Howdy, Stranger!

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


Highest UnixBench score I ever have seen - Vultr Multi-core 24cpu - 9600 points.
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.

Highest UnixBench score I ever have seen - Vultr Multi-core 24cpu - 9600 points.

myhkenmyhken Member
edited December 2014 in General

Has waited and waited to test Vultr.com biggest multi-core server, with 24vcore + 64 GB RAM.

Here is the result:

Never seen so high benchmark result on any type of server I have tested the last three years.
Over 2000 points more then their16vcore server.

My aff link is: http://www.vultr.com/ if you want to support my site, and start using their great service.

Wil update my Top20 list later on today with the new 1.place.

Good work @DaveA

Thanked by 2MarkTurner DaveA

Comments

  • Ok.

    Thanked by 3Brad Mark_R Cakey
  • Processor type?

  • So more cores = higher performance score?

    Thanked by 1sc754
  • @myhken said:
    Never seen so high benchmark result on any type of server I have tested the last three years.

    Do you test $600 servers so often?

  • @kerouac said:
    Do you test $600 servers so often?

    Hourly is the difference here :).

  • @GoodHosting said:
    Hourly is the difference here :).

    I doubt he tests so many servers hourly or monthly priced. Besides they probably gave it to him for free to test.

  • @kerouac said:
    I doubt he tests so many servers hourly or monthly priced. Besides they probably gave it to him for free to test.

    See related thread:

    http://lowendtalk.com/discussion/35159/7-month-review-of-vultr-com/p1

  • Okay.

    Thanked by 1Brad
  • Hmm..... Good!

  • Which version of UnixBench?

  • kerouac said: I doubt he tests so many servers hourly or monthly priced. Besides they probably gave it to him for free to test.

    Maybe he show these score to show off, and promote his affiliate link :D "There's no one provider can beat this vultr score, sign up now using these link"

    Thanked by 1linuxthefish
  • I have always "promoted" the companies on mu Top20 list. Doing that for years. But before, most of the companies on mu Top20 list was high end companies. So used WHT. Just search for my name there, and you will see.

    I did not get paid to write my reviews and do my benchmark. I don't write reviews on request (did it in the start back in 2011/2012). Now I just write about the companies I actually use.

    The test this night, cost me $1.80 of my own money.

  • @rds100 said:
    Which version of UnixBench?

    I have used UnixBench 5.1.3 on all my test. (On CentOS 6.x)

  • @kerouac said:
    I doubt he tests so many servers hourly or monthly priced. Besides they probably gave it to him for free to test.

    I test servers from companies I use. I use my own money, but often has huge credit with each provider I use. Like to have money for 6 - 12 month in my accounts.
    Not getting a dime for doing benchmark and/or reviews for any company.

    So I have much money in play here. Don't need to get paid to do a benchmark.

  • @myhken said:

    Never seen so high benchmark result on any type of server I have tested the last three years.

    what servers on par with this have you tested

  • LeeLee Veteran

    The relevance of a benchmark for a $600 instance on a $7 VPS site?

    Thanked by 3Mark_R kerouac Brad
  • @W1V_Lee said:
    The relevance of a benchmark for a $600 instance on a $7 VPS site?

    It's 1.80 isn't it. Probably need new rules for hourly billed limit?

    :)

    Thanked by 1Lee
  • @W1V_Lee said:
    The relevance of a benchmark for a $600 instance on a $7 VPS site?

    The point is that with Vultr, DO, Iwstack etc you can deploy and run a $600/mo server for some hours, and just pay some few bucks. Thats is a lowendtalk thing.
    I used two hours on my test, and did just pay $1.80, far below any $7 limit here.

    And with that power, you can do quite some work on a server like that, but still, just pay for the time you actually need it.

  • I've used their 16 core one before with great results, very good for quick processing using startup scripts and API.

  • @myhken windows vista??

  • @linuxthefish said:
    myhken windows vista??

    Never liked Vista, but always used their start button as avatar. Have changed it now :D

  • eLohkCalb said: It's 1.80 isn't it. Probably need new rules for hourly billed limit?

    I'm pretty sure the current rules in play already outline it's $7/mo.

  • Still, Vultr.com is a under $7/mo company, with their smallest plan. Can't see the reason why I could not test any of their bigger plans and post the result here?
    I'm not a Vultr employee or have anything with them to do, other then that I have used them for hosting services for almost 10 months now.

    Thanked by 1linuxthefish
  • LeeLee Veteran
    edited December 2014

    myhken said: The point is that with Vultr, DO, Iwstack etc you can deploy and run a $600/mo server for some hours, and just pay some few bucks. Thats is a lowendtalk thing. I used two hours on my test, and did just pay $1.80, far below any $7 limit here.

    Ok, well I suppose using the same logic the $7 limit is monthly, so the cost can therefore be no higher than $0.009 per hour.

    But that aside, it is what it is as always, a benchmark in the Grand scheme of what you intend to use it for is really only for the numbers kids not the people that will really use that size of instance.

  • Not representativer in multiple ways. Just like most pharma studies.

  • @myhken is that all credit just from affiliate sales?

  • @vonlulzweg said:
    myhken is that all credit just from affiliate sales?

    No, most of it is my own money, but used with different coupons, like give us $50 and we give $50. On iwstack, some of it is 30% off, but most is just my own money.
    On Vultr.com I have paid around $175 this year, but my servers is using for around $13/mo. Rest is coupons, and $80 is affilate sales. Just started with that in July, and I now has $140 pending. And unlike with DigitalOcean, most account pending at Vultr.com actually pays out. But this is my stats on DO, and there almost nothing is paid out:

  • Thanks for the benchmark posting! We're all about performance

    Thanked by 1myhken
  • eva2000eva2000 Veteran
    edited December 2014

    @myhken nice you managed to get hold of their multi-core vps - they're not always in stock as I found.

    But 9,600 is a bit slow compared to what I got with Unixbench and Vultr's 16 core (11k Unixbench score) 3-way compare benchmarks at https://community.centminmod.com/threads/digitalocean-vs-linode-vs-vultr-48gb-16-cpus-centminmodbench-sh-results.1389/. Maybe it's that your tests included the file copy results ?

Sign In or Register to comment.