These are some confusing feature sets. The 5016 seems like it's the higher end member, but with only a x4 (or split x2/x2) PCI-E v5 interface, it's the slower of the two as the 4016 offers x8. But in all other ways the 4016 is the lesser of the two.
I think the enterprise cares more about density than throughtput especially after applying some redundancy calculation. PCIE x8 is only theoretically supported by EDSFF connector ifrc. Using half the lanes means it can also be put in the same chassis with ordinary U.2 connectors.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
3 Comments
Back to Article
mariushm - Thursday, August 8, 2024 - link
May want to fix the title as soon as possible.dwillmore - Friday, August 9, 2024 - link
These are some confusing feature sets. The 5016 seems like it's the higher end member, but with only a x4 (or split x2/x2) PCI-E v5 interface, it's the slower of the two as the 4016 offers x8. But in all other ways the 4016 is the lesser of the two.erinadreno - Monday, August 12, 2024 - link
I think the enterprise cares more about density than throughtput especially after applying some redundancy calculation. PCIE x8 is only theoretically supported by EDSFF connector ifrc. Using half the lanes means it can also be put in the same chassis with ordinary U.2 connectors.