench.log

    System Info

    Processor : Intel(R) Xeon(R) CPU E5-1620 v2 @ 3.70GHz

    CPU Cores : 8

    Frequency : 1200.042 MHz

    Memory : 16064 MB

    Swap : 1021 MB

    Uptime : 3 days, 12:43,

    OS : CentOS release 6.9 (Final)

    Arch : x86_64 (64 Bit)

    Speedtest (IPv4 only)

    Your public IPv4 is xXxXxXxXx
    Location Provider Speed
    CDN Cachefly 99.9MB/s

    Atlanta, GA, US Coloat 14.7MB/s

    Dallas, TX, US Softlayer 12.8MB/s

    Seattle, WA, US Softlayer 3.08MB/s

    San Jose, CA, US Softlayer 9.94MB/s

    Washington, DC, US Softlayer 12.6MB/s

    Tokyo, Japan Linode 9.43MB/s

    Singapore Softlayer 4.56MB/s

    Rotterdam, Netherlands id3.net 49.2MB/s

    Haarlem, Netherlands Leaseweb 105MB/s

    Disk Speed

    I/O (1st run) : 86.5 MB/s

    I/O (2nd run) : 89.1 MB/s

    I/O (3rd run) : 94.7 MB/s

    Average I/O : 90.1 MB/s

    W3:

    [root@xXxXxXxXx~]# wget freevps.us/downloads/bench.sh -O - -o /dev/null|bash

    Benchmark started on Tue Aug 15 02:28:39 CEST 2017

    Full benchmark log: /root/bench.log

    System Info

    Processor : Intel(R) Xeon(R) CPU W3520 @ 2.67GHz

    CPU Cores : 8

    Frequency : 1600.000 MHz

    Memory : 16089 MB

    Swap : 1021 MB

    Uptime : 3 days, 12:42,

    OS : CentOS release 6.9 (Final)

    Arch : x86_64 (64 Bit)

    Speedtest (IPv4 only)

    Your public IPv4 is xXxXxXxXx

    Location Provider Speed

    CDN Cachefly 68.3MB/s

    Atlanta, GA, US Coloat 6.27MB/s

    Dallas, TX, US Softlayer 11.2MB/s

    Seattle, WA, US Softlayer 10.8MB/s

    San Jose, CA, US Softlayer 11.0MB/s

    Washington, DC, US Softlayer 558KB/s

    Tokyo, Japan Linode 8.96MB/s

    Singapore Softlayer 4.83MB/s

    Rotterdam, Netherlands id3.net 49.0MB/s

    Haarlem, Netherlands Leaseweb 35.7MB/s


    Disk Speed

    I/O (1st run) : 121 MB/s

    I/O (2nd run) : 123 MB/s

    I/O (3rd run) : 125 MB/s

    Average I/O : 123 MB/s

    You have new mail in /var/spool/mail/root

    Which one is better by this benchmark logs?

  • If you want more cpu power then go with E5.

  • Which one is better by this benchmark logs?

    You can get much more better performance on E5.

    Sometimes it's dependable on it's setup. (As i think)

    I have show you a Proxmox Linux Container benchmark in the screenshot.

  • basically this E5 is basically a E3 cpu. higher single thread and good overall performance with E5 benefits. (more ram etc)

  • Processor : Intel(R) Xeon(R) CPU E5-1620 v2 @ 3.70GHz CPU Cores : 8 Frequency : 1200.042 MHz Memory : 16064 MB Swap : 1021 MB

    Processor : Intel(R) Xeon(R) CPU W3520 @ 2.67GHz CPU Cores : 8 Frequency : 1600.000 MHz Memory : 16089 MB Swap : 1021 MB

    Frequency made me confused as E5 has low frequency as compare to W3, Thanks for the suggestion i am considering E5 for my usage.

  • MasonRMasonR Community Contributor
    edited August 2017

    @Athyl said: Frequency made me confused as E5 has low frequency as compare to W3, Thanks for the suggestion i am considering E5 for my usage.

    Don't let the frequency output confuse you, that's just the CPU governor downclocking the CPU to save power while idling. So that's not really the highest clock possible. Under high load, the frequency will reach the highest possible for that CPU (so 3.7GHz/3.9GHz burst in this case for the E5). The E5 if much higher powered frequency-wise than the W3.

  • Athyl said: Frequency made me confused as E5 has low frequency as compare to W3, Thanks for the suggestion i am considering E5 for my usage.

    This is only base as no load. The turbo of a 1620v2 is 3.9Ghz @ 130W.

    W3520 - a first generation workstation CPU in socket 1366 with no AES-NI or other advanced functions - is beaten by a simple X5650 which is also cheaper used. 2.93Ghz turbo at the same 130W.

    Performance wise the E5 beats out the W by 30-50% in any workload, with AES-NI tenfold or more. At same power usage peak.

    X5650 (3.06Ghz Turbo with 6 cores, instead of 4 as on these 2) uses less power (95W) for around 75% of the E5-1620v2.

    This gives an overview of functionality:

    http://ark.intel.com/compare/47922,75122,75056,75789,75779

    TarZZ92 said: basically this E5 is basically a E3 cpu. higher single thread and good overall performance with E5 benefits. (more ram etc)

    Sort... of. E3 is based on the i7 base plus ECC controller and - usually, except x5 SKUs - without GPU. Same PCIe amount, same core count (until Coffee Lake with 6), same memory limits.

    E5-16xx is based on a cut down (binned) E5-26xx core with 40 PCIe lanes (vs. 16 on E3/i7) and up to 6 cores, QPI is disabled but you get all RAM channels (thus double speed than E3 as well). Basically disable QPI, clock it up, disable some cores and adapt the thermal limits/TDP, much as an i5-7xxx is nearly 1:1 the same as an i7-7xxx.

    Thanked by 2Iam vimalware
Sign In or Register to comment.