Oplink 美国 AMD 7950X 测评

这家争议挺大，主要矛盾点在于安装费(黑五活动可以免设置费)。商家的意思安装费是用于筛选客户，只保留精致的用户。新用户可以购买测试3天，如果不满意可以全额退款，包括设置费。就目前买过机器的人而言，表现是很好的。确实做到了像他描述的那样，因为门槛高，所以相对而已邻居的素质也更高。网络而言并不是很好，

套餐#

Provider : Oplink
Type/Plan : Premium VPS 7950x
Processor : AMD Ryzen 9 7950X (4.5 GHz++)
Num of Core : 6 Cores
Memory : 16 GB DDR5 RAM
Storage : 500 GB NVMe(PCIe 4.0)
Bandwidth : 32TB @ 10G Mbps IN | 10G Mbps OUT
Location : US
Price : $29.95 USD + $4.95 Setup Fee

测评#

Yabs#

1
---------------------------------
2
Uptime     : 0 days, 0 hours, 20 minutes
3
Processor  : AMD Ryzen 9 7950X 16-Core Processor
4
CPU cores  : 6 @ 4499.980 MHz
5
AES-NI     : ✔ Enabled
6
VM-x/AMD-V : ✔ Enabled
7
RAM        : 15.6 GiB
8
Swap       : 0.0 KiB
9
Disk       : 484.5 GiB
10
Distro     : Ubuntu 22.04.3 LTS
11
Kernel     : 5.15.0-92-generic
12
VM Type    : KVM
13
IPv4/IPv6  : ✔ Online / ✔ Online
14

15
IPv6 Network Information:
16
---------------------------------
17
ISP        : The Optimal Link Corporation
18
ASN        : AS40156 The Optimal Link Corporation
19
Host       : The Optimal Link Corporation
20
Location   : Houston, Texas (TX)
21
Country    : United States
22

23
fio Disk Speed Tests (Mixed R/W 50/50):
24
---------------------------------
25
Block Size | 4k            (IOPS) | 64k           (IOPS)
26
  ------ | --- ---- | ---- ----
27
Read       | 479.26 MB/s (119.8k) | 5.56 GB/s    (86.9k)
28
Write      | 480.52 MB/s (120.1k) | 5.59 GB/s    (87.3k)
29
Total      | 959.79 MB/s (239.9k) | 11.15 GB/s  (174.2k)
30
           |                      |
31
Block Size | 512k          (IOPS) | 1m            (IOPS)
32
  ------ | --- ---- | ---- ----
33
Read       | 8.10 GB/s    (15.8k) | 8.45 GB/s     (8.2k)
34
Write      | 8.53 GB/s    (16.6k) | 9.01 GB/s     (8.8k)
35
Total      | 16.64 GB/s   (32.5k) | 17.47 GB/s   (17.0k)
36

37
iperf3 Network Speed Tests (IPv4):
38
---------------------------------
39
Provider        | Location (Link)           | Send Speed      | Recv Speed      | Ping
40
Clouvider       | London, UK (10G)          | busy            | busy            | 102 ms
41
Scaleway        | Paris, FR (10G)           | 1.53 Gbits/sec  | 1.73 Gbits/sec  | 104 ms
42
NovoServe       | North Holland, NL (40G)   | 1.40 Gbits/sec  | 1.38 Gbits/sec  | 115 ms
43
Uztelecom       | Tashkent, UZ (10G)        | 676 Mbits/sec   | 435 Mbits/sec   | 210 ms
44
Clouvider       | NYC, NY, US (10G)         | 5.13 Gbits/sec  | 2.03 Gbits/sec  | 35.1 ms
45
Clouvider       | Dallas, TX, US (10G)      | 7.62 Gbits/sec  | 4.44 Gbits/sec  | 5.90 ms
46
Clouvider       | Los Angeles, CA, US (10G) | 4.76 Gbits/sec  | 1.64 Gbits/sec  | 36.2 ms
47

