# ConnectX-7: format\_select\_dw\_8\_6\_ext = 0 — blocks DOCA Flow IPv6 CT pipe

**URL:** <https://forums.developer.nvidia.com/t/connectx-7-format-select-dw-8-6-ext-0-blocks-doca-flow-ipv6-ct-pipe/370527>\
**Category:** Getting Started & Resources\
**Created:** [May 18, 2026, 4:12pm UTC](https://forums.developer.nvidia.com/t/connectx-7-format-select-dw-8-6-ext-0-blocks-doca-flow-ipv6-ct-pipe/370527 "2026-05-18T16:12:18Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![maxpain96](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/maxpain96/32/499091_2.png) [@maxpain96](https://forums.developer.nvidia.com/u/maxpain96)\
**Post date:** [May 18, 2026, 4:12pm UTC](https://forums.developer.nvidia.com/t/connectx-7-format-select-dw-8-6-ext-0-blocks-doca-flow-ipv6-ct-pipe/370527/1 "2026-05-18T16:12:18Z")

</div>

We are building a commercial IaaS+PaaS cloud SDN data plane using DOCA Flow  
on ConnectX-7. We need IPv6 Connection Tracking which requires the  
format\_select\_dw\_8\_6\_ext (full\_dw\_jumbo\_support) capability bit.

We have tested three different SKUs:

- MCX75310AAS-NEA (PSID MT\_0000000838, 400G NDR OSFP)
- MCX713106AS-VEA (PSID MT\_0000000840, 200G QSFP112)
- MCX755206AS-NEA (PSID MT\_0000000892, DGX storage)

All three: format\_select\_dw\_8\_6\_ext = 0 in BOTH GET\_CUR and GET\_MAX  
on firmware 28.48.1000 (latest).

Steps we have already verified:

1. No mlxconfig variable affects this bit (tested 24 NV-config changes  
with mlxfwreset, all confirmed via -e -a query showing Current column).
2. CR-space direct access returns 0xbadacce6 (blocked on standalone CX-7).
3. SET\_HCA\_CAP from userspace is filtered by kernel mlx5 driver (EINVAL).
4. All public CX-7 PSIDs share identical MAIN\_CODE section.

Questions:

1. Is there a roadmap to enable IPv6 CT pipe on standalone ConnectX-7?
2. Is there a documented mechanism (signed CS/CRCS/DBG token) to unlock  
this capability for evaluation/research?
3. The DOCA Flow CT documentation states “BlueField-3 and above required  
for IPv6”. Is this a permanent product positioning decision or  
firmware/HW limitation?

We are happy to share our CRCS challenge tokens with NVIDIA Engineering  
support team. We have an active deployment of 3+ CX-7-based SDN nodes  
and would prefer not to migrate to BlueField-3 if there is a path forward.

---

<div class="post-metadata">

**Author:** ![abirman](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/abirman/32/367020_2.png) [@abirman](https://forums.developer.nvidia.com/u/abirman)\
**Post date:** [June 4, 2026, 8:12am UTC](https://forums.developer.nvidia.com/t/connectx-7-format-select-dw-8-6-ext-0-blocks-doca-flow-ipv6-ct-pipe/370527/2 "2026-06-04T08:12:31Z")

</div>

Hello,

Thank you for your questions.

In order to answer them, opening a new support case will be required.

To Create a New Enterprise Support Case Please Use One of the Following Methods

1. Log in and Create a Case in the Enterprise Support Portal

> **[AppHub UI](https://nvid.nvidia.com/login)**

This is recommended and will provide the fastest response

2. Call your Regional Enterprise Support Number

Find your regional phone numbers:

[https://www.nvidia.com/en-eu/support/enterprise/#contact-us](https://www.nvidia.com/en-eu/support/enterprise/#contact-us)

3. Create a Case Without a Login

> **[EnterpriseSupport](https://enterprise-support.nvidia.com/s/create-case)**

Can be used to check warranty status

Recommended to reset login passwords

Requires validation and may increase time to respond

Best Regards,

Anatoly