[Landing Node] DDPS Tokyo, Japan IIJ | Native IP | R9-9900X | 10Gbps | International Connectivity Review

DDPS is commonly seen in two data centers: CDN77 (Datacamp) and IIJ. Today I am testing DDPS's IIJ data center.

International connectivity is in section 3.

For the CDN77 data center, see my other review:

DDPS Tokyo, Japan CDN77/Datacamp | Native IP | R9-7900 | 10Gbps | International Connectivity Review

1. Configuration

The plan says R9-7900, but in practice this provider's IIJ data center should all be 9900X, with somewhat better performance.

Provider website: http://ddps.jp

🔗 Quick links

Note: the provider does not seem to care much about individual users. Provisioning and ticket response times are extremely long; you will probably receive IIJ, and tickets to change DC may not be handled.

upstreamis IIJ

2. Node Quality

💻Basic information

R9-9900X, very strong performance, with single-core GB5 28.7% higher than the other CDN77 R9-7900 machine.

Basic System Information:
---------------------------------
Uptime     : 1 days, 3 hours, 38 minutes
Processor  : AMD Ryzen 9 9900X 12-Core Processor
CPU cores  : 1 @ 4391.720 MHz
AES-NI     : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM        : 929.7 MiB
Swap       : 976.5 MiB
Disk       : 21.0 GiB
Distro     : Debian GNU/Linux 12 (bookworm)
Kernel     : 6.12.57+deb13-cloud-amd64
VM Type    : STANDARD PC (I440FX + PIIX, 1996)
IPv4/IPv6  : ✔ Online / ✔ Online

IPv6 Network Information:
---------------------------------
ISP        : DDPS Networks, LLC
ASN        : AS18526 DDPS Networks, LLC
Host       : DDPS Networks, LLC
Location   : Tokyo, Tokyo (13)
Country    : Japan

fio Disk Speed Tests (Mixed R/W 50/50) (Partition -):
---------------------------------
Block Size | 4k            (IOPS) | 64k           (IOPS)
  ------   | ---            ----  | ----           ---- 
Read       | 181.31 MB/s  (45.3k) | 2.19 GB/s    (34.2k)
Write      | 181.79 MB/s  (45.4k) | 2.20 GB/s    (34.4k)
Total      | 363.11 MB/s  (90.7k) | 4.39 GB/s    (68.6k)
           |                      |                     
Block Size | 512k          (IOPS) | 1m            (IOPS)
  ------   | ---            ----  | ----           ---- 
Read       | 3.07 GB/s     (6.0k) | 3.17 GB/s     (3.1k)
Write      | 3.24 GB/s     (6.3k) | 3.38 GB/s     (3.3k)
Total      | 6.31 GB/s    (12.3k) | 6.56 GB/s     (6.4k)

Geekbench 5 Benchmark Test:
---------------------------------
Test            | Value                         
                |                               
Single Core     | 2447                          
Multi Core      | 2424                          
Full Test       | https://browser.geekbench.com/v5/cpu/23926520

 SysBench CPU Test (Fast Mode, 1-Pass @ 5sec)
---------------------------------
 1 Thread Test (Single Core) Score:          6753 Scores
 SysBench Memory Test (Fast Mode, 1-Pass @ 5sec)
---------------------------------
 Single-thread Read Test:          105771.05 MB/s
 Single-thread Write Test:          39988.30 MB/s

🎬IP quality

Japan native IP, with good unlocking.

IPv4	IPv6

🌐Network quality

Tested during the daytime; it looks usable for direct connection then. During evening peak, definitely not.

IPv4	IPv6

📍Return route

IPv4	IPv6

3. International connectivity - iPerf3 TCP IPv4 Single-threadspeed test (10 seconds per run, 2025-11-26)

Compared with its own CDN77 model, the IIJ model has local latency 1ms+ higher, and latency to Hong Kong and the US West Coast is also higher. Local bandwidth is similar to the CDN77 model, while cross-border speeds are generally lower than CDN77.

[Test records]

Japan-US connectivity

The IIJ model has relatively high latency to the United States. BAGE LA1 and LA2 return through GTT and TATA respectively; both Seattle routes return through COGENT and have no latency advantage.

🇯🇵 IIJ <-> 🇺🇸 BAGE LA1	🇯🇵 IIJ <-> 🇺🇸 BAGE LA2	🇯🇵 IIJ <-> 🇺🇸 Misaka SEA	🇯🇵 IIJ <-> 🇺🇸 AlphaVPS SEA
Outbound	Outbound	Outbound	Outbound
Return	Return	Return	Return

Hong Kong-Japan connectivity

Pulling the IIJ model from Hong Kong also has relatively high latency. Hong Kong DMIT and RFC JINX both go through GSL<->Lumen<->IIJ, with higher latency than direct GSL. Bage HKS and Halo STD (Akari) both return via PCCW, and latency does not look good.

🇭🇰 DMIT <-> 🇯🇵 IIJ	🇭🇰 RFCHOST JINX <-> 🇯🇵 IIJ	🇭🇰 BAGE HKS <-> 🇯🇵 IIJ	🇭🇰 Halo STD <-> 🇯🇵 IIJ
Outbound	Outbound	Outbound	Outbound
Return	Return	Return	Return

4. Summary

Native IP, good unlocking, and a large port. IIJ is acceptable for Telecom/Mobile direct connection during daytime, but do not expect it at night. I recommend using it directly as a landing node. International connectivity is not as good as the CDN77 model; if you care about latency, it is better to pull it from a Japan route machine or a Shanghai-Japan private line.

If you need performance, this model is much stronger than CDN77, and you can buy a higher configuration in the same series.

5. About the tests

The machines tested in this article, including all machines used for iPerf3 speed tests, were purchased at my own expense. Results may be time-sensitive.
Kernel parameters of the tested machines (/etc/sysctl.conf) are as follows

fs.file-max = 6815744
net.ipv4.tcp_no_metrics_save=1
net.ipv4.tcp_ecn=0
net.ipv4.tcp_frto=0
net.ipv4.tcp_mtu_probing=0
net.ipv4.tcp_rfc1337=0
net.ipv4.tcp_sack=1
net.ipv4.tcp_fack=1
net.ipv4.tcp_window_scaling=1
net.ipv4.tcp_adv_win_scale=1
net.ipv4.tcp_moderate_rcvbuf=1
net.core.rmem_max=33554432
net.core.wmem_max=33554432
net.ipv4.tcp_rmem=4096 87380 33554432
net.ipv4.tcp_wmem=4096 16384 33554432
net.ipv4.udp_rmem_min=8192
net.ipv4.udp_wmem_min=8192
net.ipv4.ip_forward=1
net.ipv4.conf.all.route_localnet=1
net.ipv4.conf.all.forwarding=1
net.ipv4.conf.default.forwarding=1
net.core.default_qdisc=fq
net.ipv4.tcp_congestion_control=bbr
net.ipv6.conf.all.forwarding=1
net.ipv6.conf.default.forwarding=1