Howdy, Stranger!

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


HostBrr | Two year anniversary deals | AMD EPYC Turin + Block Storage | Flash deals ! - Page 7
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.

HostBrr | Two year anniversary deals | AMD EPYC Turin + Block Storage | Flash deals !

145791053

Comments

  • @NET18 said:

    @maverick said:

    @NET18 said:

    @barbaros said:

    @maverick said:

    @NET18 said:

    @PineappleM said:

    @NET18 said:

    @NHNHNH000 said:

    @NET18 said: poor server performance

    what? it's gb5 in your previous text not gb6

    Here are the GB5 benchmark results for the EPYC 9655 VPS from another provider

    Refund it back if you don’t like it.

    But what price do you pay for the other provider? Same/similar? Hosts can control CPU limits at the hypervisor level.

    Calling it poor performance is a large stretch, especially when most hosts here don’t even exceed 1000 on GB6.

    Dude, this ain't normal performance at all. The AES-XTS is performing like crap, which proves something's definitely wrong here. The disk read/write speeds are way off too. Maybe you're just used to those crappy low-performance servers

    while you might be onto something, man... you're in full panic mode, just don't forget that you have 7 days to refund, don't take your life because you found a bug in GB or something like that :D

    Here, do this on both your turin's and show how crap this one is, i don't have 2 turins to compare, but you can:

    # openssl speed -evp aes-128-xts
    [...]
    type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes  16384 bytes
    AES-128-XTS    1129156.79k  3811952.23k  7514381.48k 11155585.71k 13042322.09k 13371479.38k
    

    From Turin from another provider (3 core AMD EPYC 9655)

    type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes  16384 bytes
    AES-128-XTS    1206917.75k  3967885.85k  7565387.61k 11079769.43k 12758062.42k 13081133.06k
    

    I see minimal difference.

    I'm not sure why the AES-XTS performance was worse in the GB5 benchmark
    https://browser.geekbench.com/v5/cpu/23482852

    go to GB forum and tell them that their software is crap

    ... and apologize to labze for being rude here, when the problem is elsewhere

    Why should I apologize? I used a widely accepted benchmark script and got some unexpected results, so I raised questions. I don't think there's anything wrong with that

    There is correct and wrong ways to solve problems. Also your tone is important. Call it language barrier, or written text.

  • emperoremperor Member

    @lukast__ said: On another note, Turin really has great performance:>!

    Is that custom deal in spoiler ?

  • barbarosbarbaros Member
    edited April 18

    @Blembim said:

    @barbaros said:

    @Blembim said:

    This is le problemo

    Nice profile pic.

    Looks like the lad got squeezed between the sliding elevator doors

  • beanman109beanman109 Member, Megathread Squad

    @lukast__ said:

    @beanman109 said:

    @barbaros said:

    @beanman109 said:

    @beanman109 said:

    @barbaros said:

    @beanman109 said:

    @NHNHNH000 said:

    @beanman109 said:

    @NHNHNH000 said:
    im sorry
    but in which case you guys use aes-xts

    this thread to meme how low i can get my numbers

    beat me
    https://browser.geekbench.com/v5/cpu/23421120

    it's impossibel but rpi2 yabs cuming soon

    RIP

    that 200MB swap is about to do some work

    gg

    Perhaps you should try Geekbench 3?

    cant get 5 or 6 to run even with 2gb of swap and gb4 no arm compatibility this is a disaster

    Let's try this (RPI Zero (2?) W, trying to run GB5+6)...

    # ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
    #              Yet-Another-Bench-Script              #
    #                     v2025-01-01                    #
    # https://github.com/masonr/yet-another-bench-script #
    # ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
    
    Fri 18 Apr 13:33:06 CEST 2025
    
    ARM compatibility is considered *experimental*
    
    Basic System Information:
    ---------------------------------
    Uptime     : 49 days, 15 hours, 39 minutes
    Processor  : ARM1176
    CPU cores  : 1 @ 1000.0000 MHz
    AES-NI     : ❌ Disabled
    VM-x/AMD-V : ❌ Disabled
    RAM        : 427.9 MiB
    Swap       : 4.9 GiB
    Disk       : 58.5 GiB
    Distro     : Raspbian GNU/Linux 12 (bookworm)
    Kernel     : 6.6.51+rpt-rpi-v6
    VM Type    : NONE
    IPv4/IPv6  : ✔ Online / ❌ Offline
    

    On another note, Turin really has great performance:

    # ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
    #              Yet-Another-Bench-Script              #
    #                     v2025-01-01                    #
    # https://github.com/masonr/yet-another-bench-script #
    # ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
    
    Fri Apr 18 12:37:23 CEST 2025
    
    Basic System Information:
    ---------------------------------
    Uptime     : 5 days, 13 hours, 23 minutes
    Processor  : AMD EPYC 9655 96-Core Processor
    CPU cores  : 1 @ 2246.622 MHz
    AES-NI     : ✔ Enabled
    VM-x/AMD-V : ✔ Enabled
    RAM        : 1.9 GiB
    Swap       : 1024.0 MiB
    Disk       : 24.9 GiB
    Distro     : Debian GNU/Linux 12 (bookworm)
    Kernel     : 6.1.0-33-cloud-amd64
    VM Type    : KVM
    IPv4/IPv6  : ✔ Online / ✔ Online
    
    IPv6 Network Information:
    ---------------------------------
    ISP        : dataforest GmbH
    ASN        : AS58212 dataforest GmbH
    Host       : dataforest GmbH
    Location   : Kriftel, Hesse (HE)
    Country    : Germany
    
    fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/vda3):
    ---------------------------------
    Block Size | 4k            (IOPS) | 64k           (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 444.57 MB/s (111.1k) | 749.42 MB/s  (11.7k)
    Write      | 445.74 MB/s (111.4k) | 753.37 MB/s  (11.7k)
    Total      | 890.32 MB/s (222.5k) | 1.50 GB/s    (23.4k)
               |                      |                     
    Block Size | 512k          (IOPS) | 1m            (IOPS)
      ------   | ---            ----  | ----           ---- 
    Read       | 548.12 MB/s   (1.0k) | 1.28 GB/s     (1.2k)
    Write      | 577.25 MB/s   (1.1k) | 1.37 GB/s     (1.3k)
    Total      | 1.12 GB/s     (2.1k) | 2.66 GB/s     (2.6k)
    
    iperf3 Network Speed Tests (IPv4):
    ---------------------------------
    Provider        | Location (Link)           | Send Speed      | Recv Speed      | Ping           
    -----           | -----                     | ----            | ----            | ----           
    Clouvider       | London, UK (10G)          | 7.10 Gbits/sec  | 5.21 Gbits/sec  | 14.0 ms        
    Eranium         | Amsterdam, NL (100G)      | 9.27 Gbits/sec  | 10.1 Gbits/sec  | 8.05 ms        
    Uztelecom       | Tashkent, UZ (10G)        | 1.75 Gbits/sec  | 2.01 Gbits/sec  | 91.6 ms        
    Leaseweb        | Singapore, SG (10G)       | 914 Mbits/sec   | 1.10 Gbits/sec  | 161 ms         
    Clouvider       | Los Angeles, CA, US (10G) | 1.02 Gbits/sec  | 883 Mbits/sec   | 141 ms         
    Leaseweb        | NYC, NY, US (10G)         | 1.95 Gbits/sec  | 2.26 Gbits/sec  | 83.4 ms        
    Edgoo           | Sao Paulo, BR (1G)        | 724 Mbits/sec   | 840 Mbits/sec   | 194 ms         
    
