Equinix United States Los Angeles High Performance Bare Metal Server Test and Review
We recently put the Equinix United States Los Angeles High-Performance Bare Metal Server through a series of tests to check out its performance. This server runs on an Intel Xeon E-2278G 8-Core processor at 3.40GHz. It’s equipped with 32 GB RAM, two 480GB SSDs for storage, dual 10Gbps internet ports, Intel HD Graphics P630, one IPv4 address, and one IPv6 address.
We ran Basic Benchmark, Yabs, and Unix Benchmark to see how the server handles different workloads. We also checked the server’s connectivity across various locations with the Global Node Test and evaluated its response times with the Global Ping Test.
Additionally, we tested media accessibility through the Media Unlock Test and internet connection quality via the CloudFlare Internet Test. For insights into web performance, we used both PageSpeed Insights and GTmetrix.
About Equinix
Imagine a world where you can deploy powerful, interconnected bare metal infrastructure in minutes. That’s the world Equinix Metal has created with its Bare Metal On Demand Servers. These servers are not just servers; they are curated for common workloads and available on-demand in minutes in popular metros.
Equinix offers a variety of CPUs to cater to different needs. For instance, the a3.large.x86 server comes with 2 x Intel Xeon 6338, providing 64 cores at 2.0 GHz. On the other hand, the c3.large.arm64 server is equipped with an Ampere Altra Q80-30 CPU, boasting 80 cores at 3.0 GHz. These powerful CPUs ensure that your workloads run smoothly and efficiently.
But what about memory and storage? Equinix has got you covered. The a3.large.x86 server comes with a whopping 1 TB of memory and 2 x 240 GB SSD for boot. The c3.large.arm64 server, on the other hand, offers 256 GB RAM and 2 x 960 GB NVMe for boot. With such ample memory and storage, you can be sure that your applications will run without a hitch.
Equinix understands that different workloads require different network capabilities. That’s why they offer a range of network options. For example, the a3.large.x86 server provides 4 x 25 Gbps network, while the c3.large.arm64 server offers 2 x 25 Gbps network. This flexibility lets you choose the server that best fits your network requirements.
Official Website: https://www.Equinix.com
BasicBench: Server Profile, I/O Speed, Upload and Download Speed
CPU Model : Intel(R) Xeon(R) E-2278G CPU @ 3.40GHz
CPU Cores : 16 @ 927.082 MHz
CPU Cache : 16384 KB
AES-NI : ✓ Enabled
VM-x/AMD-V : ✓ Enabled
Total Disk : 439.1 GB (1.7 GB Used)
Total Mem : 31.2 GB (695.9 MB Used)
Total Swap : 1.9 GB (0 Used)
System uptime : 0 days, 0 hour 8 min
Load average : 0.22, 0.18, 0.10
OS : Debian GNU/Linux 11
Arch : x86_64 (64 Bit)
Kernel : 5.10.0-28-amd64
TCP CC : bbr
Virtualization : Dedicated
IPv4/IPv6 : ✓ Online / ✓ Online
Organization : AS54825 Packet Host, Inc.
Location : Los Angeles / US
Region : California
----------------------------------------------------------------------
I/O Speed(1st run) : 397 MB/s
I/O Speed(2nd run) : 407 MB/s
I/O Speed(3rd run) : 429 MB/s
I/O Speed(average) : 411.0 MB/s
----------------------------------------------------------------------
Node Name Upload Speed Download Speed Latency
Speedtest.net 33536.67 Mbps 47093.94 Mbps 1.36 ms
Los Angeles, US 43500.43 Mbps 46432.38 Mbps 1.44 ms
Dallas, US 3078.46 Mbps 17853.88 Mbps 29.89 ms
Montreal, CA 637.56 Mbps 921.70 Mbps 68.38 ms
Amsterdam, NL 684.58 Mbps 3122.76 Mbps 132.10 ms
Shanghai, CN 904.37 Mbps 8900.45 Mbps 171.42 ms
Chongqing, CN 0.49 Mbps 0.49 Mbps 245.66 ms
Hongkong, CN 563.72 Mbps 5611.90 Mbps 161.07 ms
Mumbai, IN 762.83 Mbps 4181.62 Mbps 252.49 ms
Singapore, SG 115.77 Mbps 29.12 Mbps 175.56 ms
Tokyo, JP 447.50 Mbps 6852.50 Mbps 153.94 ms
----------------------------------------------------------------------
Finished in : 5 min 20 sec
Timestamp : 2024-05-15 00:16:08 UTC
YabsBench: Disk Speed, iperf3 Network Speed, Geekbench 6 Benchmark
Basic System Information:
---------------------------------
Uptime : 0 days, 0 hours, 13 minutes
Processor : Intel(R) Xeon(R) E-2278G CPU @ 3.40GHz
CPU cores : 16 @ 800.240 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 31.2 GiB
Swap : 1.9 GiB
Disk : 437.2 GiB
Distro : Debian GNU/Linux 11 (bullseye)
Kernel : 5.10.0-28-amd64
VM Type : NONE
IPv4/IPv6 : ✔ Online / ✔ Online
IPv6 Network Information:
---------------------------------
ISP : Packet Host, Inc.
