Be careful when buying ram for AM5 systems, the most expensive ram is actually the worst, c26 6000 ram is junk, c40 ram wins in gaming

[Space Lynx]

@ir_cow it gets the lowest score here in this review... really shocked me... thoughts?

cl 40 ram the cheap shit actually beats it by a lot lmao wtf

G.Skill Trident Z5 NEO RGB DDR5-6000MT/s CL26 32GB Review - KitGuru

G.Skill's Trident Z5 Neo RGB DDR5-6000 memory has been designed for overclocking on AMD platforms wi

www.kitguru.net

 

[ir_cow]

Hmmm very strange to see "the worst" kits at the top. Considering everything else is within 1-2 FPS (or margin of error), my guess is those have some bad data or placed in the wrong chart. 15 FPS higher is not normal. Someone wasn't checking when they made the charts. - Im guilty of this too when I do average of 3 runs and I flip a number on one. Usually I catch it well before publication.

 

[cadaveca]

I'd say somebody's test rig isn't as stable as they think it is. assuming that's not easy numerical errors.

 

[Space Lynx]

interesting, thanks for the feedback @ir_cow strange indeed

 

[Onasi]

@ir_cow
Either that, or it’s just how Dawntrail do be. I myself jumped off FF14, but a friend who is still playing is telling me that their attempt to “upgrade” the engine resulted in absolutely shitshow things happening basically every patch since DT released - from broken new PBR materials to scuffed upsampling and other weird behaviors, so it would not surprise me at all. It’s a PS3 engine that worked iffily back then and time didn’t improve much.

 

[adilazimdegilx]

It says 'Intel Test Rig' on the pic. Probably a testing fail on their end. Or their 14700k gone really bad...

 

[ir_cow]

@ir_cow
Either that, or it’s just how Dawntrail do be

Some game be like that, but still some of that can be mitigated by doing multiple runs and removing the extreme outlier. Not the best practice, but if a game-engine is known to have random spikes (like failing to cull in or out) and you still want to include it in the benchmark suite, that is the only option really.

 

[Deleted member 245263]

This is just a very poor review. In the cons, they wrote that it only has the EXPO profile—they probably tried to set it up on an Intel platform and failed. On AMD, the kit delivers 'normal', more predictable results. They run some tests on one platform and others on a different one—if you click on AMD 'Office Productivity', you get Intel results because they only tested it there. It's simply a bad review. No meaningful conclusions can be drawn from it.

Im sorry, I'll skip my opinion on OP reading skills.

 

[ir_cow]

There is no comment section to let Simon Crisp know that there is a few errors.

This is why I always tell people to check out as many reviews as possible. Making decisions on only one opinion will burn you at some point.

 

[AsRock]

interesting, thanks for the feedback @ir_cow strange indeed

Try running your ram at cl40 and see if it's "better"

 

[chrcoluk]

At a guess motherboards are setting poor tertiary and other timings to get the lower CAS to work which ends up making it worse overall.

Although after what ir_cow told me that only the most premium boards can handle fast ram, I have decided to stop spending money on premium kits, to this day I have yet to get my samsung b-die ddr4 above 3200mhz stable lol. Although I am able to run it on tight timings and good refresh.

 

[ShrimpBrime]

At a guess motherboards are setting poor tertiary and other timings to get the lower CAS to work which ends up making it worse overall.

Although after what ir_cow told me that only the most premium boards can handle fast ram, I have decided to stop spending money on premium kits, to this day I have yet to get my samsung b-die ddr4 above 3200mhz stable lol. Although I am able to run it on tight timings and good refresh.

That's the problem, lack of data. You get CL the speed and the result.

The data looks bunk, probably bunk.

 

[phanbuey]

could be unstable rig, data collection error.. running benchmarks while windows search is indexing etc. But for sure CL40 is not faster than CL26, unless the motherboard blew out all the other timings set to auto for some reason.

