Netac Z RGB DDR5-6200 CL32 2x 16 GB

[ir_cow]

Netac brings its new DDR5 Z RGB flagship memory to market. Equipped with a 6200 MT/s XMP profile, and "bedazzling" look, Netac is starting to shine brighter than ever! In this review, we will compare, overclock, and find out the magic within. Follow along as we test this memory and see how it stacks up to the competition!

Show full review

 

[Sabishii Hito]

Crazy results for M-die!

 

[Mussels]

Could you add ranks into the first page specs for the RAM?

Some entries in the charts also fail to mention capacity/ranks, while others do.

Sure you can figure it out from the product code, but it's still a minor oopsie



Two things stand out here

1. You deserve an award for getting RAM that fast on AM5

2. AMD really dont gain much from these speeds, people truly should focus on timings and not the higher clock rates
After a certain speed single rank REALLY hurts AM5, that much is clearly still a concern - interesting that JEDEC speeds don't show that difference


2x16 5600 CL40 - 298FPS
2x32 5600 CL40 - 314 FPS (+5.3%)
6200 CL32 - 325FPS (+9%)
Low latency 5600 in 2x32 would seem to be the sweet spot between all the tested kits, but also seems to be not tested yet. Probably easier to run on most setups, too.

Intel at DDR4 3200 had 319FPS and DDR 7000 340FPS -7.2% from a *HUGE* change in RAM speeds.

It does clearly show the intels get a higher average FPS, but that the platform that OC's ram the easiest, is barely worth it (in this title at least, honestly i picked the first one shown)

 

[ir_cow]

@Mussels thanks for pointing these issues out. I will work on fixing them / making it more coherent .

DDR5-6400 works for me too, but I don't add it because that is well past what is normal. Only works on X670E ACE and X670E Taichi. Probably not anymore with the SoC voltage lockdown. It needed 1.35V for that.

Update:
1)Benchmark graphs and tables are fixed. Added DDR5-8000 as well.
2)Investigating why DOOM Eternal score was low. (Bad data?)
3) Prices have been changed
Z RGB DDR5-7200 32GB(16GB*2) C34 Sliver $189
Z RGB DDR5-6600 32GB(16GB*2) C34 Sliver $153
Z RGB DDR5-6200 32GB(16GB*2) C32 Sliver $132

 

[Mussels]

@Mussels thanks for pointing these issues out. I will work on fixing them / making it more coherent .

DDR5-6400 works for me too, but I don't add it because that is well past what is normal. Only works on X670E ACE and X670E Taichi. Probably not anymore with the SoC voltage lockdown. It needed 1.35V for that.

Update:
1)Benchmark graphs and tables are fixed. Added DDR5-8000 as well.
2)Investigating why DOOM Eternal score was low. (Bad data?)
3) Prices have been changed
Z RGB DDR5-7200 32GB(16GB*2) C34 Sliver $189
Z RGB DDR5-6600 32GB(16GB*2) C34 Sliver $153
Z RGB DDR5-6200 32GB(16GB*2) C32 Sliver $132

From what i've seen on the SoC issue, some boards even when set to 1.1V in the BIOS (and reading 1.1v) were running at ~1.5v when measured - likely some over aggressive LLC.
It'd explain burnouts, if shortcuts were made to help the boards OC further or just a genuine screwup (higher RAM speeds/EXPO scaling SoC up voltage, but past a certain point they just never stopped scaling higher)

About a minute into this "What the hell... Maybe their SoC knob is broken?"

 

[ir_cow]

When I probe the bucks to find out which mosfet is which for MB reviews, it was always whatever voltage I had set it to or auto. Never seen it go to 1.5V for SoC

But I also only had DDR5-6000 2x16 in at the time and R23 load though, so I can see like 64GB or 6400 doing weird stuff or idle. Checked my CPU yesterday, no burn marks yet

 

[mechtech]

Nice review, I like the xmp vs jedec specs.

