Looks interesting, looking forward to more reviews.

 

Quote:

Originally Posted by chibi

Looks interesting, looking forward to more reviews.

I can guarantee this is currently best ramtester there is, I've been using this software almost as long as @CptKuolio and like him, i dont have affiliations with the creator.

 

Does this also stress the Cache similar to how HCI Memtest does?

 

Quote:

Originally Posted by CptKuolio

I rate Ram Test as the best memory testing utlity at this time, bar from using linux-mint and google's memorybench. But out of standalone utilities/programs in windows this is by far the most accurate, thorough and FAST. I repeat; about 15 times faster then HCI memtest. And no need to launch multiple instaces of the program! I can push 703% coverage on 12288 megabyte memory-run in 10 minutes, when HCI memtest only does ~45% covera in same time. HUUUUUUGE! Check my video, I show you the two of them against each other

You can skip to 12:00 mark to see how they compare, the beginning of the video is pretty much just showing of the program, and telling some tales about it's origins.

It's possible to run Stress App via BASH. So are you saying this is better than running Stress App under Windows? Also no limited freeware version to test, we're meant to take your word for it?

 

I've just bought the tool and testing with known conditional settings that I use regularly with Stress App test. The tool is picking up errors in under 5 minutes where Stress App under BASH terminal normally takes 20 to 30 minutes. Obviously, more time is needed to see how well this isolates the memory but this tool is definitely worth a try.

 

Quote:

Originally Posted by Silent Scone

I've just bought the tool and testing with known conditional settings that I use regularly with Stress App test. The tool is picking up errors in under 5 minutes where Stress App under BASH terminal normally takes 20 to 30 minutes. Obviously, more time is needed to see how well this isolates the memory but this tool is definitely worth a try.

You're a trusted source of vital info on overclock.net, Scone, I'll give it s try.

 

Damn. Might give this a try! Please tell me how it goes!

 

Quote:

Originally Posted by chibi

Does this also stress the Cache similar to how HCI Memtest does?

I'm not familiar with how HCI stresses cache. RAM Test is built to test memory stability. Will ask Mustanaamio about this, though he is now aware of this thread. I dont know if he has account here.

Quote:

Originally Posted by Silent Scone

It's possible to run Stress App via BASH. So are you saying this is better than running Stress App under Windows? Also no limited freeware version to test, we're meant to take your word for it?

No, can't run it on bash. Not sure what you are referring to with Stress App under Windows. This is Tool for memory stresstesting, nothing else. Better then HCI memtest, memtest 86 etc. for sure.

Yes unfortunately you do need to take my word for it. No demo or shareware available at the moment. If 10€/$ is too much of an gamble then dont buy it yet, wait for someone you know better than me to recommend it.

 

Quote:

Originally Posted by CptKuolio

I'm not familiar with how HCI stresses cache. RAM Test is built to test memory stability. Will ask Mustanaamio about this, though he is now aware of this thread. I dont know if he has account here.
No, can't run it on bash. Not sure what you are referring to with Stress App under Windows. This is Tool for memory stresstesting, nothing else. Better then HCI memtest, memtest 86 etc. for sure.

Yes unfortunately you do need to take my word for it. No demo or shareware available at the moment. If 10€/$ is too much of an gamble then dont buy it yet, wait for someone you know better than me to recommend it.

I'm referring to Google Stress App Test ran through BASH terminal.

 

Quote:

Originally Posted by Silent Scone

I've just bought the tool and testing with known conditional settings that I use regularly with Stress App test. The tool is picking up errors in under 5 minutes where Stress App under BASH terminal normally takes 20 to 30 minutes. Obviously, more time is needed to see how well this isolates the memory but this tool is definitely worth a try.

I did also, known good config ran 1H clean. Will test some more. But yeah, worth adding to the box of ram tools. (was $12 US - a bit pricey IMO)
Quote:

Originally Posted by Silent Scone

I'm referring to Google Stress App Test ran through BASH terminal.

^^ which works very well in isolating the ram subsystem for stability testing. I still use a puppy linux USB to run it on various rigs.

 

Need to find some settings that are borderline stable to get a feel for error detection rates.
Quote:

Originally Posted by Silent Scone

I've just bought the tool and testing with known conditional settings that I use regularly with Stress App test. The tool is picking up errors in under 5 minutes where Stress App under BASH terminal normally takes 20 to 30 minutes. Obviously, more time is needed to see how well this isolates the memory but this tool is definitely worth a try.

What settings do you use with stressapptest?

On most of my systems, the default settings are not the fastest at finding memory errors.

 

Quote:

Originally Posted by Blameless

Need to find some settings that are borderline stab
What settings do you use with stressapptest?

On most of my systems, the default settings are not the fastest at finding memory errors.

that's not a problem.

Also want to be sure there's no false positives (errors in this case).
for bash GSAT, I use: stressapptest -W -M 28672 -s 7200 --pause_delay 14400

 

Will jump on the Chu-Chu train and buy it myself as well..

Always good to support new projects

 

Quote:

Originally Posted by Jpmboy

that's not a problem.

Also want to be sure there's no false positives (errors in this case).
for bash GSAT, I use: stressapptest -W -M 28672 -s 7200 --pause_delay 14400

Have you tried adjusting the thread counts and or omitting the -W? Sometimes 1.5 threads per logical CPU, or the less CPU intensive memory copy tests in combination with the inversion tests can prompt errors faster.

