Sharon’s New GoMami Hong Kong AMD 9950X VPS Benchmark

Related background can be found in Sharon’s New GoMami Hong Kong AMD 7763 VPS Benchmark

Let’s skip the chatter and jump straight into the benchmarks.

The launch promo code this time is HappyBirthday

Official Site#

Try it now → https://gomami.io

With AFF → https://gomami.io/aff.php?aff=4

Plan Details#

Provider : GoMami
Type/Plan : HKG.Peak.X5.Mini
Processor : AMD Ryzen 9 9950X 16-Core Processor
Num of Core : 2 Cores
Memory : 4 GB
Storage : 4****0 GB NVMe
Bandwidth : 1****TB(only out) @ 1 Gbps IN | 1 Gbps OUT ( Shared )
Location : HK
Price : $59.00 USD / $50.15 USD (Use HappyBirthday)

Benchmark#

Compared to the AMD 7763 lineup, this instance has IO capped at about half. The CPU scores match a nearly full AMD 9950X and are very close to my bare-metal DuFu box.

Yabs#

1
Basic System Information:
2
---------------------------------
3
Uptime     : 0 days, 0 hours, 12 minutes
4
Processor  : AMD Ryzen 9 9950X 16-Core Processor
5
CPU cores  : 2 @ 4291.912 MHz
6
AES-NI     : ✔ Enabled
7
VM-x/AMD-V : ✔ Enabled
8
RAM        : 3.8 GiB
9
Swap       : 0.0 KiB
10
Disk       : 39.9 GiB
11
Distro     : Debian GNU/Linux 12 (bookworm)
12
Kernel     : 6.1.0-9-amd64
13
VM Type    : KVM
14
IPv4/IPv6  : ✔ Online / ❌ Offline
15

16
IPv4 Network Information:
17
---------------------------------
18
ISP        : Next Hop LLC
19
ASN        : AS36002 Next Hop LLC
20
Host       : Telescope Communications
21
Location   : Hong Kong, Kowloon ()
22
Country    : Hong Kong
23

24
fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/sda3):
25
---------------------------------
26
Block Size | 4k            (IOPS) | 64k           (IOPS)
27
  ------ | --- ---- | ---- ----
28
Read       | 399.31 MB/s  (99.8k) | 430.46 MB/s   (6.7k)
29
Write      | 400.36 MB/s (100.0k) | 432.73 MB/s   (6.7k)
30
Total      | 799.67 MB/s (199.9k) | 863.20 MB/s  (13.4k)
31
           |                      |
32
Block Size | 512k          (IOPS) | 1m            (IOPS)
33
  ------ | --- ---- | ---- ----
34
Read       | 520.34 MB/s   (1.0k) | 561.11 MB/s    (547)
35
Write      | 547.99 MB/s   (1.0k) | 598.48 MB/s    (584)
36
Total      | 1.06 GB/s     (2.0k) | 1.15 GB/s     (1.1k)
37

38
Geekbench 5 BenchMark Test:
39
---------------------------------
40
Test            | Value
41
                |
42
Single Core     | 2566
43
Multi Core      | 4520
44
Full Test       | https://browser.geekbench.com/v5/cpu/23723730
45

46
Geekbench 6 BenchMark Test:
47
---------------------------------
48
Test            | Value
49
                |
50
Single Core     | 3419
51
Multi Core      | 5974
52
Full Test       | https://browser.geekbench.com/v6/cpu/13293079

PassMark PerformanceTest Linux#

1
AMD Ryzen 9 9950X 16-Core Processor (x86_64)
2
2 cores @ 0 MHz  |  3.8 GiB RAM
3
Number of Processes: 2  |  Test Iterations: 1  |  Test Duration: Medium
4
--------------------------------------------------------------------------------
5
CPU Mark:                          8451
6
  Integer Math                     18926 Million Operations/s
7
  Floating Point Math              18796 Million Operations/s
8
  Prime Numbers                    71.0 Million Primes/s
9
  Sorting                          10974 Thousand Strings/s
10
  Encryption                       3959 MB/s
11
  Compression                      80655 KB/s
12
  CPU Single Threaded              4660 Million Operations/s
13
  Physics                          1221 Frames/s
14
  Extended Instructions (SSE)      6789 Million Matrices/s
15

16
Memory Mark:                       2058
17
  Database Operations              3237 Thousand Operations/s
18
  Memory Read Cached               42468 MB/s
19
  Memory Read Uncached             41153 MB/s
20
  Memory Write                     19809 MB/s
21
  Available RAM                    2000 Megabytes
22
  Memory Latency                   65 Nanoseconds
23
  Memory Threaded                  45291 MB/s
24
--------------------------------------------------------------------------------

UnixBench#

1
========================================================================
2
   BYTE UNIX BenchMarks (Version 6.0.0)
3

4
   System: catcat.blog: GNU/Linux
5
   OS: GNU/Linux -- 6.1.0-9-amd64 -- #1 SMP PREEMPT_DYNAMIC Debian 6.1.27-1 (2023-05-08)
6
   Machine: x86_64 (unknown)
7
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
8
   CPU 0: AMD Ryzen 9 9950X 16-Core Processor (8583.8 bogomips)
9
          x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
10
   CPU 1: AMD Ryzen 9 9950X 16-Core Processor (8583.8 bogomips)
11
          x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
12
   11:23:12 up 44 min,  2 users,  load average: 0.00, 0.13, 0.17; runlevel 2025-08-13
13

14
------------------------------------------------------------------------
15
BenchMark Run: Wed Aug 13 2025 11:23:12 - 11:51:30
16
2 CPUs in system; running 1 parallel copy of tests
17

