Is Y-Cruncher Pass required to be considered stable?

[ir_cow]

To keep this short. Y-Cruncher is a DRAM benchmark http://www.numberworld.org/y-cruncher/ to compute Pi. I have found that when the system memory or CPU OC is not stable, the test will fail either by BSOD / Instant Restart or a test failure message saying the cache is too large (or something of that nature).

There have been a number of instances where the memory will "pass" TestMem5 or Karhu stress (1000%+), but fail the benchmark. Usually the remedy is to tweak the timings, raise the CPU memory related voltages or occasionally just DRAM voltage.

Where do you draw the line? For example, DDR5-6400 can boot into windows for AMD, but only one motherboard so far has passed y-cruncher 2.5B for me. DDR5-8000 works on a number of SK Hynix-A die for Intel, yet only passes y-cruncher with a cold boot (or much loser secondary timings / tertiary when warmed up).

Do you believe passing Y-Cruncher 2.5B / 10B (32GB) / 64GB) should be a requirement to consider it stable system memory?

 

[freeagent]

To call it stable I would think so. It should be able to run anything like a stock machine. But if you are just messing around.. nah.

 

[ThrashZone]

Hi,
I'd likely call it a day and stable if all is well using blender opendata "full test" which can be switched between cpu or gpu testing.
If all good with this gaming or benchmarks should go smoothly


Blender Open Data — blender.org

 

[Solaris17]

Is 2.5b the standard measurement? (asking because im curious what the meta is) as for considering it stable, I honestly always chase the dream. At one point super-pi was a stability test. When it wasnt as effective the goal posts moved etc etc etc etc.

If I want my things to be stable, I will always test them with the tool most likely to fail it.

 

[ir_cow]

Is 2.5b the standard measurement? (asking because im curious what the meta is) as for considering it stable, I honestly always chase the dream. At one point super-pi was a stability test. When it wasnt as effective the goal posts moved etc etc etc etc.

If I want my things to be stable, I will always test them with the tool most likely to fail it.

2.5B because it eats up more ram. Only works on 32GB or above. 10B for 64GB. I found ycuncher to be the hardest by accident while trying to submit scores to HWBot when DDR5 first came out. Was wondering why people weren't submitting 2.5B scores, while the lower 1B and 25m was full of super high frequencies. Turns out, its much harder to get a passing result the higher you go.

So while, you could do everything gamers do on the daily. Chances are its not entirely stable. I've resolved this by changing the timings, changing the DRAM voltage and messing with CPU voltages. It could be all 3, just one or none because that MB or CPU just can't handle it.

But if its the ONLY program to fail, is it truly the best program to set the bar to, or is it kinda like Prime95 for CPUs these days?

 

[freeagent]

So while, you could do everything gamers do on the daily

Honestly I would say gamers run the most unstable systems just for the simple fact that if it plays their game it’s all good. Lots of people think certain program loads are not real world.. but they are.. just not in their world.

 

[igralec84]

I did 140 hours of gaming on a RAM OC that would fail karhu at 850-1000% and TM5 6 errors in 25 cycles without a single crash as the crashes came on idle randomly or in benchmarks

Can now pass everything with the new ram, 6400 CL38 @ 6200 CL32 Also get to the logon screen of windows then BSOD at 6400 c36

 

[mechtech]

I usually run p95 for 24hours and a run of memtest86 (+) But I don't OC though, stock factory defaults.

 

[ir_cow]

passive aggressive

 

[RJARRRPCGP]

Also get to the logon screen of windows then BSOD at 6400 c36

Let me guess? WHEA for memory or "PAGE_FAULT_IN_NON_PAGED_AREA" BSOD. Maybe "PFN_LIST_CORRUPT" BSOD?

("non paged" means the RAM, so there's RAM corruption)

 

[qubit]

My system has been super stable since I installed the Corsair AX1600i. The dodgy RAM stick i'd finally replaced a couple of years ago. The PC isn't overclocked.

Intrigued to give this a go and see if it passes.

 

[DemonicRyzen666]

No.
24+ hours on idle should be a test for stablity, or even light loads for a whole day.
seen too many people crash even after passing all these only to crash while gaming or watching youtube video.

