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.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Bero host Ryzen 9950x vs Turin
itachikonoha
Member
in Reviews
With Black Friday offers coming in, I wanted to test the Berohost Ryzen 9950x vps vs Turin server and results were as follows.
Both have 4 cores. Ryzen VPS is a year old while Epyc Turin is in empty node (have confirmed it with Berohost). Both running debian 13
Ryzen 9950x
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
# Yet-Another-Bench-Script #
# v2025-04-20 #
# https://github.com/masonr/yet-another-bench-script #
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
Tue Dec 2 12:23:46 IST 2025
Basic System Information:
---------------------------------
Uptime : 1 days, 3 hours, 52 minutes
Processor : AMD Ryzen 9 9950X 16-Core Processor
CPU cores : 4 @ 4291.948 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 7.8 GiB
Swap : 4.0 GiB
Disk : 49.1 GiB
Distro : Debian GNU/Linux 13 (trixie)
Kernel : 6.12.57+deb13-amd64
VM Type : KVM
IPv4/IPv6 : ✔ Online / ✔ Online
IPv6 Network Information:
---------------------------------
ISP : Oliver Horscht is trading as \"SYNLINQ\"
ASN : AS44486 synlinq.de
Host : Rubv6
Location : Frankfurt am Main, Hesse (HE)
Country : Germany
fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/sda1):
---------------------------------
Block Size | 4k (IOPS) | 64k (IOPS)
------ | --- ---- | ---- ----
Read | 388.56 MB/s (97.1k) | 1.99 GB/s (31.2k)
Write | 389.59 MB/s (97.3k) | 2.00 GB/s (31.3k)
Total | 778.16 MB/s (194.5k) | 4.00 GB/s (62.5k)
| |
Block Size | 512k (IOPS) | 1m (IOPS)
------ | --- ---- | ---- ----
Read | 390.75 MB/s (763) | 4.45 GB/s (4.3k)
Write | 411.51 MB/s (803) | 4.74 GB/s (4.6k)
Total | 802.27 MB/s (1.5k) | 9.19 GB/s (8.9k)
iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Clouvider | London, UK (10G) | 2.84 Gbits/sec | 3.02 Gbits/sec | 12.0 ms
Eranium | Amsterdam, NL (100G) | 4.78 Gbits/sec | 4.81 Gbits/sec | 6.80 ms
Uztelecom | Tashkent, UZ (10G) | 806 Mbits/sec | 2.32 Gbits/sec | 88.7 ms
Leaseweb | Singapore, SG (10G) | 706 Mbits/sec | 1.50 Gbits/sec | 155 ms
Clouvider | Los Angeles, CA, US (10G) | 645 Mbits/sec | 1.37 Gbits/sec | 140 ms
Leaseweb | NYC, NY, US (10G) | 921 Mbits/sec | 2.99 Gbits/sec | 82.5 ms
Edgoo | Sao Paulo, BR (1G) | 508 Mbits/sec | 1.29 Gbits/sec | 191 ms
iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Clouvider | London, UK (10G) | 4.31 Gbits/sec | 4.50 Gbits/sec | 11.8 ms
Eranium | Amsterdam, NL (100G) | 4.99 Gbits/sec | 4.73 Gbits/sec | 6.78 ms
Uztelecom | Tashkent, UZ (10G) | 773 Mbits/sec | 2.16 Gbits/sec | 88.6 ms
Leaseweb | Singapore, SG (10G) | 589 Mbits/sec | 1.51 Gbits/sec | 155 ms
Clouvider | Los Angeles, CA, US (10G) | 594 Mbits/sec | 1.32 Gbits/sec | 138 ms
Leaseweb | NYC, NY, US (10G) | 749 Mbits/sec | 2.97 Gbits/sec | 82.5 ms
Edgoo | Sao Paulo, BR (1G) | 362 Mbits/sec | 1.14 Gbits/sec | 197 ms
Geekbench 6 Benchmark Test:
---------------------------------
Test | Value
|
Single Core | 3159
Multi Core | 9003
Full Test | https://browser.geekbench.com/v6/cpu/15346201
YABS completed in 10 min 48 sec
Epyc turin
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
# Yet-Another-Bench-Script #
# v2025-04-20 #
# https://github.com/masonr/yet-another-bench-script #
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
Tue Dec 2 06:51:06 UTC 2025
Basic System Information:
---------------------------------
Uptime : 0 days, 4 hours, 41 minutes
Processor : AMD EPYC 9275F 24-Core Processor
CPU cores : 4 @ 4100.000 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 11.7 GiB
Swap : 0.0 KiB
Disk : 147.6 GiB
Distro : Debian GNU/Linux 13 (trixie)
Kernel : 6.12.57+deb13-amd64
VM Type : KVM
IPv4/IPv6 : ✔ Online / ✔ Online
IPv6 Network Information:
---------------------------------
ISP : Oliver Horscht is trading as \"SYNLINQ\"
ASN : AS44486 synlinq.de
Host : Rubv6
Location : Frankfurt am Main, Hesse (HE)
Country : Germany
fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/sda1):
---------------------------------
Block Size | 4k (IOPS) | 64k (IOPS)
------ | --- ---- | ---- ----
Read | 548.12 MB/s (137.0k) | 5.37 GB/s (84.0k)
Write | 549.57 MB/s (137.3k) | 5.40 GB/s (84.4k)
Total | 1.09 GB/s (274.4k) | 10.78 GB/s (168.4k)
| |
Block Size | 512k (IOPS) | 1m (IOPS)
------ | --- ---- | ---- ----
Read | 7.62 GB/s (14.8k) | 6.63 GB/s (6.4k)
Write | 8.02 GB/s (15.6k) | 7.07 GB/s (6.9k)
Total | 15.65 GB/s (30.5k) | 13.70 GB/s (13.3k)
iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Clouvider | London, UK (10G) | 3.97 Gbits/sec | 4.86 Gbits/sec | 11.7 ms
Eranium | Amsterdam, NL (100G) | 4.67 Gbits/sec | 4.97 Gbits/sec | 6.77 ms
Uztelecom | Tashkent, UZ (10G) | 580 Mbits/sec | 2.35 Gbits/sec | 84.6 ms
Leaseweb | Singapore, SG (10G) | 422 Mbits/sec | 1.48 Gbits/sec | 155 ms
Clouvider | Los Angeles, CA, US (10G) | 487 Mbits/sec | 1.42 Gbits/sec | 138 ms