 

Quote:

Originally Posted by Silent Scone

I'm referring to Google Stress App Test ran through BASH terminal.

I have 'limited' amount of experience with that, then! I began my memory OC tries and "getting into memory beyond XMP profile" stuff by reading guides on this forums, like the 'Official Intel Memory stability thread' and the 'Comprehensive Memory Overclocking Guide'. Those threads were pretty much my starting point for me about 8-9 months ago, little after Ryzen launch and little after The Stilt was hyping memory speeds on Ryzen at io-tech forums. Afterwards I closely followed ryzen overclocking and ryzen memory threads here.

I followed the procedures given in those threads, and so I also tried the Google Stress App, with linux mint on USB drive. Also used the same drive to run /kill ryzen and determined that mine was one of those that segfault, but that is another story. But also that was something I learned from this forum. Nevertheless I found it 'cumbersome' to say the least to boot into Linux and run stress-test while having all other OC-tools on windows (ryzen timing checker, at that time ryzen master because nothing else showed some values correct, hwinfo etc.) so I gave up on that and started lookin at HCI memtest. But with HCI-Memtest the whole process of testing memory was so slow and felt so 'inaccurate', that it was a total pain to hunt for the correct settings with Ryzen.

This all changed after Mustanaamio released the first iteration of this software and stilt was raving over it, because it made his workflow a lot faster with the ryzen timing things he was doing back then. RAM Tests first iteration was fast, a lot faster then other windows utilities. Mustanaamio has been updating RAM Test over the past months, and it has been getting even more faster all the time. I used to run a lot of HCI memtest after running RAM test, to verify that the RAM is 'hci-stable', and from that I noticed that many times RAM Tester showed up errors that HCI did not.

So I avoid comparing RAM Test to Google's Stress App, because I have too limited experience with it. I've only ran it a handful of times. I see that people here are testing RAM Tester and comparing it to Stress App, and it is very interesting to follow! Thanks guys

But I also think that perhaps Stress App is not the 'main competitor' here, but the windows platform programs are. Nevertheless, this was what I was expecting, to see OC guys and gals to get arms deep to testing RAM Test

My motivation to bringing this RAM Test software to this forum is simple; I've learned so much from threads here that I feel like it's only fair to give back. RAM Test revolutionized how I feel about memory overclocking. Prior to to RAM Test it was a big and slow process to OC memory, mainly because it took so long to stress test between steps. Now I can literally run RAM Test for ten minutes, get to 700% coverage, and that gives me enough info if I'm in between steps and want to go further, or finetune and continue benching with current values. With other tools it takes much longer, every step you bench between different settings, even when changing one value by 1, takes hours. With ram-test it's a lot more like OC'ing your CPU in a way, where you hit prime running for 20 minutes between steps and see if it is "stable enough", and boot to bios to up the frequency even more

 

Quote:

Originally Posted by Blameless

Need to find some settings that are borderline stable to get a feel for error detection rates.
What settings do you use with stressapptest?

On most of my systems, the default settings are not the fastest at finding memory errors.

What settings are you using? There are arguments that work better with cache, I've not used any that find memory errors quicker.
Quote:

Originally Posted by Jpmboy

I did also, known good config ran 1H clean. Will test some more. But yeah, worth adding to the box of ram tools. (was $12 US - a bit pricey IMO)
^^ which works very well in isolating the ram subsystem for stability testing. I still use a puppy linux USB to run it on various rigs.

It's simple here, as I know that simply dropping tRCD one clock is enough to throw 2 or 3 errors in GSAT.

 

Quote:

Originally Posted by Silent Scone

What settings are you using? There are arguments that work better with cache, I've not used any that find memory errors quicker.

Depends on the platform.

My LGA-2011v3 parts seem to crap out fastest simply by mixing some inversion threads in to the test, while my Vishera parts seem to fail faster with the less intensive copy test (probably a bottleneck somewhere using -W on them).

 

Thumbs up for this software from me as well. Very precise and fast as hell.

 

sub for later use

 

I just got this last night.Works fast in finding errors..When running this tester how much free memory should you leave running in the background so windows does not crash.

 

Quote:

Originally Posted by NIK1

I just got this last night.Works fast in finding errors..When running this tester how much free memory should you leave running in the background so windows does not crash.

I close all programs and then dial in the same amount of memory to be tested as the program shows to be available.

 

K..Thanks for the info...

 

friend i cant use paypal how can we do it ?

 

Quote:

Originally Posted by NIK1

I just got this last night.Works fast in finding errors..When running this tester how much free memory should you leave running in the background so windows does not crash.

It doesn't matter. Windows will crash if your memory is unstable enough.
Quote:

Originally Posted by ScriptKiddie

friend i cant use paypal how can we do it ?

Unfortunately PayPal is currently the only way for international customers to buy a license for the software. It will take some time to integrate with the APIs of other PSPs.

 

Quote:

Originally Posted by Mustanaamio

It doesn't matter. Windows will crash if your memory is unstable enough.
Unfortunately PayPal is currently the only way for international customers to buy a license for the software. It will take some time to integrate with the APIs of other PSPs.

It does matter. Allocating more RAM than what is available is a great way of testing the page file, but not DRAM.