Download Download NVIDIA MBF2M516A-CECO DPU Firmware 24.35.1012 for OS Independent Free

Firmware is added at the time of manufacture to run user programs on the device, which can be thought of as the software that allows the hardware to run. Embedded firmware is used to control the functions of various hardware devices and systems in the same way that a computer’s operating system (OS) controls the functions of software applications. Firmware can be written in read-only memory (ROM), erasable programmable read-only memory (EPROM), or flash memory.

NVIDIA BlueField-2 DPUs deliver innovative acceleration, security, and efficiency to every host. BlueField-2 data center infrastructure combines the power of NVIDIA ConnectX-6 Dx with programmable Arm cores and hardware offload for software-defined storage, networking, security and management workloads.

NVIDIA BlueField-2 also delivers superior performance, security, and lower total cost of ownership for cloud computing platforms, enabling organizations to efficiently build and operate virtualized, containerized, and bare-metal infrastructure at scale.

Changes and new features:

– HPCC related configuration is now supported through the mlxconfig utility.
– Added support for copy modify header control operations for UDP fields.
– Added the following resource dump segment: SEG_HW_STE_FULL, which includes dumping to STE and all its dependencies.
– Added the following resource dump segment: SEG_FW_STE_FULL, which includes dumping to FW_STE and HW_STE_FULL in scope
– Since software needs additional space to handle strided RQs before and after packet scatter, the hardware will allocate the required space for spare copies while scatter packets.
– Improved connections per second (CPS) rate for secure offload using Generic Objects DEK (PSP TLS, etc.).
– Increased VirtIO hardware offload message rate to 20/20 MPPS for 256 virtual devices by optimizing datapath application code.
– Open SNAPI steering hop optimization: only dmac/dlid 0 packets will enter the SNAPI steering channel.This means that if an Open SNAPI communication channel exists, all other traffic will only have a 1 turn hop penalty
– Open SNAPI steering jump optimization: SW must use dlid/dmac == 0 in address vector of SNAPI QP
– Open SNAPI steering jump optimization: across eSwitch, if LAG is enabled, now allows open SNAPI
– Added support for SHA calculation via MMO engine.
– Optimized latency for UPT traffic.
– Added LAG support on NODNIC interfaces to enable traffic on other ports when LAG is enabled.
– QoS priority trust default state can now be changed using new nvconfig: QOS_TRUST_STATE_P1 or QOS_TRUST_STATE_P2. The values ​​that can be used to set the default state are: TRUST_PORT, TRUST_PCP, TRUST_DSCP, TRUST_DSCP_PCP
– Implemented a transition device that supports spec 1.x compliant drivers and allows legacy drivers to support virtio legacy drivers (virtio 0.95).

Bug fixes:

– Fixed an issue that caused RDE (Redfish) PATCH operations on LLDPTransmit properties “ManagementAddressIPv4”, “ManagementAddressIPv6”, and “ManagementAddressMAC” to be applied only on the first attempt and fail on the next.
– Fixed an issue that caused the MLNX_OFED driver to fail to send commands to the NIC when the VirtIO driver was loaded.
– In strongSwann or OVS IPsec solutions, BlueField’s support for IPSec acceleration (i.e. transparent IPSec full offload) of the host data path stops passing more than about 6.2*10e9 packets over an IPSec session at an Rx or Tx MTU of 1500 bytes host data packets. Whether to use the ESN flag.
– Added synchronous reset support in case RShim PF is exposed but not connected to RShim PCIe driver.
– Fixed Bluefield-2’s PCI gateway configuration to enable MCTP/PCI.
– Fixed an issue that caused the PLDM AEN event sink medium to change unexpectedly and the target BDF to be overwritten with garbage when certain PLDM packets were received from the SMBus layer.
– Modified compression GGA behavior to prevent QP from being moved to “error” state when the compressed size is larger than the destination buffer.
– CPU handling synchronization needs to be separated (run ptp4l with taskset-c [cpu #] prefix) during heavy traffic.

About network adapter drivers:

Once connected, the operating system usually installs generic drivers that help the computer recognize the newly connected device.

However, if you want to take advantage of all the features your network adapter has to offer, you must apply the appropriate software. This task also allows the computer to correctly identify all device characteristics such as manufacturer, chipset, technology, etc.

Updating the adapter’s driver and utility versions may improve overall performance and stability, increase transfer speeds, fix various compatibility issues and some network-related bugs, and bring various other changes.

To install this release, simply grab the package, unpack it if necessary, run the installer, and follow the on-screen instructions. Don’t forget to reboot your system and reconnect your network adapters when you’re done to make sure all changes take effect properly.

Without further ado, if you plan to use this release, click the download button and install the package. Also, please visit our website as often as possible so you don’t miss a single new release.