ASN : AS54825 Packet Host, Inc.
Host : Packet Host, Inc.
Location : Los Angeles, California (CA)
Country : United States
fio Disk Speed Tests (Mixed R/W 50/50):
---------------------------------
Block Size | 4k (IOPS) | 64k (IOPS)
------ | --- ---- | ---- ----
Read | 151.38 MB/s (37.8k) | 153.18 MB/s (2.3k)
Write | 151.78 MB/s (37.9k) | 153.99 MB/s (2.4k)
Total | 303.16 MB/s (75.7k) | 307.18 MB/s (4.7k)
| |
Block Size | 512k (IOPS) | 1m (IOPS)
------ | --- ---- | ---- ----
Read | 203.63 MB/s (397) | 221.66 MB/s (216)
Write | 214.45 MB/s (418) | 236.42 MB/s (230)
Total | 418.09 MB/s (815) | 458.09 MB/s (446)
iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
Clouvider | London, UK (10G) | 1.36 Gbits/sec | 1.73 Gbits/sec | 131 ms
Eranium | Amsterdam, NL (100G) | 3.04 Gbits/sec | 2.32 Gbits/sec | --
Telia | Helsinki, FI (10G) | busy | busy | 151 ms
Uztelecom | Tashkent, UZ (10G) | 1.76 Gbits/sec | 1.39 Gbits/sec | 235 ms
Leaseweb | Singapore, SG (10G) | 2.68 Gbits/sec | 2.16 Gbits/sec | 163 ms
Clouvider | Los Angeles, CA, US (10G) | 9.36 Gbits/sec | 8.21 Gbits/sec | 1.04 ms
Leaseweb | NYC, NY, US (10G) | 7.81 Gbits/sec | 6.00 Gbits/sec | 60.2 ms
Edgoo | Sao Paulo, BR (1G) | 1.87 Gbits/sec | 1.55 Gbits/sec | 221 ms
iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
Clouvider | London, UK (10G) | 1.37 Gbits/sec | 1.74 Gbits/sec | 131 ms
Eranium | Amsterdam, NL (100G) | busy | busy | --
Uztelecom | Tashkent, UZ (10G) | 1.00 Gbits/sec | 1.33 Gbits/sec | 235 ms
Leaseweb | Singapore, SG (10G) | 2.65 Gbits/sec | 2.11 Gbits/sec | 163 ms
Clouvider | Los Angeles, CA, US (10G) | 9.07 Gbits/sec | 7.95 Gbits/sec | 1.03 ms
Leaseweb | NYC, NY, US (10G) | 7.38 Gbits/sec | 5.76 Gbits/sec | 60.1 ms
Edgoo | Sao Paulo, BR (1G) | 1.82 Gbits/sec | 1.51 Gbits/sec | 224 ms
Geekbench 6 Benchmark Test:
---------------------------------
Test | Value
|
Single Core | 1888
Multi Core | 8578
Full Test | https://browser.geekbench.com/v6/cpu/6102314
YABS completed in 15 min 53 sec
UnixBench: CPU Performance, A Score Higher is Better
Benchmark Run: Wed May 15 2024 00:32:46 - 01:00:43
16 CPUs in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 64124006.0 lps (10.0 s, 7 samples)
Double-Precision Whetstone 9590.7 MWIPS (9.9 s, 7 samples)
Execl Throughput 8644.6 lps (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 1808888.7 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 481224.1 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 5111428.4 KBps (30.0 s, 2 samples)
Pipe Throughput 3475674.3 lps (10.0 s, 7 samples)
Pipe-based Context Switching 359344.7 lps (10.0 s, 7 samples)
Process Creation 19022.0 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 5722.5 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 10336.9 lpm (60.0 s, 2 samples)
