fiber mallSFP 10G SR for HPE ProCurve J9150A HPE Aruba J9150D J9150DCM JL748A 10GBASE-SR SFP+ 850nm 300m MMF DOM Transceiver Module

I use it for netgear RX700 really good and give me fullspeed

This has likely caused problems in my network when getting downstream connection from a 10GbE device to 5GbE which significantly slowed down the connection to even less than 200mbps in the 10GbE device. I have to force the 10GbE device to operate at 5GbE to avoid this drastic slowdown to have around 3.8 GbE afterwards.

I have previously purchased and used 10GBase-T SFP+ 30m (Marvell 88X3310) modules from two other vendors (Wiitek & MikroTik), as well as a 10GBase-T SFP+ 80m (Broadcom BCM84891) module from one other vendor (10Gtek). These modules have been used in MikroTik hardware; specifically the MikroTik CCR2004-1G-12S+2XS ("CCR2004") and the MikroTik CRS317-1G-16S+RM ("CRS317").I purchased two of these Fiber Mall 10GBase-T SFP+ 80m (Broadcom BCM84891) modules in the hopes of getting a Broadcom BCM84891-based module with working temperature/voltage reporting, as that is something that is missing from both the Wiitek Marvell 88X3310-based and 10GTek Broadcom BCM84891-based modules.While the MikroTik Marvell 88X3310-based modules do have working temperature/voltage reporting, they are unfortunately extremely hot and power hungry (far more so than every other Marvell 88X3310-based module out there — which is particularly annoying given that the MikroTik Marvell 88X3310-based module is the only SFP+ 10GBase-T module with fully working NBASE-T support!).Anyways, among all of these modules, these Fiber Mall Broadcom BCM84891-based modules are easily & by far my favorites so far! Why?#1: They have working temperature/voltage reporting! This is a rarity on Marvell 88X3310-based modules, and the only other Broadcom BCM84891-based module I have been able to test so far also lacked working temperature/voltage reporting, so this is a very big advantage!#2: They have had no trouble whatsoever with reliably achieving a 10GBase-T link on my longest run of cat5e (which cannot be said of any of the Marvell 88X3310-based modules, which struggle to manage a 5GBase-T link at the absolute most, and will typically settle at only 1000Base-T on that run).(Please note that I have been able to successfully push 10 Gbps over a cat5e run approximately ~3x this distance by going from my port 1 on my Intel X550-AT2 back to port 2 on my Intel X550-AT2, so the issues with achieving full speed links on this run with Marvell 88X3310-based modules are presumably due primarily to a combination of power limitations coupled with the inferior power efficiency of the Marvell 88X3310-based modules.)#3: Broadcom BCM84891-based modules are dramatically superior to Marvell 88X3310-based modules in certain applications.Yes, all of the Broadcom BCM84891-based modules on the market are priced at a significant (35–100%) price premium over the Marvell 88X3310-based modules, however these particular Broadcom BCM84891-based modules are one of the cheapest Broadcom BCM84891-based modules available.Furthermore, the Broadcom BCM84891 is well worth the price premium over the Marvell 88X3310 in situations involving:• Achieving 10G links on ≥30m runs of cat6a cabling• Achieving 10G links on runs of cat5e/cat6 cabling that are too long for Marvell 88X3310-based modules to reliably achieve 10G link speeds over. Note that in this case "too long" is often MUCH lower than 30m — for example, my use case is a link that should be somewhere in the ~15–20m range, and I've seen many other users online reporting similar issues (and similar resolutions!) at a very wide range of different link lengths.• Situations where module power consumption is a significant consideration. Note that Broadcom BCM84891-based modules are generally specified for a power draw of ~1.6–2.0 W for ≤30m links (or 2.0–2.5 W for ≤80m), while Marvell 88X3310-based modules are generally specified for a power draw of ~2.4–3.4 W for ≤30m links — this can be a rather dramatic (~17–41%) difference in power consumption for ≤30m links!• Situations where module temperature is a significant consideration• Situations where the increased fan noise resulting from the elevated module temperatures & power consumption would be undesirable, or where supplementary cooling would otherwise be required (e.g. with fully-passively-cooled hardware such as a MikroTik CRS305 or CRS309).One example of a situation like the last example is my use case for my CRS317. I have it located in my bedroom/office, and so having it frequently kick its (normally quietly idling) fans up to max speed (which is very loud!) for an extended period of time to try to cool down the overheating Marvell 88X3310-based MikroTik modules operating at 10G is obviously a significant annoyance.I even had to relocate the position of the MikroTik modules to ensure they were placed directly in front of the rear fans in order to increase airflow to them. While this significantly reduced how often & how long the fan would spin up to max speed, it unfortunately did not do so by anywhere near enough to make it tolerable.Yet the Fiber Mall Broadcom BCM84891-based module that is acting as the switch's uplink port (which is also in the absolute worst SFP+ port on the switch from an airflow point of view!) is running FAR cooler than the MikroTik module that sitting directly in the path of the fans (and therefore receiving FAR more airflow from them), even though it is operating at the exact same link speed, and is (at least currently) passing the exact same amount of traffic through it!Of course, if you're working with cat6a links under 30m with actively cooled host hardware located in an area where noise is of no concern (e.g. datacenter, MDF, IDF), and have sufficient port group power budgets, then the cheaper Marvell 88X3310-based modules are obviously going to be a better fit for you.#4: The price is right — and it's available right here on Amazon.com with fast and free Prime shipping!The only possible downside I can think of is that the multi-packs sold on Amazon currently aren't Prime-eligible and have extremely long shipping windows, however I suspect part of the reason for the lengthy shipping window is pandemic-related shipping delays, and it's possible that the seller may begin to stock those (and/or other multi-pack options) in US warehouses down the road. Regardless, this minor downside is not something that I care much about, but I thought it was still worth mentioning.Overall, I am EXTREMELY satisfied with these modules, and plan to purchase more of them in the near future!

Got the intel flavor first, my es16xg was fine with it, but not my er8xg. Got the Cisco flavor instead, both router and switch are happy with it. Cherry on the cake, the sfp does report some dom/ddm info (like sfp temp).Also, do yourself a favor and spend a bit more dough to buy the 80m flavor (uses a Broadcom chip). The 30m version of this sfp uses a Marvell chip, it’s finicky, and it runs bloody hot (the same goes for all other 30m 10baseT sfp, not just this brand).

This 80 m transceiver is not recognized by the Intel on-board SFP+ ports on my Netgate 6100 firewall. The transceiver is recognized by a Mellanox ConnectX-4 card. When plugging the other end into a regular 10GBase-T RJ45 port on a Netgear switch, the link is up but data doesn't work. When plugging the other end into another 10 G transceiver in the SFP+ port of the same Netgear switch, the connection works.

I tested these with a Mikrotik CRS305-1G-4S+IN connected to a QNAP TS-932PX NAS via AmazonBasics OM3 cable - hot plugged it and it went straight to work. No problems at all, and will probably buy more for similar applications.