New cl26 DDR5 RAM for Ryzen coming in January.

[WARburton]

Nope.

Not enough to warrant the cost. I mean, if you haven't bought any RAM yet, then sure, knock yourself out, but I wouldn't replace existing RAM. Going from 6000 CL36 to 6000 CL26 is fairly worthwhile (1-2%), but it very much depends on the cost. Again, you won't notice a difference, but people will pay for a small percentage increase even if they can't tell.

thanks for the reply guys, I'll not bother

 

[Capitano]

Im sure there are testing videos out (or about to be out) that you can compare and see if its worthwhile

 

[monkethicc]

You can get gold standard RAM if you know where to look, you don’t need a manufacturer to bin a kit for you and charge extra.

I dont mean gold standard i mean the physically gold trident z royals

 

[zaphod-159]

Any news on when/where the CL26 kits might be available?

 

[Jay85]

My review sample arrived today to go in a build mod, unfortunately can't test this yet until the mod is complete!

 

[Jay85]

My usual quick photos and an unboxing shortly!

 

[eeii]

My usual quick photos and an unboxing shortly!

Are these likely to need a special motherboard to get working ?

 

[Jay85]

Are these likely to need a special motherboard to get working ?

I will test, I have a few X870 and X870E boards plus ann X870 itx board. Will compare with the B680 boards too when they come through to me.

I think minimum will be a Ryzen 9000 series processor though.

 

[Jay85]

 

[GoRedwings19]

I was under the general impression due to the AMD platform having weaker IMC and higher latency compared to intels in general it is not worth it.

It seems it would be more worthwhile to see if you can run your existing memory at slightly higher mem speed and tighten up the secondary timings to bring the latency down a bit. Rather than go out and buy an expensive mem kit.

Am I wrong?

 

[ATIorNvidia]

I will test, I have a few X870 and X870E boards plus ann X870 itx board. Will compare with the B680 boards too when they come through to me.

I think minimum will be a Ryzen 9000 series processor though.

Primary timings are 100% down to the sticks so 7000 series will be fine provided it can do 6000 which basically all of them do.

 

[SpellowHouse]

I was under the general impression due to the AMD platform having weaker IMC and higher latency compared to intels in general it is not worth it.

It seems it would be more worthwhile to see if you can run your existing memory at slightly higher mem speed and tighten up the secondary timings to bring the latency down a bit. Rather than go out and buy an expensive mem kit.

Am I wrong?

You are not wrong, BUT it's an out-of-the-packet solution. And many people don't care about an extra £100 on top of the cost of a new system.

 

[pewpew]

You are not wrong, BUT it's an out-of-the-packet solution. And many people don't care about an extra £100 on top of the cost of a new system.

surely not £100 premium on top of existing kits? do we have a rough price estimate?

 

[Jay85]

surely not £100 premium on top of existing kits? do we have a rough price estimate?

Don't have a price for these but i can ask my contact.

 

[Capitano]

any benchmarks out yet?

 

[jp7152]

The funny thing is that you'll probably get more benefit from tuning the subtimings than getting a kit with slightly faster primary timings

 

[JollyJamma]

The funny thing is that you'll probably get more benefit from tuning the subtimings than getting a kit with slightly faster primary timings

Maybe but even if a CL26 kit was faster, if you purchased a good quality CL30 kit and overclocked it, it might be close to or as fast as a CL26 kit (it might even be faster) but it would definitly be much cheaper.

A CL26 kit is going to be much more expensive than a CL30 kit.

 

[Mike25]

This video compares 6000/6400/8000 MT/s.

This guys bar charts are bad - cut off because there's so little difference and has to make it look like a score of 15,859 is half of 16,635 for example.
Also saying that a 64GB kit is the way to go because the 2x32GB 6000MT CL 28 2200FCLK is faster than 2x16GB 6000MT CL 30 2000FCLK.
Goes on to say "it's not a pure 1:1 example but, I wouldn't expect there to be that big of a difference, who knows."
Why not test it then as you have the kits utterly bizarre.
No mention of if the 6400MHz kit is running in 1:1 or 2:1 mode, or any other settings.

This CL26 kit probably won't make much difference if any, would be interesting if it did though.

 

[megatron]

any benchmarks out yet?

 

[Capitano]

thanks! looks like not too much of a difference then. Not the best charts either but it'll have to do for now