• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

AMD Ryzen Thread: Affordable Core Act

Thanks. This should be an option in the BIOS settings, right? I'll do some searching.

CnQ should be in the BIOS, yeah. Check under CPU Options or CPU Features or some sub-menu with a name similar to that.

I know this isn't an all-inclusive AMD thread, but anyone have thoughts on the Vega launches? Seems like the Vega 56 will be pretty solid with 3rd party coolers, but the Vega 64 seems a bit overpriced for what it gives you. I'm glad they aren't total duds, but it still feels like AMD's GPU division is playing catch-up with Nvidia at this point.
 

JohnnyFootball

GerAlt-Right. Ciriously.
Put together a new Ryzen build on the weekend but have unfortunately been encountering constant BSOD crashes, often only minutes after having boooted into Windows.

My build is; https://pcpartpicker.com/list/HmF4bj
BIOS screens; http://imgur.com/a/YAzrO

I have attempted;
  • Reseating all components,
  • Removing all optional drives (leaving m.2),
  • Updating all drivers to latest,
  • Updating Bios to latest,
  • Formatting and reinstalling Windows fresh

Even with the above, on a fresh clean install with latest manufacturer drivers I'm seeing the same behaviour. The machine will noticeably hitch, frequently noticeable when playing a game for example, and eventually freeze into a reboot or BSOD.

I'd consider the GPU a culprit, but it's the one component I kept from my past build and has always been completely stable.

Most probable culprit seems to be RAM, especially since Ryzen is known to be sensitive and it's not a kit on the boards QoL (though before buying I've seen others have success). Aware the RAM is known to have corruption issues, but haven't seen and aren't using RGB software.

Failing that it could be a motherboard/PSU issue, which I'm hoping not.

Considered next steps;
  • Manually clock the RAM, particularly raising voltage (advice here appreciated, see BIOS screens above),
  • Try to stress test with single RAM (s) inserted,
  • Run another mem check (last one was error free at 95%, then froze on the reboot so I missed the results, and took several hours -_-),
  • Try my prior build PSU,
  • Cry into a pillow and return parts (whilst I can)

Anyone faced similar issues or have any possible solutions I'm not currently considering?

EDIT: Missed the part where you are using PC part picker to describe your build instead of being one you just purchased.

Thank god. I thought you had lost your mind spending $400 on a Geforce 970.
 

Wozzer

Member
I'd verify the vCore values with HWiNFO under Windows. Start a stress test like RealBench, Prime95 or Cinebench R15 single/multi-core test, and double click the vcore line in HWiNFO to get a value diagram.
It's possible the board isn't supplying enough voltage regardless of the BIOS setting (excessive voltage droop under load for ex.).
Have you saved any previous BSOD dumps? If yes, are the BugCheck codes the same always (0x0000003b in this case)
It's very weird since you're pretty much running stock BIOS settings, so there shouldn't be issues like this.

Bumped the SOC to 1.0 from 0.9, and RAM voltage to 1.3 from 1.0.

Ran one stick (8gb) at 15minutes stress using up to 4gb test on RealBench no problem, though motherboard TMPIN2 hit 98*.

Tried the same with 4 sticks (32gb) using up to 16gb in the test and it froze the machine 5minutes into it.

So I'm currently doing every stick for 15mins, and if that goes without locking up I'll try 2 sticks. Potential issue running 4stick/32gig?

AS of today, I am able to reach the speed of 3066 just fine. there is not tweaking. The only thing you need to do is change to DOCP and pick the 3066 speed. the voltage will be increased automatically to 1.3. Everything else I left the same. Basically just change DOCP

Using DOCP with standard settings had same result. I am assuming you're using dual sticks at 16gig instead of quad at 32?
 

Wozzer

Member
Crashed on the third stick, with this photo taken just before it locked up and restarted;
728FRTH.jpg
 

3x0

Neo Member
Bumped the SOC to 1.0 from 0.9, and RAM voltage to 1.3 from 1.0.

Ran one stick (8gb) at 15minutes stress using up to 4gb test on RealBench no problem, though motherboard TMPIN2 hit 98*.

Tried the same with 4 sticks (32gb) using up to 16gb in the test and it froze the machine 5minutes into it.

