Ryzen 2400

[pomkon]

They all say Ryzen heavily benefits from faster ram, those tests are done from gaming comparisons. What about non-gaming tasks, do they have a practical use?

If not, can you get away with 2400 on R7?

 

rephrasing, what % increase in rendering, virtualization, editing, etc...?

 

there are just too many ram to choose from, some that wouldnt oc, dif latency, too much reading ewquired

[just_dave]

Well, if your mobo can do more (2933) there is no reason to stay at 2400. 2933+ will give you huge performance bump, i think even per core performance because the CCXs can communicate faster.

[pomkon]

Just now, dave_k said:

Well, if your mobo can do more (2933) there is no reasons to stay at 2400. 2933+ will give you huge performance bump, i think even per core performance because the CCXs can communicate faster.

ccxs??

[Guest]

It was previously difficult to get RAM to run at rated speeds. It is not the case anymore. Grab a kit of Corsair Vengeance LPX. The DOCP (XMP) profile should work right out of the gate. Though I do recommend watching a video or two on YT and set the settings yourself for the best stability. At the end of the day, if you just don't want to be bothered, just leave it at 2400. It is technically fine. Just would be a shame not to take a half hour and get the best out of what you paid for.

 

Edit: I'm merely suggesting LPX because I've used it half a dozen times with no issues. There are, as you've said, tons of kits that'll do the job.

[Nocte]

2 minutes ago, pomkon said:

ccxs??

A CCX is a Core Complex which contains Ryzen cores. There are several CCXs in a Ryzen CPU (number of activated CCXs depends on the model), and they communicate through what AMD calls "Infinity Fabric". Infinity Fabric communication performance is linked to memory (RAM) speeds.

[just_dave]

11 minutes ago, Eibe said:

A CCX is a Core Complex which contains Ryzen cores. There are several CCXs in a Ryzen CPU (number of activated CCXs depends on the model), and they communicate through what AMD calls "Infinity Fabric". Infinity Fabric communication performance is linked to memory (RAM) speeds.

Whut? Your explanation is somehow wrong

@jdwii There are two CCXs in a desktop Ryzen die (4 in Threadripper which is made of two desktop dies, 8 in epyc because it contains 4 desktop dies), CCX contains 4 cores and 8mb of L3 Cache. The CCXs have interconnect together and with SoC and shit like that using Infinity Fabric which is very scalable and runs from RAM speed, so the higher speed you have, the faster the interconnect > the faster the cache acces and communication between the CCXs

AMD disables cores symmetrically, so fully enabled and binned Ryzen has 8 cores, with one core in each CCX disabled you get 6 core and with two in each CCX you get 4 core.

[Nocte]

1 minute ago, dave_k said:

Whut? Your explanation is somehow wrong

@jdwii There are two CCXs in a desktop Ryzen die (4 in Threadripper which is made of two desktop dies, 8 in epyc because it contains 4 desktop dies), CCX contains 4 cores and 8mb of L3 Cache. The CCXs have interconnect together and with SoC and shit like that using Infinity Fabric which is very scalable and runs from RAM speed, so the higher speed you have, the faster the interconnect > the faster the cache acces and communication between the CCXs

Yeah you are correct. I got confused with the cores inside the CCXs being deactivated, not the CCXs themselves my bad.