48
iperf3 Network Speed Tests (IPv6):
49
---------------------------------
50
Provider        | Location (Link)           | Send Speed      | Recv Speed      | Ping
51
Clouvider       | London, UK (10G)          | 1.51 Gbits/sec  | busy            | 102 ms
52
Scaleway        | Paris, FR (10G)           | busy            | busy            | 104 ms
53
NovoServe       | North Holland, NL (40G)   | 1.37 Gbits/sec  | 1.45 Gbits/sec  | 115 ms
54
Uztelecom       | Tashkent, UZ (10G)        | busy            | busy            | 210 ms
55
Clouvider       | NYC, NY, US (10G)         | 5.25 Gbits/sec  | 4.32 Gbits/sec  | 35.1 ms
56
Clouvider       | Dallas, TX, US (10G)      | 8.12 Gbits/sec  | 4.84 Gbits/sec  | 5.87 ms
57
Clouvider       | Los Angeles, CA, US (10G) | 4.68 Gbits/sec  | 1.67 Gbits/sec  | 36.3 ms
58

59
Geekbench 6 BenchMark Test:
60
---------------------------------
61
Test            | Value
62
                |
63
Single Core     | 2740
64
Multi Core      | 9614

Bench#

1
 CPU Model          : AMD Ryzen 9 7950X 16-Core Processor
2
 CPU Cores          : 6 @ 4499.980 MHz
3
 CPU Cache          : 512 KB
4
 AES-NI             : ✓ Enabled
5
 VM-x/AMD-V         : ✓ Enabled
6
 Total Disk         : 484.5 GB (1.6 GB Used)
7
 Total Mem          : 15.6 GB (221.6 MB Used)
8
 System uptime      : 0 days, 0 hour 2 min
9
 Load average       : 0.39, 0.13, 0.05
10
 OS                 : Ubuntu 22.04.3 LTS
11
 Arch               : x86_64 (64 Bit)
12
 Kernel             : 5.15.0-92-generic
13
 TCP CC             : bbr
14
 Virtualization     : KVM
15
 IPv4/IPv6          : ✓ Online / ✓ Online
16
 Organization       : AS40156 The Optimal Link Corporation
17
 Location           : Houston / US
18
 Region             : Texas
19
----------------------------------------------------------------------
20
 I/O Speed(1st run) : 1.5 GB/s
21
 I/O Speed(2nd run) : 1.8 GB/s
22
 I/O Speed(3rd run) : 2.2 GB/s
23
 I/O Speed(average) : 1877.3 MB/s
24
----------------------------------------------------------------------
25
 Node Name        Upload Speed      Download Speed      Latency
26
 Speedtest.net    7586.17 Mbps      8161.37 Mbps        0.75 ms
27
 Los Angeles, US  2138.41 Mbps      3854.96 Mbps        43.46 ms
28
 Dallas, US       7399.73 Mbps      5130.61 Mbps        5.25 ms
29
 Montreal, CA     880.51 Mbps       882.29 Mbps         38.90 ms
30
 Paris, FR        750.25 Mbps       5508.00 Mbps        107.29 ms
31
 Amsterdam, NL    735.68 Mbps       2748.66 Mbps        112.95 ms
32
 Shanghai, CN     16.49 Mbps        1684.88 Mbps        183.91 ms
33
 Mumbai, IN       369.22 Mbps       2146.11 Mbps        224.30 ms
34
 Singapore, SG    395.75 Mbps       3723.43 Mbps        201.61 ms
35
 Tokyo, JP        604.17 Mbps       2078.46 Mbps        136.35 ms
36
----------------------------------------------------------------------

