• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** Official Ryzen Owners Thread ***

eyetrip said:
your system is still responsive running tpu memtest?

How long do you run it for our of curiosity?

full 16 threads?
Click to expand...

Yes, in the sense the mouse doesn't lock up, but i don't really do anything whilst benchmarks and stress test apps are running. Just screenshots and maybe a bit of browsing.
Full 16 threads, only 20 minutes as i use it for quick testing only when settling the settings. I use HCI for at least 500% for end stability on memory as it's far better at throwing up errors.
Prime 95 and RealBench for at least 2 hours.
 
Rossi~ said:
Yes, in the sense the mouse doesn't lock up, but i don't really do anything whilst benchmarks and stress test apps are running. Just screenshots and maybe a bit of browsing.
Full 16 threads, only 20 minutes as i use it for quick testing only when settling the settings. I use HCI for at least 500% for end stability on memory as it's far better at throwing up errors.
Prime 95 and RealBench for at least 2 hours.
Click to expand...

What Prime test Rossi?

I'm thinking of ditching IBT AVX and moving to Prime 95 or Aida, and HCI.
 
Hi guys, new to ryzen yesterday and must say I'm quite impressed.

Setup is

Gigabyte ab350m gaming 3
Ryzen 1600
2 x 4gb team group dark 3200
R9 380 sapphire nitro 4gb
750w evga 80+bronze

Had a good mess around yesterday and finally have the ram running at 3200mhz and a mild oc on the cpu at 3.8ghz with an offset of +0.090 which gives a max bore of 1.356v temp top out at 60c after an hour or so in prime.

Question is what is a safe 24/7 volts for these chips as even at 1.4v 3.9 crashes within mins of stress tests and don't think the extra volts is worth it for an extra 100mhz.

Any advice at all
 
Rossi~ said:
Yes, in the sense the mouse doesn't lock up, but i don't really do anything whilst benchmarks and stress test apps are running. Just screenshots and maybe a bit of browsing.
Full 16 threads, only 20 minutes as i use it for quick testing only when settling the settings. I use HCI for at least 500% for end stability on memory as it's far better at throwing up errors.
Prime 95 and RealBench for at least 2 hours.
Click to expand...

Using it for a quick test is a bit pointless as I found yesterday when trying to test things.

I ran it after starting at one point and it crashed in a couple of minutes.

Now you may be inclined to think it's doing a good job of throwing up errors quickly - I then immediately ran it again to see how quickly it failed again.

7 hours it took.

Plus, the lock-ups get worse the longer it runs for me.

Currently at 45 min, and lockups appear to be around 5 seconds.
 
eyetrip said:
Using it for a quick test is a bit pointless as I found yesterday when trying to test things.

I ran it after starting at one point and it crashed in a couple of minutes.

Now you may be inclined to think it's doing a good job of throwing up errors quickly - I then immediately ran it again to see how quickly it failed again.

7 hours it took.

Plus, the lock-ups get worse the longer it runs for me.

Currently at 45 min, and lockups appear to be around 5 seconds.
Click to expand...

That's the joy of all this stuff, everyone has different experiences :D :p

Just out of interest, what MEMCLK are you running? Just wondering whether it's related to the DF on high loads moving threads about the shop. This runs at MEMCLCK so maybe if it's too low and slow or too high and unstable on shifting the workload about? Don't know, just food for thought.
 
Last edited:
Rossi~ said:
That's the joy of all this stuff, everyone has different experiences :D :p

Just out of interest, what MEMCLK are you running? Just wondering whether it's related to the DF on high loads moving threads about the shop. This runs at MEMCLCK so maybe if it's too low and slow or too high and unstable on shifting the workload about? Don't know, just food for thought.
Click to expand...

This stopped being joyful a couple of months ago :/

It's been the most frustratingly inconsistent system I have ever had the displeasure of owning.

I mean, I disliked my z97/4790k a bit as it was a little temperamental initially, but not even close to being in the same league as this.
 
eyetrip said:
This stopped being joyful a couple of months ago :/

It's been the most frustratingly inconsistent system I have ever had the displeasure of owning.

I mean, I disliked my z97/4790k a bit as it was a little temperamental initially, but not even close to being in the same league as this.
Click to expand...

I know what you mean, a lot of people have had many issues. I haven't, just been lucky maybe.I must admit though, I haven't enjoyed the fact that we've been beta testers for this platform.

If I had as many issues as some out there, i would've sold up by now and gone Intel.
 
AMDMatt said:
What's the highest stable CPU frequency you've achieved with the highest memory frequency?
Click to expand...

As a combination of the two? I can do:

4.1GHz and 3466MHz, but with high volts.
4.0GHZ 3600MHz, with high volts and loose timings

More realistically, I'm settling with either of these:

CPU @ 4040MHz (101x40)
RAM @3366MHz with tightened timings

or

CPU @ 4025MHz (100x40.25)
RAM @ 3333MHz with tightened timings

3200-3466MHz is the sweet spot. Latency means a lot to Ryzen at this point, so it's worth hitting the breaks on speed and getting the timings down.

It's a slightly different ballgame if you play with high bclk/refclk as this puts less requirement on the memclk so you should be able to reach higher frequencies this way.
 
so, currently observing the behaviour of different procodt settings, in a near stable setup running tpu memtest

used to use just 48, as seemed to be the only stable one before.

so, I was experiencing the long lock-ups. went up to 53, and lock-ups don't appear to be worse than about 5 seconds. mouse does not move completely smoothly.