So I'm currently doing every stick for 15mins, and if that goes without locking up I'll try 2 sticks. Potential issue running 4stick/32gig?



Using DOCP with standard settings had same result. I am assuming you're using dual sticks at 16gig instead of quad at 32?

Download Thaiphoon Burner http://www.softnology.biz/files.html and run it as administrator. Click the Read button and post a screen shot of the output.
It's very likely your RAM chips are from Hynix, which aren't very compatible with Ryzen, especially in 32GB configs.

Some misc. remarks. Don't run multiple HW monitoring programs, as it can lead to various glitches. For ex. RealBench doesn't seem to be able to read the temps of the CPU. HWMonitor as well, as well as voltages (not sure if you're running the latest version).
Are you sure RAM was at 1 volt before upping it to 1.3? Default JEDEC value is 1.2v for 2133MHz DDR4, 1.35v for DOCP.
 

Wozzer

Member
Download Thaiphoon Burner http://www.softnology.biz/files.html and run it as administrator. Click the Read button and post a screen shot of the output.
It's very likely your RAM chips are from Hynix, which aren't very compatible with Ryzen, especially in 32GB configs.

Some misc. remarks. Don't run multiple HW monitoring programs, as it can lead to various glitches. For ex. RealBench doesn't seem to be able to read the temps of the CPU. HWMonitor as well, as well as voltages (not sure if you're running the latest version).
Are you sure RAM was at 1 volt before upping it to 1.3? Default JEDEC value is 1.2v for 2133MHz DDR4, 1.35v for DOCP.

AnWt5os.png
 

Wozzer

Member
Yup, Hynix. Try getting them exchanged for Samsung B-die versions as they're much more likely to run in 4x8 configs. You can see the list of Samsung B-die RAM here: http://www.overclock.net/t/1627555/ryzen-memory-ic-collection-thread

Good to know. Will see if I can return to Newegg and get some replacement. The motherboard QVL (here) coupled with your link has the G.Skill F4-3200C16Q-32GFX/F4-3200C16Q-32GFXR looking as good contenders.

Also ran a 1hour stress test, and after about 40minutes it halted with 'Instability Detected (7z)'; but no further context beyond that.
 

Wozzer

Member
Just purchased "F4-3200C14D-16GTZ", which is listed both as Samsung die and on the motherboard qvl list. Means a drop from 32 to 16gig, but if that comes at stability and saves me a little cash then it'll suffice.

Will start the refund process with the RGB sticks. Should have new ones by tomorrow.
 

kotodama

Member
Finally got my 1950x built.

The Corsair Crystal 460X was annoying to build in, but it does look nice. My cable management, on the other hand, looks like utter trash. RGB fans are stupid, but I can't stop looking at them and changing the colors. Case supports three more it seems. Thinking now, I should've got the RGB ram....or not.

Anyways, I was kind of freaked out about getting Threadripper set in it's LGA slot since I saw HardOCP's video with the bent pins. That part was actually pretty darn easy to do without any issue. The problem was that afterwards the three screws on the slot cover wouldn't thread through. I would crank and crank and they wouldn't catch. 2 Pints of Belgian Strong (Alesmith Horny Devil) later, and putting a lot of pressure on the CPU cover and the board itself, I got it to thread through. However then it wouldn't be aligned to the other two screws. I had to just get it to catch slightly on a screw and then go from 1 -> 2 -> 3 and repeat, cranking them like 6 short turns a piece a round till the torx wrench finally pulls back. When it catches it's not a click as I would have thought, but it really is a nice pull saying it's set. Seriously, I thought I might have gotten a board with a defective CPU slot mounting mechanism.

After that it was a comedy of building errors all in some ways case related. Even when I got to the Windows install, since there is no USB 2.0 on the tower nor the back of the ASROCK X399 motherboard, I had to make a new GPT Windows USB install, yah.

Anyways, downloading and setting up my stuff atm. My ram is rated at 3600 mhz, but can't get it passed 2100 mhz. However I really haven't tried to manually mess with it too much yet.

The 1950x is resting mostly at 2195 mhz according to HWiNFO with one or two cores shooting up to 4191.3 mhz once in a while. Average temp from my Corsair H100iV2 is 35.7C at the moment with a max of 52.6C. Offset on the Tctl is at +27C. Seems alright at the moment without full coverage on chip. However I haven't downloaded anything that'll truly stress it yet.
 