byte-unixbench 性能测试#

1
6 CPUs in system; running 1 parallel copy of tests
2

3
Dhrystone 2 using register variables       76446801.9 lps   (10.0 s, 7 samples)
4
Double-Precision Whetstone                    12709.8 MWIPS (9.7 s, 7 samples)
5
Execl Throughput                               7853.7 lps   (29.8 s, 2 samples)
6
File Copy 1024 bufsize 2000 maxblocks       2370804.7 KBps  (30.0 s, 2 samples)
7
File Copy 256 bufsize 500 maxblocks          636404.6 KBps  (30.0 s, 2 samples)
8
File Copy 4096 bufsize 8000 maxblocks       6747492.3 KBps  (30.0 s, 2 samples)
9
Pipe Throughput                             3531683.8 lps   (10.0 s, 7 samples)
10
Pipe-based Context Switching                 257257.9 lps   (10.0 s, 7 samples)
11
Process Creation                              14310.5 lps   (30.0 s, 2 samples)
12
Shell Scripts (1 concurrent)                  15682.1 lpm   (60.0 s, 2 samples)
13
Shell Scripts (8 concurrent)                   8165.5 lpm   (60.0 s, 2 samples)
14
System Call Overhead                        2829431.9 lps   (10.0 s, 7 samples)
15

16
System BenchMarks Index Values               BASELINE       RESULT    INDEX
17
Dhrystone 2 using register variables         116700.0   76446801.9   6550.7
18
Double-Precision Whetstone                       55.0      12709.8   2310.9
19
Execl Throughput                                 43.0       7853.7   1826.4
20
File Copy 1024 bufsize 2000 maxblocks          3960.0    2370804.7   5986.9
21
File Copy 256 bufsize 500 maxblocks            1655.0     636404.6   3845.3
22
File Copy 4096 bufsize 8000 maxblocks          5800.0    6747492.3  11633.6
23
Pipe Throughput                               12440.0    3531683.8   2839.0
24
Pipe-based Context Switching                   4000.0     257257.9    643.1
25
Process Creation                                126.0      14310.5   1135.8
26
Shell Scripts (1 concurrent)                     42.4      15682.1   3698.6
27
Shell Scripts (8 concurrent)                      6.0       8165.5  13609.2
28
System Call Overhead                          15000.0    2829431.9   1886.3
29
                                                                   ========
30
System BenchMarks Index Score                                        3263.2
31

32
------------------------------------------------------------------------
33

34
6 CPUs in system; running 6 parallel copies of tests
35

36
Dhrystone 2 using register variables      453129785.0 lps   (10.0 s, 7 samples)
37
Double-Precision Whetstone                    76366.0 MWIPS (10.0 s, 7 samples)
38
Execl Throughput                              26883.7 lps   (30.0 s, 2 samples)
39
File Copy 1024 bufsize 2000 maxblocks       1740142.7 KBps  (30.0 s, 2 samples)
40
File Copy 256 bufsize 500 maxblocks          498479.6 KBps  (30.0 s, 2 samples)
41
File Copy 4096 bufsize 8000 maxblocks       3541943.4 KBps  (30.0 s, 2 samples)
42
Pipe Throughput                            20177471.4 lps   (10.0 s, 7 samples)
43
Pipe-based Context Switching                1435090.5 lps   (10.0 s, 7 samples)
44
Process Creation                              52064.3 lps   (30.0 s, 2 samples)
45
Shell Scripts (1 concurrent)                  73699.7 lpm   (60.0 s, 2 samples)
46
Shell Scripts (8 concurrent)                  10572.2 lpm   (60.0 s, 2 samples)
47
System Call Overhead                        7923054.7 lps   (10.0 s, 7 samples)
48

49
System BenchMarks Index Values               BASELINE       RESULT    INDEX
50
Dhrystone 2 using register variables         116700.0  453129785.0  38828.6
51
Double-Precision Whetstone                       55.0      76366.0  13884.7
52
Execl Throughput                                 43.0      26883.7   6252.0
53
File Copy 1024 bufsize 2000 maxblocks          3960.0    1740142.7   4394.3
54
File Copy 256 bufsize 500 maxblocks            1655.0     498479.6   3012.0
55
File Copy 4096 bufsize 8000 maxblocks          5800.0    3541943.4   6106.8
56
Pipe Throughput                               12440.0   20177471.4  16219.8
57
Pipe-based Context Switching                   4000.0    1435090.5   3587.7
58
Process Creation                                126.0      52064.3   4132.1
59
Shell Scripts (1 concurrent)                     42.4      73699.7  17382.0
60
Shell Scripts (8 concurrent)                      6.0      10572.2  17620.4
61
System Call Overhead                          15000.0    7923054.7   5282.0
62
                                                                   ========
63
System BenchMarks Index Score                                        8294.1
64