    iperf3 Network Speed Tests (IPv6):
    ---------------------------------
    Provider        | Location (Link)           | Send Speed      | Recv Speed      | Ping           
    -----           | -----                     | ----            | ----            | ----           
    Clouvider       | London, UK (10G)          | 8.07 Gbits/sec  | 8.77 Gbits/sec  | 13.9 ms        
    Eranium         | Amsterdam, NL (100G)      | 9.20 Gbits/sec  | 14.9 Gbits/sec  | 7.90 ms        
    Uztelecom       | Tashkent, UZ (10G)        | 560 Mbits/sec   | 2.01 Gbits/sec  | 91.6 ms        
    Leaseweb        | Singapore, SG (10G)       | 914 Mbits/sec   | 1.09 Gbits/sec  | 161 ms         
    Clouvider       | Los Angeles, CA, US (10G) | 1.08 Gbits/sec  | 1.28 Gbits/sec  | 140 ms         
    Leaseweb        | NYC, NY, US (10G)         | 1.95 Gbits/sec  | 2.24 Gbits/sec  | 83.4 ms        
    Edgoo           | Sao Paulo, BR (1G)        | 721 Mbits/sec   | 890 Mbits/sec   | 198 ms         
    
    Geekbench 6 Benchmark Test:
    ---------------------------------
    Test            | Value                         
                    |                               
    Single Core     | 2468                          
    Multi Core      | 2473                          
    Full Test       | https://browser.geekbench.com/v6/cpu/11572542
    
    YABS completed in 12 min 39 sec
    

    5GB of swap

    Thanked by 2lukast__ Blembim
  • lukast__lukast__ Member, Megathread Squad

    @emperor said:

    @lukast__ said: On another note, Turin really has great performance:>!

    Is that custom deal in spoiler ?

    No, an older deal that was migrated because the hypervisor it was on was overloaded. It might get migrated back to another hypervisor, dunno, but I'll enjoy those nice YABSes while it lasts.

  • @NHNHNH000 said:

    @beanman109 said:

    @NHNHNH000 said:
    im sorry
    but in which case you guys use aes-xts

    this thread to meme how low i can get my numbers

    beat me
    https://browser.geekbench.com/v5/cpu/23421120

    epic

    Thanked by 1Blembim
  • beanman109beanman109 Member, Megathread Squad

    @Blembim said:

    @barbaros said:

    @Blembim said:

    This is le problemo

    Nice profile pic.

    Thanked by 1Blembim
  • @lukast__ said: On another note, Turin really has great performance:>!

    yeah, can agree, i knew i wouldn't resist picking it just to see how it performs, and it delivers above expectations

    compared to EPYC 9654, which is the fastest epyc i had 'til now (and looks as if it's the same 96 core cpu, just a generation older of course), these 4 cores are a match for 6 cores of the genoa, which is amazing

    of course, i perfectly understand that the new host is yet to be fully filled, so the results will drop in time, but still... turin delivers, definitely

    Thanked by 1lukast__
  • amazing power. if only i can afford Turin prices

    Thanked by 2Blembim beanman109
  • @cybertech said:
    amazing power. if only i can afford Turin prices

    Just sell one of your kidneys.

    Thanked by 2Blembim beanman109
  • BlembimBlembim Member

    @beanman109 said:

    @Blembim said:

    @barbaros said:

    @Blembim said:

    This is le problemo

    Nice profile pic.

    The triple regards

    Thanked by 1beanman109
  • beanman109beanman109 Member, Megathread Squad

    @NHNHNH000 said:

    @beanman109 said:

    @NHNHNH000 said:
    im sorry
    but in which case you guys use aes-xts

    this thread to meme how low i can get my numbers

    beat me
    https://browser.geekbench.com/v5/cpu/23421120

    FUCK i was close, i think this can be fine tuned better though

    Geekbench 5 Benchmark Test:
    ---------------------------------
    Test            | Value
                    |
    Single Core     | 73
    Multi Core      | 73
    Full Test       | https://browser.geekbench.com/v5/cpu/23482967
    
  • lukast__lukast__ Member, Megathread Squad

    @lukast__ said: Let's try this (RPI Zero (2?) W, trying to run GB5+6)...

    Okay, failed:

    @NHNHNH000 said:

    @beanman109 said:

    @NHNHNH000 said:
    im sorry
    but in which case you guys use aes-xts

    this thread to meme how low i can get my numbers

    beat me
    https://browser.geekbench.com/v5/cpu/23421120

    I throttled a 9950X VPS to 0.5% of the original performance using cgroups, let's see...