System Call Overhead 3828026.6 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 64124006.0 5494.8
Double-Precision Whetstone 55.0 9590.7 1743.8
Execl Throughput 43.0 8644.6 2010.4
File Copy 1024 bufsize 2000 maxblocks 3960.0 1808888.7 4567.9
File Copy 256 bufsize 500 maxblocks 1655.0 481224.1 2907.7
File Copy 4096 bufsize 8000 maxblocks 5800.0 5111428.4 8812.8
Pipe Throughput 12440.0 3475674.3 2794.0
Pipe-based Context Switching 4000.0 359344.7 898.4
Process Creation 126.0 19022.0 1509.7
Shell Scripts (1 concurrent) 42.4 5722.5 1349.6
Shell Scripts (8 concurrent) 6.0 10336.9 17228.1
System Call Overhead 15000.0 3828026.6 2552.0
========
System Benchmarks Index Score 2988.2
------------------------------------------------------------------------
Benchmark Run: Wed May 15 2024 01:00:43 - 01:28:46
16 CPUs in system; running 16 parallel copies of tests
Dhrystone 2 using register variables 539037518.1 lps (10.0 s, 7 samples)
Double-Precision Whetstone 125264.5 MWIPS (10.2 s, 7 samples)
Execl Throughput 60121.1 lps (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 1646987.9 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 425318.6 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 5288483.1 KBps (30.0 s, 2 samples)
Pipe Throughput 26950806.9 lps (10.0 s, 7 samples)
Pipe-based Context Switching 3498736.6 lps (10.0 s, 7 samples)
Process Creation 133046.7 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 131571.0 lpm (60.0 s, 2 samples)
Shell Scripts (8 concurrent) 17416.5 lpm (60.0 s, 2 samples)
System Call Overhead 10791181.7 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 539037518.1 46190.0
Double-Precision Whetstone 55.0 125264.5 22775.4
Execl Throughput 43.0 60121.1 13981.6
File Copy 1024 bufsize 2000 maxblocks 3960.0 1646987.9 4159.1
File Copy 256 bufsize 500 maxblocks 1655.0 425318.6 2569.9
File Copy 4096 bufsize 8000 maxblocks 5800.0 5288483.1 9118.1
Pipe Throughput 12440.0 26950806.9 21664.6
Pipe-based Context Switching 4000.0 3498736.6 8746.8
Process Creation 126.0 133046.7 10559.3
Shell Scripts (1 concurrent) 42.4 131571.0 31030.9
Shell Scripts (8 concurrent) 6.0 17416.5 29027.5
System Call Overhead 15000.0 10791181.7 7194.1
========
System Benchmarks Index Score 12757.5
======= Script description and score comparison completed! =======
Looking Glass
None
Global Latency Test
Query: ping 136.144.61.163, interval: 5s, dense mode: no
Probing 136.144.61.163 – RIPENCC /23 AS54825 PACKET – United States, California, Los Angeles
Global Node Test
## CDN Speedtest
CacheFly : 79.24 MiB/s | 633.91 Mbps | ping 0.813ms
Gdrive : 8.47 KiB/s | 0.07 Mbps | ping 1.080ms
## North America Speedtest
Softlayer, Washington, USA : 1.21 MiB/s | 9.71 Mbps | ping 60.549ms
SoftLayer, San Jose, USA : 74.31 MiB/s | 594.45 Mbps | ping 11.329ms
SoftLayer, Dallas, USA : 0 B/s | N/A | ping error!