65
======= Script description and score comparison completed! =======

全局网络测试#

1
 ## CDN Speedtest
2

3
 CacheFly :  84.76 MiB/s |  678.07 Mbps  | ping  27.885ms
4
 Gdrive   :   9.10 KiB/s |    0.07 Mbps  | ping   6.064ms
5

6
 ## North America Speedtest
7

8
 Softlayer, Washington, USA :  13.61 MiB/s |  108.91 Mbps  | ping  30.331ms
9
 SoftLayer, San Jose, USA   :  25.47 MiB/s |  203.74 Mbps  | ping  43.611ms
10
 SoftLayer, Dallas, USA     :       0 B/s |      N/A       | ping error!
11
 Vultr, New Jersey, USA     :   4.10 MiB/s |   32.83 Mbps  | ping  33.675ms
12
 Vultr, Seattle, USA        :  41.21 MiB/s |  329.69 Mbps  | ping  55.324ms
13
 Vultr, Dallas, USA         :  50.77 MiB/s |  406.18 Mbps  | ping   5.651ms
14
 Vultr, Los Angeles, USA    :  61.74 MiB/s |  493.89 Mbps  | ping  35.765ms
15
 Ramnode, New York, USA     :  37.10 MiB/s |  296.78 Mbps  | ping  34.492ms
16
 Ramnode, Atlanta, USA      :  87.22 MiB/s |  697.75 Mbps  | ping  18.817ms
17

18
 ## Europe Speedtest
19

20
 Vultr, London, UK            :  22.15 MiB/s |  177.16 Mbps  | ping 101.015ms
21
 LeaseWeb, Frankfurt, Germany :  39.69 KiB/s |    0.31 Mbps  | ping 114.489ms
22
 Hetzner, Germany             :     297 B/s |      N/A       | ping 114.407ms
23
 Ramnode, Alblasserdam, NL    :  10.65 MiB/s |   85.21 Mbps  | ping 113.979ms
24
 Vultr, Amsterdam, NL         :  19.18 MiB/s |  153.47 Mbps  | ping 112.635ms
25
 EDIS, Stockholm, Sweden      :     400 B/s |      N/A       | ping 137.679ms
26
 OVH, Roubaix, France         :     363 B/s |      N/A       | ping 110.873ms
27
 Online, France               : 501.25 KiB/s |    3.92 Mbps  | ping 104.054ms
28
 Prometeus, Milan, Italy      :     418 B/s |      N/A       | ping 121.831ms
29

30
 ## Exotic Speedtest
31

32
 Sydney, Australia     :   6.32 MiB/s |   50.54 Mbps  | ping 233.600ms
33
 Lagoon, New Caledonia :     292 B/s |      N/A       | ping 235.014ms
34
 Hosteasy, Moldova     :   4.86 MiB/s |   38.85 Mbps  | ping 149.269ms
35
 Prima, Argentina      : 280.71 KiB/s |    2.19 Mbps  | ping error!
36

37
 ## Asia Speedtest
38

39
 SoftLayer, Singapore : 141.00 KiB/s |    1.10 Mbps  | ping 205.761ms
40
 Linode, Tokyo, Japan :  12.03 MiB/s |   96.24 Mbps  | ping 141.534ms
41
 Linode, Singapore    :   6.70 MiB/s |   53.58 Mbps  | ping 232.179ms
42
 Vultr, Tokyo, Japan  :  14.27 MiB/s |  114.16 Mbps  | ping 150.413ms