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.
Comments
I re-ran this benchmark since @MAXKO_Hosting was kind enough to do CPU passthrough on my VM. It's slightly better now performance-wise.
cc: @forest
https://www.robovps.biz/en/hi_cpu_vps_in_finland/372-antares.html
~26.93$ /month
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Yet-Another-Bench-Script
v2025-04-20
https://github.com/masonr/yet-another-bench-script
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Thu Mar 19 09:27:24 PM MSK 2026
Basic System Information:
Uptime : 65 days, 1 hours, 49 minutes
Processor : AMD Ryzen 9 9950X 16-Core Processor
CPU cores : 8 @ 4291.920 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 7.8 GiB
Swap : 0.0 KiB
Disk : 177.1 GiB
Distro : Debian GNU/Linux 13 (trixie)
Kernel : 6.12.57+deb13-amd64
VM Type : KVM
IPv4/IPv6 : ✔ Online / ✔ Online
IPv6 Network Information:
ISP : Scalaxy B.V.
ASN : AS58061 Scalaxy B.V.
Host : Gtelcom LLC
Location : Helsinki, Uusimaa (18)
Country : Finland
fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/vda2):
iperf3 Network Speed Tests (IPv4):
iperf3 Network Speed Tests (IPv6):
Running GB6 benchmark test... cue elevator music
Geekbench 6 Benchmark Test:
Test | Value
|
Single Core | 3245
Multi Core | 13842
Full Test | https://browser.geekbench.com/v6/cpu/17162600
YABS completed in 11 min 27 sec
€60 / year AMD Ryzen VPS:
https://www.robovps.biz/en/vps_in_russia/206-mercury.html
~ 2,25$ / month
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Yet-Another-Bench-Script
v2025-04-20
https://github.com/masonr/yet-another-bench-script
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Thu Mar 19 10:21:43 PM MSK 2026
Basic System Information:
Uptime : 0 days, 6 hours, 3 minutes
Processor : Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz
CPU cores : 2 @ 2397.222 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 967.1 MiB
Swap : 1024.0 MiB
Disk : 24.6 GiB
Distro : Debian GNU/Linux 13 (trixie)
Kernel : 6.12.57+deb13-amd64
VM Type : KVM
IPv4/IPv6 : ✔ Online / ❌ Offline
IPv4 Network Information:
ISP : JSC \"ER-Telecom Holding\"
ASN : AS25446 JSC ER-Telecom Holding
Host : TK Rustelekom LLC
Location : Moscow, Moscow (MOW)
Country : Russia
fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/vda2):
iperf3 Network Speed Tests (IPv4):
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Clouvider | London, UK (10G) | 221 Kbits/sec | busy | 50.2 ms
Eranium | Amsterdam, NL (100G) | 218 Mbits/sec | 189 Mbits/sec | 41.6 ms
Uztelecom | Tashkent, UZ (10G) | 216 Mbits/sec | 148 Mbits/sec | 55.4 ms
Leaseweb | Singapore, SG (10G) | 199 Mbits/sec | 107 Mbits/sec | 176 ms
Performing IPv4 iperf3 recv test from Clouvider (Attempt #3 of 3)...
Clouvider | Los Angeles, CA, US (10G) | 219 Kbits/sec | busy | 188 ms
Leaseweb | NYC, NY, US (10G) | 208 Mbits/sec | 143 Mbits/sec | 107 ms
Edgoo | Sao Paulo, BR (1G) | 193 Mbits/sec | 91.2 Mbits/sec | 229 ms
Running GB6 benchmark test... cue elevator music
Geekbench test failed and low memory was detected. Add at least 1GB of SWAP or use GB4 instead (higher compatibility with low memory systems).
YABS completed in 14 min 21 sec
Hetzner Cloud, CX23 in Falkenstein, Germany
DeluxHost T-3
SWCVPS NYC & AU 15$/year
NYC
AU
@SolidVPS Intel Core i3-4150 / 8GB RAM / 250GB SSD / 1Gbps Unmetered / 1 IPv4 / Instant Delivery / Deal Price $12.95/m
zgh.cl ~60$/yr
I think it's my personal record for the amount of time that GB6 needed, but makes sense as missing RAM is SWAP + that ANCIENT CPU.
seedhost.eu VS 2 in Frankfurt
With GB results:
infofractal AMD SEV-SNP enabled VPS in Santiago, Chile 🇨🇱 from this comment. @jfrac
swiotlb set to 16MB to save the RAM (swiotlb=8192). IO performance is great even with smaller swiotlb buffers (it was 64MB by default).
SolidVPS $12.95 USD/monthly
onidel's ONI-2 plan in Vietnam (🇻🇳) a few weeks after launch, now with IPv6
PipeHost's VPS S with Ryzen 9000 in Maribor, Slovenia 🇸🇮
@MattBGP
ps. They have IPv6 prefix allocated, will be enabled on VPS servers a few days from what I've been told.
edit: it was enabled within a few minutes; YABS updated with IPv6 now
@oloke please redo again but with ipv6 disabled
Wanna also try my Ryzen 9000-powered PipeBombs for free?
Then could you enable another preset such as Haswell? Using the default qemu64 is a Pentium-era preset. Even Sandy Bridge would be enough to get quite a few new features without causing breakage.
Greencloud 12th Birthday sale
@berohost RYZEN POWER package
Guess I'm on a pretty busy node
UltraVPS 25Y-Storage-1 in NL
@MAXKO_Hosting in Belgrade, Serbia 🇷🇸 with 10 Gbps unmetered networking
test ipv4: 150.40.126.1
test ipv6: 2a0e:bfc7:300::1
Still showing the location and country incorrectly. Which GeoIP database is used by the YABS test?
YABS is doing some weird things and inaccurate GeoIP issues are pretty common in results. In case of Maxko Serbia, it tries to query ip-api.com with its IPv6 address (as IPv6 is default network on newer Debians):
If it queried IPv4 instead, this would get correctly recognized as Serbia:
Definitely not the first time seeing wrong location on YABS. I feel this would be more accurate using @Operable4829 's ipinfo.io API, but there must've been a reason authors of yabs.sh didn't go this route.
Another YABS of @MAXKO_Hosting new Belgrade, Serbia 🇷🇸 loc
Over 10 Gbps
(don't mind my broken IPv6 i'm testing stuff hehe)
And yours geoip is correct since no IPv6
DON'T BLAME IT ON IPV6 😭
(even though you're right)
@zGato , @oloke since you have same deal, is that custom made or ? Cant find it on their site. Too bad we don't have direct peering with Serbia, goes to Slovakia and back.
Good clue, the team has submit geo correction for IPv6 2a0e:bfc7:300::/40.
We waiting your BGP in Serbia to be up
Continuous 24/7 benchmarking can become more of a stress test for the entire environment than a meaningful performance measurement. It doesn’t always reflect real-world workloads, and in shared or small VPS environments it can end up impacting neighbors, disk I/O, or even triggering automated abuse systems.