Speaking of jedec, what is the fastest ddr5 standard now? Still 4800?

 

[ir_cow]

Speaking of jedec, what is the fastest ddr5 standard now? Still 4800?

I don't know honestly. A lot of motherboards will do that auto boost to 5200 if you don't enable XMP/ EXPO. I assume it's still 4800 since all the memory has that as the highest JEDEC.

 

[mechtech]

I don't know honestly. A lot of motherboards will do that auto boost to 5200 if you don't enable XMP/ EXPO. I assume it's still 4800 since all the memory has that as the highest JEDEC.

Hmmmmm after some digging on Jedec.............seems you need to be a member to get any docs..........annoucements are open.

"JESD79-5A expands the timing definition and transfer speed of DDR5 up to 6400 MT/s for DRAM core timings and 5600 MT/s for IO AC timings to enable the industry to build an ecosystem up to 5600 MT/s. "

JEDEC Publishes Update to DDR5 SDRAM Standard Used in High-Performance Computing Applications | JEDEC

www.jedec.org

looks like -5B might be latest now https://www.jedec.org/standards-documents/docs/jesd79-5b

and possibly reserved up to 8400

Insights into DDR5 Sub-timings and Latencies

www.anandtech.com

interesting...........

 

[Mussels]

When I probe the bucks to find out which mosfet is which for MB reviews, it was always whatever voltage I had set it to or auto. Never seen it go to 1.5V for SoC

But I also only had DDR5-6000 2x16 in at the time and R23 load though, so I can see like 64GB or 6400 doing weird stuff or idle. Checked my CPU yesterday, no burn marks yet

Having read more of the findings, it's limited to various boards, especially on older BIOSes.

I'll keep editing this post instead of making new ones as it's related to high speed RAM, but by the time anyone buys more DDR5 RAM there will be updates to fix this to install and be able to use the ram safely.


"As soon as EXPO is enabled, the CPU SOC and CPU VDDIO/MC voltages increase to 1.36-1.4v, sometimes boosting to 1.5V in Windows, causing instant death on the X3D"
(Quoted from another TPU user who quoted someone else etc)

And then some boards have safeties disabled when PBO is enabled by using "motherboard defaults", so enabling PBO to access curve optimiser - which RAM reviews wouldnt use - would unlock all the amperages and wattages, and disable temperature based safeties too. Those "motherboard defaults" are basically "infinity"

Then theres a genuine bug that may be AMD related, where sleep or hibernate is triggering issues with temperature detection and fan speeds - which may be semi-related, or just a coincidence (confirmed on MSI boards)

Seems like the issues vary between the motherboard brands, but AMD is expected to provide a safety net for them in new AGESA.

DDR5 Expo voltages being cloned to VSOC has been reported on Asus and Gibabyte boards so far, so yeah definitely keep an eye out on your review system(s)

Edit: Asus was automatically setting SoC 1.35v when EXPO was enabled, but over time was creeping up higher over time to 1.5v as the CPU aged - the longer it ran at higher voltages, the less resistance the CPU had internally and the more power went through.

Edit 2: Asus boards have an insanely high OCP (overcurrent protection) effectively infinite.

TL;DR: You'd get a crash or hang, the system would try and reboot (the famous ryzen slow boot bug) and just keep feeding more and more power until the system powered on, or exploded

 

[ir_cow]

Interesting. So PBO is to blame as it disables some of the safeties?

I guess I'm lucky that with all these memory testing hours, my CPU is fine. No bulge or burn marks. DDR5-6200 2x 16GB works fine at 1.25v. don't remember what DR rank needed though. Time to check.

 

[Mussels]

PBO can raise the motherboards safeties, at the very least.
If the system is frozen and not working normally, the CPU's safeties cant kick in.


If the CPU is in a frozen state and the motherboard assumes the CPU will control the safety features, i can totally see how this is happening - high voltages cause a crash, the CPU fails to recover without user intervention, she burns up.