 

[freeagent]

This just proves my point about toobers.

Anyone can run a bench, not hard at all, but not everyone can tune a system.

 

[Solaris17]

@ir_cow
Either that, or it’s just how Dawntrail do be. I myself jumped off FF14, but a friend who is still playing is telling me that their attempt to “upgrade” the engine resulted in absolutely shitshow things happening basically every patch since DT released - from broken new PBR materials to scuffed upsampling and other weird behaviors, so it would not surprise me at all. It’s a PS3 engine that worked iffily back then and time didn’t improve much.

XIV is the worst possible example to use as the OP did to talk about a product because XIV runs like garbage on my system since DT. I can push similar frames in outlaws which is saying something. Understanding of course, I am not saying I get low fps. Its just certainly disproportionate to anything else.

 

[Sunny and 75]

@OP
An outlier, for sure

 

[Dr. Dro]

It says 'Intel Test Rig' on the pic. Probably a testing fail on their end. Or their 14700k gone really bad...

Unlikely, it's just looking like the timings are too low and there are stability issues causing error correction to kick in.

Please, let's just bury the hatchet, 13/14th Gen CPUs don't die like flies and it's extremely unlikely (not to use the heavy word impossible) for these chips to degrade any further. If your CPU survived through those microcodes, and doesn't crash, it's fine and will remain fine forever.

 

[Space Lynx]

Try running your ram at cl40 and see if it's "better"

I'm just leaving mine on EXPO. These results just surprised me, even in games where its only 5 fps difference or so, this kit is not leading the pack, and you would expect it to do so. So I don't know, cl 26 6000 just isn't as impressive as I was thinking, while this game is an outlier true, the rest of the games in that review also don't paint an amazing picture by any means

 

[freeagent]

When looking at the min/max of those results, I am not sure what is going on.

The min/max is all over the spectrum, it literally makes no sense.

 

[Space Lynx]

When looking at the min/max of those results, I am not sure what is going on.

The min/max is all over the spectrum, it literally makes no sense.

One thing that is consistent in all of the games in this kitguru review are that min frames are better with the faster ram across the board. So, that's something I guess.

 

[ShrimpBrime]

When looking at the min/max of those results, I am not sure what is going on.

The min/max is all over the spectrum, it literally makes no sense.

Trident Z CL26 6000mhz at the bottom.

 

[iameatingjam]

Perhaps there's other factors in the ram, or elsewhere affecting this. Its hard to see 6000 cl40 beating 6000 cl30 if all else is equal... at least by that much.

 

[ShrimpBrime]

Perhaps there's other factors in the ram, or elsewhere affecting this. Its hard to see 6000 cl40 beating 6000 cl30 if all else is equal... at least by that much.

Looks like someone's elbow bumped the 4060 frequency slider 50mhz.
That's 15 FPS !! at 1080P over CL 26 at the same frequency.
I would love to see the entire timing sets of both kits and what settings for both kits top and bottom of that screenshot at OP.

 

[Toothless]

Don't know how it's not obvious that the benchmark is either borked or doesn't care about speed/timings.

oOoOoOo MoRe LaTeNcY iS gOoDeR no whoever did the chart is dumb.

 

[Braegnok]

They are using AMD EXPO 6000CL26 1.45v kit, and running it @ Intel DOCP on Z790 board with i7-14700K.

The AMD EXPO profile must not be training correctly for the clock cycles other then the primary timings on the Intel system, most likely lots of errors, correcting going on. AIDA64 Cache & Memory Benchmark results would show this as very bad scores, needing tuned, adjusted to run correctly on the Intel system.

The 1.40v kit has much better ICs vs the 1.45v kit and can be adjusted, tuned for maximum transfer rate and stability at fairly low voltage.

1.40v kit:



1.40v kit tuned profile:



TM5 90-min no errors:



Anyone who believes c26 ram is junk and c40 ram wins in gaming,.. I have some ocean front property in Colorado for sale.