Do both SCCM and SCOM provide monitoring features? What differences between?

SCCM stands for System Center Configuration Manager whereas SCOM stands from System Center Operation Manager.
SCCM is used application in scenario where you have thousand of servers or desktop machine. Consider you need to deploy patch or deploy a script or install software on multiple servers, so instead of going to each and every machine you can just create a package and deploy the installation from single console.
there are also many reporting features which will help you to know on how many servers it has been installed and how many have failed and etc.
Whereas SCOM is monitoring tool say you have thousand servers and you want get alert for C Drive when space remaining on it is less than 80% then you can set it through scom only once after that you will receive alerts/mail as configured for less disk space.
You can also set alert for production servers, so if you do not get ping response from server you will receive an alert in scom console.


Hope it clears your doubt !! :)

Does SCCM provide a monitoring feature? Otherwise how come MS website says it can do "Client health and monitoring"? I'm kind eager to know what kind of monitoring feature it provides.

Besides SCOM, do you recommend any other server monitoring tools?

There are various monitoring tools depending upon cost, such as HPOV and you can go for open source tools such as Nagios which is one of the good option for windows server as well as Unix servers

How about SolarWinds SAM?

We are kind of looking for a monitoring tool to detect the deteriorating hardware issue so we can take proactive action in advance. Can you recommend some?

If you are looking for free tool without any cost then you can go for Nagios as it is open source, and it will help you to monitor hardware issues. If tech guy has good knowledge further more improvements also can be done using customized scripts.
You can check with other tools but they have licensing cost and you need to compare each and every feature to get for what you paid.
There are list of tools, SCOM, Solarwinds, HPOV, etc.. list goes on. :)

Sorry for the late response but do you know if any of the above mentioned tools can proactively detect hardware stress, such as a port being dying on a network switch or a failing physical hard drive on a RAID. The point is, we want to know the coming problem/failure in advance instead of sitting there waiting the alert to tell you the hardware is "already" dead. Rather, we want to get a alert telling you the hardware "is dying" so you can fix the issue before it totally dies.

Define a dying port? How would you diagnose it using a script?
A failing disk will  have disk events in the system event log. Your raid vendor may also have management packs that integrate with SCOM