Vultr, New Jersey, USA : 956.04 KiB/s | 7.47 Mbps | ping 59.928ms
Vultr, Seattle, USA : 133.87 MiB/s | 1070.95 Mbps | ping 28.834ms
Vultr, Dallas, USA : 111.73 MiB/s | 893.85 Mbps | ping 31.110ms
Vultr, Los Angeles, USA : 276.24 MiB/s | 2209.94 Mbps | ping 0.994ms
Ramnode, New York, USA : 36.75 MiB/s | 294.00 Mbps | ping 60.043ms
Ramnode, Atlanta, USA : 48.51 MiB/s | 388.07 Mbps | ping 49.012ms
## Europe Speedtest
Vultr, London, UK : 27.69 MiB/s | 221.48 Mbps | ping 132.461ms
LeaseWeb, Frankfurt, Germany : 20.68 KiB/s | 0.16 Mbps | ping 139.107ms
Hetzner, Germany : 135 B/s | N/A | ping 150.463ms
Ramnode, Alblasserdam, NL : 15.29 MiB/s | 122.31 Mbps | ping 146.722ms
Vultr, Amsterdam, NL : 25.92 MiB/s | 207.36 Mbps | ping 140.631ms
EDIS, Stockholm, Sweden : 305 B/s | N/A | ping 156.595ms
OVH, Roubaix, France : 352 B/s | N/A | ping 809.371ms
Online, France : 24.73 MiB/s | 197.81 Mbps | ping 146.506ms
Prometeus, Milan, Italy : 69 B/s | N/A | ping 164.873ms
## Exotic Speedtest
Sydney, Australia : 8.59 MiB/s | 68.74 Mbps | ping 241.552ms
Lagoon, New Caledonia : 182 B/s | N/A | ping 184.335ms
Hosteasy, Moldova : 5.72 MiB/s | 45.80 Mbps | ping 176.986ms
Prima, Argentina : 144.20 KiB/s | 1.13 Mbps | ping error!
## Asia Speedtest
SoftLayer, Singapore : 262.21 KiB/s | 2.05 Mbps | ping 184.578ms
Linode, Tokyo, Japan : 18.42 MiB/s | 147.36 Mbps | ping 106.354ms
Linode, Singapore : 8.10 MiB/s | 64.83 Mbps | ping 189.972ms
Vultr, Tokyo, Japan : 34.98 MiB/s | 279.82 Mbps | ping 109.602ms
Benchmark finished in 133 seconds
Media Unlock Test
** Test Starts At: Wed May 15 00:06:53 UTC 2024
** Checking Results Under IPv4
--------------------------------
** Your Network Provider: Packet Host (136.144.*.*)
============[ Multination ]============
Dazn: Yes (Region: US)
Disney+: No
Netflix: Originals Only
YouTube Premium: Yes
Amazon Prime Video: Yes (Region: US)
TVBAnywhere+: Yes
iQyi Oversea Region: US
YouTube CDN: Los Angeles, CA
Netflix Preferred CDN: Los Angeles, CA
Spotify Registration: No
Steam Currency: USD
ChatGPT: Yes
Bing Region: US
Wikipedia Editability: Yes
Instagram Licensed Audio: No
---Forum---
Reddit: No
=======================================
** Checking Results Under IPv6
--------------------------------
** Your Network Provider: Packet Host (2604:1380:4611:*:*)
============[ Multination ]============
Dazn: Failed (Network Connection)
Disney+: No
Netflix: Originals Only
YouTube Premium: Yes
Amazon Prime Video: Unsupported
TVBAnywhere+: Failed (Network Connection)
iQyi Oversea Region: Failed
YouTube CDN: Los Angeles, CA
Netflix Preferred CDN: Los Angeles, CA
Spotify Registration: No
Steam Currency: Failed (Network Connection)
ChatGPT: Failed
Bing Region: US
Wikipedia Editability: Yes
Instagram Licensed Audio: Yes
---Forum---
Reddit: Failed (Network Connection)
=======================================
Testing Done! CloudFlare Internet Test
PageSpeed Insights
GTmetrix Performance
Admin Control Panel
The Equinix Bare Metal portal provides full visibility of your environment. This transparency lets you perform high-end testing and validate complex networking configurations within private cloud solutions. Imagine having a bird’s eye view of your entire infrastructure, enabling you to monitor and manage your resources effectively.
Secondly, the portal gives you full control over Layer 2 networking. This means you can manage how data is transferred over a network, which is crucial for maintaining high performance and security. It’s like having the steering wheel to your network infrastructure, allowing you to navigate and control your data traffic.