Leaseweb | NYC, NY, US (10G) | 520 Mbits/sec | 2.98 Gbits/sec | 83.1 ms
Edgoo | Sao Paulo, BR (1G) | 427 Mbits/sec | 1.27 Gbits/sec | 193 ms
iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Clouvider | London, UK (10G) | 3.16 Gbits/sec | 4.63 Gbits/sec | 11.7 ms
Eranium | Amsterdam, NL (100G) | 4.93 Gbits/sec | 4.90 Gbits/sec | 6.71 ms
Uztelecom | Tashkent, UZ (10G) | 672 Mbits/sec | 2.38 Gbits/sec | 84.6 ms
Leaseweb | Singapore, SG (10G) | 474 Mbits/sec | 1.40 Gbits/sec | 155 ms
Clouvider | Los Angeles, CA, US (10G) | 533 Mbits/sec | 1.53 Gbits/sec | 138 ms
Leaseweb | NYC, NY, US (10G) | 859 Mbits/sec | 2.84 Gbits/sec | 83.0 ms
Edgoo | Sao Paulo, BR (1G) | 396 Mbits/sec | 1.09 Gbits/sec | 197 ms
Geekbench 6 Benchmark Test:
---------------------------------
Test | Value
|
Single Core | 3001
Multi Core | 9453
Full Test | https://browser.geekbench.com/v6/cpu/15346162
YABS completed in 10 min 41 sec
For ryzen 9950x GB6 score of 3000+ was expected but didn't think 9275f could match the single score.
So i thought of using it real life usage.
Transcode file:
Handbrake config:
x265
slower preset
crf 26
resolution 1080
9275f completed encoding in 12 minutes 37 seconds
Ryzen 9950x completed encoding in 18 minutes 31 seconds
In terms of single score, even though 9950x outperforms turin, Berohost turin is giving a good performance.
Expecting the GB 6 to go below 2500 as the node is starting to fill up.
But at ~8 euro per month, it is a very good deal and value for money.
Comments
I think my Turin chicken never have 3000+ score, worse I get about 2000+ now.
Indeed, this score for epyc Turin is for an empty node. I have two hero hosts (one bday, the other one bf) on 9275F and I am in the range of ~2100 and ~2300 single core.
Tough Geekbench is only an indication and workload depends on other specs, as you demonstrated you got better results with the Epic one compared to the Ryzen one even though the score was higher for the other.
Still a solid offer tough!
also please stop running GB6, GB5 is max what you should use for 'servers'.
GB6 went full balls deep into desktop things.
Hi,
this is most probably because the physical hostmaschine is quiet empty. So you benefit fully from turbo and all kind of fancy stuff that is available as long as the work pressure wont become too hard.
The Ryzen should be usually faster under similar conditions which is pretty clear as it has 1 GHz more cpu frequency.
I thought turin was about 2300 to 2500
In the end, which processor is best for single core score? Possibly also for hosting websites on that server, let's say full-node?
That's hard to say, Ryzen has less cores which divide between less customers but EPYC has higher TDP.
I'd say the ryzen one more likely to have better single core performance on full node.
The epyc is also good but the more there are cores, more likely the ghz base clock is reduced, that will be all the time for the case of a full node. You can look at that on epyc turin specsheets.
Well, one thing is that the Ryzen CPU is an desktop one and EPYC it's for servers?
If you need performance in single core, it is ryzen 9950x.
Only exception I found was that of BERO. I think node is still empty so shooting up the GB6. I will give a month or two more to check the real performance under stress.
My past experiences using Bero was there were some kind of lags quite noticeable using Windows Server or Linux under XRDP. That was under Ryzen on early 2025
A week passed, I tried this one, no issue I found like earlier ..
Just a low bandwidth capacity to South East Asia compared to European Region or NYC US and UK.
I agreed on the month or two for testing the vps.
Usually after two, its hard to maintain for three or four
im right,u have problem with almost all provider u paid
Performance remains stable over time
Basic System Information:
---------------------------------
Uptime : 150 days, 15 hours, 11 minutes
Processor : AMD Ryzen 9 9950X 16-Core Processor
CPU cores : 4 @ 4291.948 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 7.8 GiB
Swap : 4.0 GiB
Disk : 145.5 GiB
Distro : Ubuntu 24.04.2 LTS
Kernel : 6.8.0-63-generic
VM Type : KVM
IPv4/IPv6 : ✔ Online / ✔ Online
Today there is Ryzen promo offer.
Go go go!
Price slightly increased compared to the birthday period
I guess you will fill the whole forum with just problem
Mine got the same for 3rd week
Is it back to normal?
My VPS has been restarting randomly a few times recently.
From my location in Indonesia
I am seeing connection fine, no reboots happening for at least these three weeks.
Their upstream speed to Singapore and South East Asia also good now.
400-500Mbps after I raised a ticket to ask them for good upstream speed.
Hopefully it worked for next month or next year.
Yes, we have made further optimizations to the network in recent weeks and months. Connections to Asia, especially China and Singapore, should now be significantly faster and have better latency.