60 results in some long lock-ups fairly soon, mouse very, very juddery

68 results in long lock-ups before long, and mouse very juddery

80 results in mouse smooth, but did notice a minute+ lock-up fairly early.

Bear in mind, this is based on first few minutes of running the test.

Need to go back to 48 now, as I cannot remember what mouse movement like, and need to pay attention early test lock-ups.

Edit:
48 - mouse smoother than 53, but system locks up more often and had a longer lock-up fairly early. Also just had display cut out on this one.

Going to play with 53 and 80... I hate this.

Edit 2:
System doesn't seem to like cold booting at 80, so playing with 53. For ***** and giggles, have lowered my dram voltage too.

Edit 3:
Dropping timings down to 16-18-18-18-38 form default 15-17-17-17-36, seems to result in slightly less lock-up. This is at 2933 for ram.
 
Last edited:
Yep my 1700 sweet spot seemed to be 39.25 x 100 with 3333c14. Beyond that on the clock needs excess volts although I've not yet played with LLC.

Can't get higher than 3333 without sporadic boot looping so have left it at that and dropped my timings down.
 
eyetrip said:
so, currently observing the behaviour of different procodt settings, in a near stable setup running tpu memtest

used to use just 48, as seemed to be the only stable one before.

so, I was experiencing the long lock-ups. went up to 53, and lock-ups don't appear to be worse than about 5 seconds. mouse does not move completely smoothly.

60 results in some long lock-ups fairly soon, mouse very, very juddery

68 results in long lock-ups before long, and mouse very juddery

80 results in mouse smooth, but did notice a minute+ lock-up fairly early.

Bear in mind, this is based on first few minutes of running the test.

Need to go back to 48 now, as I cannot remember what mouse movement like, and need to pay attention early test lock-ups.

Edit:
48 - mouse smoother than 53, but system locks up more often and had a longer lock-up fairly early. Also just had display cut out on this one.

Going to play with 53 and 80... I hate this.

Edit 2:
System doesn't seem to like cold booting at 80, so playing with 53. For ***** and giggles, have lowered my dram voltage too.

Edit 3:
Dropping timings down to 16-18-18-18-38 form default 15-17-17-17-36, seems to result in slightly less lock-up. This is at 2933 for ram.
Click to expand...


I had these mini lock ups a while back. I tried to narrow the cause down but could not find it. I made no changes in the bios and reinstalled windows.
I spent hours installing one program at a time and running realbench inbetween but I could not find the source. But it did fix my pausing problem. Note that my memory can only boot at 60 ohms, anything else is an instant f9.
 
eyetrip said:
so, currently observing the behaviour of different procodt settings, in a near stable setup running tpu memtest

used to use just 48, as seemed to be the only stable one before.

so, I was experiencing the long lock-ups. went up to 53, and lock-ups don't appear to be worse than about 5 seconds. mouse does not move completely smoothly.

60 results in some long lock-ups fairly soon, mouse very, very juddery

68 results in long lock-ups before long, and mouse very juddery

80 results in mouse smooth, but did notice a minute+ lock-up fairly early.

Bear in mind, this is based on first few minutes of running the test.

Need to go back to 48 now, as I cannot remember what mouse movement like, and need to pay attention early test lock-ups.

Edit:
48 - mouse smoother than 53, but system locks up more often and had a longer lock-up fairly early. Also just had display cut out on this one.

Going to play with 53 and 80... I hate this.

Edit 2:
System doesn't seem to like cold booting at 80, so playing with 53. For ***** and giggles, have lowered my dram voltage too.

Edit 3:
Dropping timings down to 16-18-18-18-38 form default 15-17-17-17-36, seems to result in slightly less lock-up. This is at 2933 for ram.
Click to expand...


I had these mini lock ups a while back. I tried to narrow the cause down but could not find it. I made no changes in the bios and reinstalled windows.
I spent hours installing one program at a time and running realbench inbetween but I could not find the source. But it did fix my pausing problem. Note that my memory can only boot at 60 ohms, anything else is an instant f9.

Sorry for the quad post! Damn mobile.
 
gavinh87 said:
I had these mini lock ups a while back. I tried to narrow the cause down but could not find it. I made no changes in the bios and reinstalled windows.
I spent hours installing one program at a time and running realbench inbetween but I could not find the source. But it did fix my pausing problem. Note that my memory can only boot at 60 ohms, anything else is an instant f9.

Sorry for the quad post! Damn mobile.
Click to expand...

Am going to test on a clean install of creator update now, with no apps, just stress it and see.

Can't seem find any setting in bios so far that improves on the changes I already made.
 
eyetrip said:
Am going to test on a clean install of creator update now, with no apps, just stress it and see.

Can't seem find any setting in bios so far that improves on the changes I already made.
Click to expand...

Thats the only way I cured mine mate. Took me a while and I never found the problem although I suspect it was AIsuite as I never reinstalled that. Or a corruption on the windows install :/
 
Ste said:
1.4v in BIOS which is 1.335-1.35v under load after drop and droop. I will be sorting LLC soon so hoping to get that down a bit

About 60c but that's on a 360mm AIO.
Click to expand...

Sounds similar to mine, I run 1.35v in the BIOS but also not played with LLC yet, my temps are 65c on 240mm AIO under Realbench testing.
 
You must log in or register to reply here.
Back
Top Bottom