APX-SYNC300 Data Center Time Synchronization System
Product Overview
The APEX APX-SYNC300 is a data centre time synchronization system. It takes its reference from GPS and BeiDou and distributes time inside the building through PTP and NTP, so servers, storage systems, switches and management platforms all work from one traceable clock.
A modern data centre runs on records that have to line up. Application logs, database transactions, job schedules, security events and network telemetry are collected on different machines and later merged into a single timeline. If each machine adjusts its own clock against a public server, the merged timeline contains ordering errors that no amount of log tooling can repair, and the cause is nearly impossible to prove after the fact.
The APX-SYNC300 removes that variable. It locks to the satellite reference, disciplines its internal oscillator, and then serves the estate: PTP over Ethernet for equipment that needs tight time and phase, NTP and SNTP for the majority of servers and controllers, and 1PPS and 10MHz outputs for instruments and test equipment that expect a physical reference. A rooftop mounted antenna, cable and surge protection are supplied as a matched set, and the unit reports reference and synchronization status so a problem in the chain is visible before it reaches the applications.
It is used in enterprise and cloud data centres, in colocation halls, in AI and high performance computing clusters where distributed jobs need a shared time base, and in pairs of sites where the same timeline has to hold across a primary and a secondary hall.
Key Features
- One time source for servers, storage, switches and management platforms
- GNSS reference from GPS and BeiDou with oscillator disciplining
- PTP over Ethernet for latency sensitive and time stamping equipment
- NTP and SNTP service sized for a large population of clients
- 1PPS and 10MHz outputs for instruments and test equipment
- Reference priority with automatic switching and holdover if the satellite reference is lost
- Status and alarm reporting so the reference chain is monitored, not assumed
- Antenna, low loss cable and surge protection supplied as one matched set
- Suitable for enterprise, cloud, colocation and AI cluster deployments
- Configuration recorded per site, so a second hall or a replacement unit matches the first
- Optional redundant power supply and a second reference input
- CE and RoHS compliant, supplied with a test report and a 3 year warranty
Technical Specifications
| Specification | Detail |
|---|
| Part Number | APX-SYNC300 |
| Product Type | Data centre time synchronization system |
| GNSS Reference | GPS and BeiDou timing configurations |
| Network Time | NTP and SNTP for servers, storage and management systems |
| Packet Time | PTP over Ethernet for equipment that needs time and phase on the wire |
| Physical Outputs | 1PPS and 10MHz outputs for instruments and test equipment |
| Holdover | Internal oscillator disciplined to the satellite reference, holdover maintained when the reference is lost |
| Reference Inputs | GNSS antenna input, external reference input for a second source |
| Network Interfaces | Multiple Gigabit Ethernet ports |
| Redundancy | Reference priority with automatic switching, redundant power option |
| Management | Web interface, SNMP and alarm reporting |
| Antenna Set | GNSS antenna, low loss cable and surge protection supplied as one matched set |
| Configuration | Interfaces and profiles set for the site before shipment, configuration recorded |
| Application | Enterprise and cloud data centres, colocation halls, AI and HPC clusters |
| Mounting | Rack mount chassis |
| Compliance | CE, RoHS |
Applications
Enterprise and Cloud Data Centres
Gives application, database and security logs one ordered timeline, which is what incident review, audit and capacity analysis actually depend on.
AI and High Performance Computing Clusters
Keeps the nodes of a distributed job inside a common time base, so job events, scheduler traces and accelerator telemetry can be correlated across the whole cluster.
Colocation Halls and Secondary Sites
Uses the same reference design in every hall, including sites with no satellite view of their own, by taking a second reference input from the site that has one.
Deployment and Selection Notes
Start with an inventory instead of a product number. Count how many devices need time, split that list into equipment that will take NTP and equipment that needs PTP, and note any instrument that expects a physical reference. The APX-SYNC300 is configured against that inventory, and the same form is then used for the next hall.
Plan the antenna and the fallback before installation. A rooftop or high level mounting point with a short run to the entry panel keeps the signal healthy, and where two halls are connected, the second unit should be set to take an external reference input rather than depend on its own antenna alone. Reference priority is configured so the intended order of fallback is the order that is actually used.
For wider installations the timing range covers the other roles: GNSS and NTP network time servers for smaller sites, PTP grandmaster clocks where phase must be generated for the transport path, and synchronization systems where a multi site deployment has to be planned and configured as one project.
Compatibility
The system serves standard NTP and SNTP clients and standard PTP equipment over Ethernet, plus discrete 1PPS and 10MHz consumers. Tell us the operating systems, the network equipment and the number of clients, and we confirm the interfaces, the polling capacity and the antenna set in one document so the site can be checked before the order is placed.
Related Products
Frequently Asked Questions
Is NTP accurate enough when the applications only need ordering?
For log ordering and transaction sequence, network time served from a local traceable source is normally sufficient, and it is far more consistent than each machine polling an internet server. Equipment that needs tighter time and phase is served by PTP from the same unit.
Can one unit serve a whole hall?
That depends on the client count and the network design. Send us the device count and the switch layout and we confirm the interfaces and capacity in writing, rather than leaving you to estimate it.
What if the building has no usable satellite view?
The unit accepts a second reference input, so a hall without a satellite view can be referenced from the site that has one. Reference priority is configured so that fallback order is deliberate.
Is a rooftop antenna always required?
The antenna has to see the satellites, which in practice means roof level or higher. Site surveys usually settle it quickly, and the antenna, cable and surge protection are supplied as one matched set.
Why Choose Apex
- One traceable clock for the whole estate instead of per machine adjustment against public servers
- Antenna, cable and surge protection quoted as a matched set rather than as separate line items
- Client count and interface list confirmed with you before the order is released
- Reference source, lock and holdover state reported to your monitoring, so the chain is visible
- Backed by a 3 year warranty and technical support on synchronization planning and topology
