AMD's Ryzen Will Really Like A Newer Linux Kernel

[Michael]

Thanks guys for the support and some of you pressuring AMD on Twitter and other forums... Just heard from AMD from some new person I should be receiving hardware now.... Will keep you guys posted. For those that did tip for the cause, I will still be using the funds for Ryzen-related hardware whether it be any Ryzen 3/5 hardware that I can't get, related hardware expenses for those systems, etc.

[pheldens]

Nice , my phenom x4 is on its last legs too

[dungeon]

I'd rather see MSI would spend that routing, parts and software dev money for better quality boards (parts and manufacturing/soldering) instead of a rotary switch ...

The only thing this switch does is setting the multiplicator in hardware instead of the bios. I saves you 10s when testing new oc settings and has a risk attached to it.

Well, there are always risk of crashing with overclock rotary or manually bios or... even for AMD it crashed live when they demoed XFR and OC (at about 19:50 minute of video)

[pipe13]

Nice , my phenom x4 is on its last legs too

I get a grin out of you "Never say die!!!" AMD chuckle heads.

[vortex]

Thanks guys for the support and some of you pressuring AMD on Twitter and other forums... Just heard from AMD from some new person I should be receiving hardware now.... Will keep you guys posted. For those that did tip for the cause, I will still be using the funds for Ryzen-related hardware whether it be any Ryzen 3/5 hardware that I can't get, related hardware expenses for those systems, etc.

Why no new news post stating this?

[drSeehas]

It does not work like that, if the interface mandates that the slot MUST provide X, Y and Z interfaces, ...

The interface of the PCIe SSD mandates only PCIe lanes and no SATA port at all.

I don't have experience with M.2 slots on mainboards yet. I am using PCIe to M.2 adapters without any SATA capability. So if a M.2 PCIe SSD mandates a SATA interface too, it wouldn't work in my case.

... the slot MUST provide them regardless of what is actually used.

So the shared SATA ports are for nothing, as they are always occupied by the M.2 slot?

... disable the pcie lines if I connect a Sata-only drive in a Sata-Express slot.

I am talking/writing about a legacy SATA port, not SATA Express.

[L_A_G]

Nice , my phenom x4 is on its last legs too

Same applies to my i7 950 from around the same time. Not sure if it's just failing components a bug in the BIOS exposed by using a PCIe SSD (something rather uncommon in 2009), but the thing literally cannot reboot the normal way and instead requires me to shut it down fully and then booting it again (BIOS reflash or putting the SSD in another PCIe slot does nothing). My overclocks (ratcheting up the CPU to 3.52 GHz and the RAM to 1600 MHz) may also have had something to do with it.

[NotMine999]

Phoronix: AMD's Ryzen Will Really Like A Newer Linux Kernel

AMD's Ryzen CPU is finally shipping in a few days! If you are planning to be an early adopter of AMD Ryzen processors, you will really want to be running a newer Linux kernel release for proper support and performance...

http://www.phoronix.com/scan.php?pag...n-Newer-Kernel

Don't worry. There will some some n00bs out there somewhere that will try to load an old Debian or Ubuntu release on a board using this CPU (because that's the only CD or DVD that they can get their hands on! REALLY?) and then go whining across the online forums about how much Linux outright sucks and Windoze is worlds better.

The newer Linux kernels have code or simply code tweaks that make new hardware work under Linux. Some Linux drivers will require PCI "id" updates (yep, lots of driver code actually contains PCI ids so it can recognize the appropriate hardware inside a PC) to recognize the new hardware and apply any appropriate tweaks. For an example of different tweaks driven by different hardware types, all you need to do is read the Linux source code for the Realtek r8169 network driver. Of course there is always the chance that a completely new driver might be needed for something.

As for motherboard manufacturers adding new chips just to differentiate themselves from others, I expect that of the "fruit cult" and much less in the PC world. Ask yourself how many PC motherboard vendors use some variation of Realtek networking chips or ASmedia SATA and USB chips on their boards? Most of the PC motherboard vendors use the same chip vendors for networking, SATA, and USB. ASUS has some "special chips" (ITE 8771 sensor chip) for which they have Windoze-only drivers and Linux is generally ignorant of, but I don't think you are missing out on anything by not having them supported in Linux.

There is a wide choice of "sensors" chips, as in LM Sensors, for fan speed, voltage, and temperature monitoring. Laptops due that sort of stuff differently than standard PC motherboards. Sometimes an "unique" sensor chip appears, like the ITE 8771, that was made just for ASUS and no public specs exist. Still, most boards will use sensor chips that are already known to Linux via the LM sensors package, and that package gets periodic updates to add new devices and fix bugs in drivers.

I think we can all agree that support of new hardware in video drivers, especially for 2D and 3D support, will always be a "pain". IMHO that hardware contains some of the most secretive hardware and firmware in the computer industry.

Still, Michael's statement is very accurate. Use a newer Linux kernel release. I can only add, do your research thoroughly if you want to run Ryzen and Linux and "live on the bleeding edge" at the retail introduction of those parts. As for me, I can wait a while for Ryzen support to mature in Linux.

[cj.wijtmans]

Curious about that Game Boost Knob seems MSI boards have,

am i the only one who thinks this is not really a good idea? The MOBO should overclock automatically as possible within thermal limits...

[cj.wijtmans]

I think we can all agree that support of new hardware in video drivers, especially for 2D and 3D support, will always be a "pain". IMHO that hardware contains some of the most secretive hardware and firmware in the computer industry.

This is why non iGPUs drivers should be in user space...