18
Dhrystone 2 using register variables      106400208.4 lps   (10.0 s, 7 samples)
19
Double-Precision Whetstone                    13708.8 MWIPS (10.0 s, 7 samples)
20
Execl Throughput                               8881.1 lps   (30.0 s, 2 samples)
21
File Copy 1024 bufsize 2000 maxblocks       2442971.4 KBps  (30.0 s, 2 samples)
22
File Copy 256 bufsize 500 maxblocks          617173.5 KBps  (30.0 s, 2 samples)
23
File Copy 4096 bufsize 8000 maxblocks       8594560.0 KBps  (30.0 s, 2 samples)
24
Pipe Throughput                             4437671.4 lps   (10.0 s, 7 samples)
25
Pipe-based Context Switching                 299976.3 lps   (10.0 s, 7 samples)
26
Process Creation                              27085.2 lps   (30.0 s, 2 samples)
27
Shell Scripts (1 concurrent)                  28053.5 lpm   (60.0 s, 2 samples)
28
Shell Scripts (8 concurrent)                   5735.7 lpm   (60.0 s, 2 samples)
29
System Call Overhead                        4178434.1 lps   (10.0 s, 7 samples)
30

31
System BenchMarks Index Values               BASELINE       RESULT    INDEX
32
Dhrystone 2 using register variables         116700.0  106400208.4   9117.4
33
Double-Precision Whetstone                       55.0      13708.8   2492.5
34
Execl Throughput                                 43.0       8881.1   2065.4
35
File Copy 1024 bufsize 2000 maxblocks          3960.0    2442971.4   6169.1
36
File Copy 256 bufsize 500 maxblocks            1655.0     617173.5   3729.1
37
File Copy 4096 bufsize 8000 maxblocks          5800.0    8594560.0  14818.2
38
Pipe Throughput                               12440.0    4437671.4   3567.3
39
Pipe-based Context Switching                   4000.0     299976.3    749.9
40
Process Creation                                126.0      27085.2   2149.6
41
Shell Scripts (1 concurrent)                     42.4      28053.5   6616.4
42
Shell Scripts (8 concurrent)                      6.0       5735.7   9559.5
43
System Call Overhead                          15000.0    4178434.1   2785.6
44
                                                                   ========
45
System BenchMarks Index Score                                        3988.6
46

47
------------------------------------------------------------------------
48
BenchMark Run: Wed Aug 13 2025 11:51:30 - 12:19:48
49
2 CPUs in system; running 2 parallel copies of tests
50

51
Dhrystone 2 using register variables      210625400.2 lps   (10.0 s, 7 samples)
52
Double-Precision Whetstone                    27238.4 MWIPS (10.0 s, 7 samples)
53
Execl Throughput                              16785.1 lps   (30.0 s, 2 samples)
54
File Copy 1024 bufsize 2000 maxblocks       4798590.8 KBps  (30.0 s, 2 samples)
55
File Copy 256 bufsize 500 maxblocks         1220281.8 KBps  (30.0 s, 2 samples)
56
File Copy 4096 bufsize 8000 maxblocks      13384231.3 KBps  (30.0 s, 2 samples)
57
Pipe Throughput                             8766242.1 lps   (10.0 s, 7 samples)
58
Pipe-based Context Switching                 861876.9 lps   (10.0 s, 7 samples)
59
Process Creation                              56212.2 lps   (30.0 s, 2 samples)
60
Shell Scripts (1 concurrent)                  43194.2 lpm   (60.0 s, 2 samples)
61
Shell Scripts (8 concurrent)                   5899.4 lpm   (60.0 s, 2 samples)
62
System Call Overhead                        8260592.4 lps   (10.0 s, 7 samples)
63

64
System BenchMarks Index Values               BASELINE       RESULT    INDEX
65
Dhrystone 2 using register variables         116700.0  210625400.2  18048.4
66
Double-Precision Whetstone                       55.0      27238.4   4952.4
67
Execl Throughput                                 43.0      16785.1   3903.5
68
File Copy 1024 bufsize 2000 maxblocks          3960.0    4798590.8  12117.7
69
File Copy 256 bufsize 500 maxblocks            1655.0    1220281.8   7373.3
70
File Copy 4096 bufsize 8000 maxblocks          5800.0   13384231.3  23076.3
71
Pipe Throughput                               12440.0    8766242.1   7046.8
72
Pipe-based Context Switching                   4000.0     861876.9   2154.7
73
Process Creation                                126.0      56212.2   4461.3
74
Shell Scripts (1 concurrent)                     42.4      43194.2  10187.3
75
Shell Scripts (8 concurrent)                      6.0       5899.4   9832.3
76
System Call Overhead                          15000.0    8260592.4   5507.1
77
                                                                   ========
78
System BenchMarks Index Score                                        7398.1

IO Test#

Compared to the 7763-based plan, the IO difference is quite large.

1
root@catcat:~# dd if=/dev/zero of=testfile bs=1M count=10240
2
10240+0 records in
3
10240+0 records out
4
10737418240 bytes (11 GB, 10 GiB) copied, 12.6975 s, 846 MB/s
5
root@catcat:~#

IP Quality#

I won’t benchmark this again, since the IPs are from the same batch and should be identical.

WordPress Benchmark#

PHP8.0#

PHP8.4#

Summary#

Compared to the HKG.Pulse.Mini AMD 7763 plan, the extra $20 here basically all goes into CPU performance. The only real downside is the capped IO — running the same fio command to generate a 10G test file shows a 3x gap. That said, for non-extreme workloads like WordPress and typical app services, it’s still more than sufficient.