 

[ir_cow]

No.
24+ hours on idle should be a test for stablity, or even light loads for a whole day.
seen too many people crash even after passing all these only to crash while gaming or watching youtube video.

Game drivers is a whole different thing.

 

[DemonicRyzen666]

Game drivers is a whole different thing.

I'm not talking about game drivers.
I'm talking about general stablity during daily tasks.

 

[ir_cow]

I'm not talking about game drivers.
I'm talking about general stablity during daily tasks.

If you get BSOD on the desktop, that could be anything. but if your saying it passes all these stability tests for memory and crashes while gaming or watching youtube, that's probably unrelated to system memory.

 

[The King]

Unstable CO can also also cause Y-cruncher 2.5B to fail so its not just about the system memory it goes to overall system stability.

OCCT is also a good to run for overall system stability testing but that needs at least 1 hour and it detects WHEA errors which is something other tests don't do.

I would called Y-cruncher 2.5B a quick stability test and 1 hour OCCT a proper system stability.

If its not Y-cruncher 2.5B and 1 hour OCCT stable its unstable end of story. IMHO

 

[Frick]

If you get BSOD on the desktop, that could be anything. but if your saying it passes all these stability tests for memory and crashes while gaming or watching youtube, that's probably unrelated to system memory.

Random crashes but passing stability/memory tests absolutely sounds like RAM problems to me, depending on the platform.

 

[igralec84]

Let me guess? WHEA for memory or "PAGE_FAULT_IN_NON_PAGED_AREA" BSOD. Maybe "PFN_LIST_CORRUPT" BSOD?

("non paged" means the RAM, so there's RAM corruption)

Actually first it was Kernel_security_check_failure, then i think ACPI_bios_error or something like that. Changing back to 6200 boots like nothing happened, sfc /scannow and DISM don't find anything corrupted.

After posting that yesterday, i got 3 random hard reboots in the space of 15 minutes for the first time after a week lol. Must be the CO and PBO combination then, i can replicate it while running 3Dmark CPU profile with affinity to core 3 thread 2 with PBO +100 in 4/5 times, while PBO +75 doesn't crash in 5/5 times on that thread

 

[mrthanhnguyen]

Good luck for using ygrinder to test stability on Intel 13th gen where it pulls 400w or so.

 

[Ferrum Master]

There is no universal answer as each case is platform and your sample specific. IMC's are capricious and act differently in each board, that's normal, now we have higher frequencies and random impendence changes and thus much more variables. And the great unknown... like for AMD what magic AGESA does sometimes, we never know it fully.

Basically for RAM I use OCCT for whole night at various modes. RAM testing and core cycler. The problem with Y crunchers etc... those are not dynamic enough, use specific instruction sets and the CPU prediction brancher get used to it and the tests are not random enough to trigger instabilities unlike gaming does.

Usually you feed more voltage than you need and turn it down till you start to become unstable in stress tests, not vice versa. Then you know what part is actually responsible for instability faster.

 

[ThrashZone]

My system has been super stable since I installed the Corsair AX1600i. The dodgy RAM stick i'd finally replaced a couple of years ago. The PC isn't overclocked.

Intrigued to give this a go and see if it passes.

Hi,
Hard to believe you would get a 1600w psu for the system on your spec's page

 

[3x0]

Do you believe passing Y-Cruncher 2.5B / 10B (32GB) / 64GB) should be a requirement to consider it stable system memory?

For me, yes. However it's worth considering that some firmware/BIOS bugs affect stability of a system that's beyond our control. I remember having trouble with 6700k OCing, it would randomly fail in Prime95 when all other stress tests passed. Later, Intel issued a firmware fix as it turns out the CPU had a bug with Prime95

 

[ir_cow]

Good luck for using ygrinder to test stability on Intel 13th gen where it pulls 400w or so.

You can lower the power limit you know

 

[mrthanhnguyen]

You can lower the power limit you know

Only if u run stock or downclock your cpu, any oc will pull the insanse wattage in ycruncher.

 

[ir_cow]

Only if u run stock or downclock your cpu, any oc will pull the insanse wattage in ycruncher.

No more than Cinebench R23.