Moreover, the portal enables rapid provisioning. You can set up and deploy a single service on Equinix Metal in just minutes. A full end-to-end deployment can be up and running in an hour. This means you can market applications faster. It’s like having a fast-forward button for your deployments, saving you valuable time.
In addition, the portal offers on-demand connectivity. You can connect a database or application tier on any cloud to another one in the cloud and internal and external cloud services. This feature lets you connect your resources as needed, ensuring optimal performance.
Lastly, the portal ensures security. Equinix Metal offers 100% dedicated single-tenancy so that you can isolate sensitive workloads and access hardware-level security protocols. It’s like having a fortress for your data, providing peace of mind knowing your data is secure.
Server Pricing
Official Website: https://www.Equinix.com
Payments
Equinix offers a flexible billing structure that caters to your specific needs. You can choose between On Demand and On Contract billing types. With On Demand, you pay only for what you need and get billed per hour. On the other hand, On Contract offers competitive pricing with monthly payments. This flexibility lets you choose the billing type best suits your budget and usage.
The cost of On Demand servers is a fixed rate per hour. Billing begins when the server is successfully provisioned and ends when the server has been deleted from your project. This means you only pay for the time your server is active, providing you with cost-effective solutions.
Moreover, Equinix offers a Pricing Calculator. This tool allows you to estimate the cost of an On-Demand deployment. It’s like having a financial advisor at your fingertips, helping you make informed decisions about infrastructure costs.
In addition, Equinix charges for additional services like Managed Server IP and Global Elastic IP. The first Managed Server IP is free, while additional Elastic IPs with Equinix Metal IP space cost $0.005/hr ($3.60/mo). Global Elastic IPs cost $0.15/hr ($109.50/mo). These charges ensure you have the necessary resources for your workload.
Coupon
$250 credit* Use code DEPLOYNOW to test
*A credit card is required, but you’ll only be charged for usage beyond any applied promo credit. Purchase orders and offline payment methods (wire transfer, etc.) are available via sales engagement.
Customer Support
Equinix provides 24/7 live chat or email support with technical experts. You can contact them anytime or night, ensuring that help is always available. It’s like having a dedicated support team right at your fingertips, ready to assist you whenever you need.
They offer many resources to help you get started and maximize their services. These resources provide valuable insights and guidance, from product documentation to technical guides. It’s like having a comprehensive library of knowledge, helping you easily navigate the world of Equinix Bare Metal On Demand Servers.
Moreover, Equinix has a dedicated Contact Sales page. You can contact their sales team to discuss your needs and find the best solutions here. It’s like having a personal consultant who understands your requirements and can guide you towards the right products and services.
In addition, Equinix has a vibrant community where you can find a wealth of information and insights. The community is a treasure trove of knowledge, from discussions on deployment options to videos comparing On-Demand and Reserved Server options. It’s like being part of a global network of users and experts sharing their experiences and insights.
Conclusion
Our I/O speed tests averaged 411 MB/s, which, while not reaching the excellent mark, still maintains a decent standard above 400 MB/s. Network performance revealed high speed and low latency nearby nodes like those in Los Angeles and Dallas. In contrast, international nodes, particularly in Asia, showed varied results with slower speeds and higher latencies, some being minimal.
The Yabs performance benchmark focused on mixed read/write operations, demonstrating adequate performance, hovering around the good baseline for larger block sizes, notably achieving 458 MB/s for mixed reads and writes with a 1 MB block size. Meanwhile, the Geekbench 6 scores presented a mixed picture; the single core score of 1888 indicates good performance that falls just shy of excellent. However, a very strong multi-core score of 8578 reflects the server’s ability to handle intensive multi-threaded processes efficiently.
From our UnixBench results, the server demonstrated exceptional scalability when all 16 CPUs were engaged, dramatically increasing the handling of tasks across more demanding contexts. Highlights include significantly higher scores in file copying for large buffers and superior throughput in pipelines, which are crucial for data-heavy operations.
In conclusion, while the Equinix server showcases strong multi-threading and good I/O performance, improvements in single-threading capabilities could enhance its versatility. Ideal for applications requiring robust multi-thread processing, it remains a solid choice for businesses needing reliable and efficient performance in high-demand scenarios. Future considerations might include optimizing settings to boost individual core performance and elevate usability in single-threaded applications.
Official Website: https://www.Equinix.com
Related Posts
