FAQ

​

Technical Support & Procurement FAQ Overview

  • •

    Design-in Support: Providing factory datasheets, schematic reviews, and engineering test sample requests.

  • •

    Comprehensive Product Answers: Covering technical specs for High-Speed Interconnects, PoE, Networking, and Industry 4.0 applications.

  • •

    Procurement & Alternative Evaluation: Offering Minimum Order Quantity (MOQ) details, lead time verification, and EOL replacement solutions.

  • •

    Can't find thru this FAQ ? : please feel free to send us an e-mail to us, then we will try to feedback you shortly. 

  • Q

    What is the insertion loss and return loss budget of these shuffles?

    A
    Both 4x4 MPO08 and 8x8 MPO16 shuffles achieve a maximum end-to-end Insertion Loss (IL) of ≤ 0.35 dB (with typical connector loss around 0.18–0.20 dB using Elite low-loss ferrules). Return Loss (RL) is guaranteed ≥ 60 dB due to 8-degree Angled Physical Contact (APC) polishing.
  • Q

    What is an Optical Shuffle in an AI GPU tray?

    A
    An Optical Shuffle is a pre-terminated, passive fiber optic cross-connect matrix installed inside a GPU server tray. It maps fiber optic signals from multiple host transceiver modules (e.g., OSFP 800G/1.6T) directly across multiple independent network switch planes (rails) without manual cross-patching or loose fiber cable clutter.
  • Q

    How do Loopbacks assist with switch power budget and thermal validation?

    A

    Next-generation 800G and 1.6T optical transceivers draw substantial power—often 20W to 30W+ per port.

    • Passive loopbacks supports uptot 45W+ power spots with programmable function (from 0W- 45W) that can test maximum system loads.

    • DSP Active Loopbacks combine the actual power consumption of the DSP chip (13W ~ 19W max.) with programmable power spots (23W max.). This allows data center engineers to stress-test fan speeds, power supply limits, and chassis airflow under simulated 100% full-power transceiver loads across all faceplate ports simultaneously. 

  • Q

    Why is a DSP-based Active Loopback mandatory for 112G and 224G PAM4 link bring-up?

    A

    At 1.6T (8x200G), high-frequency attenuation over switch PCB traces causes severe eye closure before the signal even reaches the port faceplate.

    • A passive loopback doubles the channel distance (host-to-port and port-to-host), causing total link failure due to combined insertion loss. and also with programmalbe power for port validation 

    • A DSP active loopback acts as a retimer at the port edge. It cleans up distorted signals coming from the host switch ASIC, calculates link Bit Error Rates (BER) via its internal DSP engine, and sends a regenerated, clean signal back into the host. This enables engineers to isolate whether link issues stem from the host ASIC/board design or the optical transceiver.

  • Q

    What is the fundamental difference between a Passive Loopback and a DSP-Based Active Loopback?

    A
    ​Passive Loopbacl Module: 
    • Architecture: Utilizes direct copper trace loops inside the module shell without active silicon.

    • Function: Simply routes the electrical output from the host TX directly back into the host RX.

    • Primary Use Case: Port trace continuity verification, basic low-power I2C/EEPROM handshake validation, and baseline link detection where host channel loss is negligible.
    • Programmalbe Power consumption: Upto45W+ 


    DSP-Based Active Loopback Module (ALB): 

    • Architecture: Houses an onboard PAM4 DSP chip (major from Marvell / Broadcom) along with programmable microcontrollers.

    • Function: Actively terminates the host 112G/224G SerDes signals, performs Clock and Data Recovery (CDR), RX Equalization (CTLE/FFE/DFE), and TX Pre-Emphasis, re-transmitting clean PAM4 signals back to the host.

    • Primary Use Case: Full high-loss host channel validation (overcoming 10dB–30dB+ PCB trace loss), real-world thermal/power profiling (simulating 20W–40W+ optics), and complex CMIS 5.x runtime state machine negotiation
  • Q

    What is a Module Compliance Board (MCB), and how is it used?

    A

    A Module Compliance Board (MCB) acts as the test fixture for optical transceivers, Active Optical Cables (AOCs), or Direct Attach Copper (DAC) modules. The optical or copper module inserts directly into the MCB’s cage.

    • Primary Function: Validates the module's electrical interface, power rail noise immunity, and host-to-module signal paths.

    • Management & Control: Equipped with control interfaces to read, write, and verify CMIS (Common Management Interface Specification) registers, power consumption, and I2C/MDIO communication.

  • Q

    What is a Host Compliance Board (HCB), and what does it test?

    A

    A Host Compliance Board (HCB) is a precision test fixture designed to plug directly into host switch or line card ASIC ports (e.g., OSFP/QSFP-DD slots). It breaks out high-speed TX and RX electrical channels into standard coaxial RF connectors.

    • Primary Function: Isolates and evaluates the host switch port's transmitter output, receiver jitter tolerance, and signal integrity without adding host-side cable distortion.

    • Key Feature: Engineered with ultra-low insertion loss PCB materials and strict trace-length matching to adhere to IEEE and OIF compliance specs.

      A Host Compliance Board (HCB) is a precision test fixture designed to plug directly into host switch or line card ASIC ports (e.g., OSFP/QSFP-DD slots). It breaks out high-speed TX and RX electrical channels into standard coaxial RF connectors.

    • Primary Function: Isolates and evaluates the host switch port's transmitter output, receiver jitter tolerance, and signal integrity without adding host-side cable distortion.

    • Key Feature: Engineered with ultra-low insertion loss PCB materials and strict trace-length matching to adhere to IEEE and OIF compliance specs.

  • Q

    If an existing part number faces End-of-Life (EOL), can MITS Component & System Corp. assist in finding alternative solutions (Drop-in / Pin-to-Pin replacements)?

    A

    Yes. The experienced MITS FAE team can assist in evaluating discontinued part numbers to find Pin-to-Pin or spec-compatible alternative solutions, while providing test samples and technical comparison support. Feel free to bring your legacy part specifications and consult MITS for alternative solutions.

  • Q

    How can I confirm the Minimum Order Quantity (MOQ) and lead time for products?

    A

    Standard Packaging Quantities (SPQ), Minimum Order Quantities (MOQ), and production lead times vary by manufacturer and part number. You can first browse our Products Overview page to identify the required part numbers and submit your estimated order quantities. Our sales team will promptly check stock availability and factory lead times for you.

  • Q

    What is the process and timeline for R&D teams to request engineering samples for testing?

    A

    You can submit a request via our online contact form or by reaching out directly to your sales representative. Once our FAE and sales teams verify your project specifications and estimated mass production schedule, we will arrange factory sample shipment as quickly as possible.