    Thanked by 2Blembim beanman109
  • beanman109beanman109 Member, Megathread Squad

    @lukast__ said:

    @lukast__ said: Let's try this (RPI Zero (2?) W, trying to run GB5+6)...

    Okay, failed:

    @NHNHNH000 said:

    @beanman109 said:

    @NHNHNH000 said:
    im sorry
    but in which case you guys use aes-xts

    this thread to meme how low i can get my numbers

    beat me
    https://browser.geekbench.com/v5/cpu/23421120

    I throttled a 9950X VPS to 0.5% of the original performance using cgroups, let's see...

    ur internet speed is premo

    Thanked by 2lukast__ Blembim
  • beanman109beanman109 Member, Megathread Squad

    @lukast__ said: I throttled a 9950X VPS to 0.5% of the original performance using cgroups, let's see...

    throttled mine down to 3% of a thread for the next run and i made the mistake of rebooting the VM

    it's been 5 minutes

  • _MS__MS_ Member

    @beanman109 said:

    @NHNHNH000 said:

    @beanman109 said:

    @NHNHNH000 said:
    im sorry
    but in which case you guys use aes-xts

    this thread to meme how low i can get my numbers

    beat me
    https://browser.geekbench.com/v5/cpu/23421120

    FUCK i was close, i think this can be fine tuned better though

    Geekbench 5 Benchmark Test:
    ---------------------------------
    Test            | Value
                    |
    Single Core     | 73
    Multi Core      | 73
    Full Test       | https://browser.geekbench.com/v5/cpu/23482967
    

    Geekbench 6 Benchmark Test:
    ---------------------------------
    Test            | Value                        
                    |                              
    Single Core     | 32                           
    Multi Core      | 26                           
    Full Test       | https://browser.geekbench.com/v6/cpu/3906403
    
  • beanman109beanman109 Member, Megathread Squad
    edited April 18

    @MS said:

    @beanman109 said:

    @NHNHNH000 said:

    @beanman109 said:

    @NHNHNH000 said:
    im sorry
    but in which case you guys use aes-xts

    this thread to meme how low i can get my numbers

    beat me
    https://browser.geekbench.com/v5/cpu/23421120

    FUCK i was close, i think this can be fine tuned better though

    Geekbench 5 Benchmark Test:
    ---------------------------------
    Test            | Value
                    |
    Single Core     | 73
    Multi Core      | 73
    Full Test       | https://browser.geekbench.com/v5/cpu/23482967
    

    Geekbench 6 Benchmark Test:
    ---------------------------------
    Test            | Value                        
                    |                              
    Single Core     | 32                           
    Multi Core      | 26                           
    Full Test       | https://browser.geekbench.com/v6/cpu/3906403
    

    3% CPU gb5 coming soon

  • beanman109beanman109 Member, Megathread Squad

    POV: @labze coming back to this thread and seeing everyone trying to get as low as possible geekbench scores

  • nvmenvme Member

    I have 5950x from last year anniversary sale probably, can I order block storage and attach to that one?

  • @sh97 said:
    Happy 2nd anniversary @labze !
    Any chance to see some DA/Reseller dealz for the flash sales?

    @labze Are you thinking?

    I think it would be great if there was a flash deal.

  • gigazilgigazil Member

    begone you mjjs, leave those crappy flash deals to me

  • JabJabJabJab Member

    Shit image. There are deals (a good one I will say) in first post! If you mean FLASH deals then uhmmm put some ChatGPT magic into this image (-:

    Thanked by 2NHNHNH000 Blembim
  • beanman109beanman109 Member, Megathread Squad
  • _MS__MS_ Member

    RIP eyes.

  • beanman109beanman109 Member, Megathread Squad

    @Blembim said: Flash me a deal

    ffs this gif is fkd

  • @beanman109 said:

    @Blembim said: Flash me a deal

    ffs this gif is fkd

    It blinds your eyes so you miss the flashing deals.

Sign In or Register to comment.