Finally got my 1950x built.

The Corsair Crystal 460X was annoying to build in, but it does look nice. My cable management, on the other hand, looks like utter trash. RGB fans are stupid, but I can't stop looking at them and changing the colors. Case supports three more it seems. Thinking now, I should've got the RGB ram....or not.

Anyways, I was kind of freaked out about getting Threadripper set in it's LGA slot since I saw HardOCP's video with the bent pins. That part was actually pretty darn easy to do without any issue. The problem was that afterwards the three screws on the slot cover wouldn't thread through. I would crank and crank and they wouldn't catch. 2 Pints of Belgian Strong (Alesmith Horny Devil) later, and putting a lot of pressure on the CPU cover and the board itself, I got it to thread through. However then it wouldn't be aligned to the other two screws. I had to just get it to catch slightly on a screw and then go from 1 -> 2 -> 3 and repeat, cranking them like 6 short turns a piece a round till the torx wrench finally pulls back. When it catches it's not a click as I would have thought, but it really is a nice pull saying it's set. Seriously, I thought I might have gotten a board with a defective CPU slot mounting mechanism.

After that it was a comedy of building errors all in some ways case related. Even when I got to the Windows install, since there is no USB 2.0 on the tower nor the back of the ASROCK X399 motherboard, I had to make a new GPT Windows USB install, yah.

Anyways, downloading and setting up my stuff atm. My ram is rated at 3600 mhz, but can't get it passed 2100 mhz. However I really haven't tried to manually mess with it too much yet.

The 1950x is resting mostly at 2195 mhz according to HWiNFO with one or two cores shooting up to 4191.3 mhz once in a while. Average temp from my Corsair H100iV2 is 35.7C at the moment with a max of 52.6C. Offset on the Tctl is at +27C. Seems alright at the moment without full coverage on chip. However I haven't downloaded anything that'll truly stress it yet.

I just finished putting together my 1950x build last night as well. I was able to get my RAM to 3066 with a 8x4GB setup, but no matter what I do I can't get it to run at 3200 like it's rated at.

I haven't noticed my cores jumping up to 4200 yet, highest I've seen is 4100. Did you change anything in your bios or is it just running at stock? Is there something I need to do to allow XFR to kick it up to 4.2?

I do agree that the slot was interesting, mine made some interesting sounds that I was sure meant I had damaged my CPU. Thankfully I didn't!

I'm running at Thermaltake Water 3.0 Ultimate 360mm, and the highest temps I've seen are 68C while running Prime95 Small FFT test for 4 hours. Those temps seem pretty good to me so I'm happy with where I'm at so far, just wish I could figure out how to get the CPU to boost higher.

Edit: Just a few edits...

I realized my fans were refusing to run at 100%, which was causing my CPU performance to suffer and the chip to run hotter than it should. Fixed the issue and now I can OC to 4.0 on 1.3125 volts and it's stable. Max temps I'm seeing during stress testing is about 70C.

If I don't overclock I'm still not seeing the performance I was seeing from reviews in Cinebench. Multi-core is about right, but single-core is only getting about 160 and I was seeing 166 or so in reviews. I noticed that only the cores that weren't running Cinebench were boosting during the test, while the cores Cinebench was using were staying at 3.7GHz. Not sure why that would be.
 

kotodama

Member
I just finished putting together my 1950x build last night as well. I was able to get my RAM to 3066 with a 8x4GB setup, but no matter what I do I can't get it to run at 3200 like it's rated at.

I haven't noticed my cores jumping up to 4200 yet, highest I've seen is 4100. Did you change anything in your bios or is it just running at stock? Is there something I need to do to allow XFR to kick it up to 4.2?

I do agree that the slot was interesting, mine made some interesting sounds that I was sure meant I had damaged my CPU. Thankfully I didn't!

I'm running at Thermaltake Water 3.0 Ultimate 360mm, and the highest temps I've seen are 68C while running Prime95 Small FFT test for 4 hours. Those temps seem pretty good to me so I'm happy with where I'm at so far, just wish I could figure out how to get the CPU to boost higher.

Edit: Just a few edits...

I realized my fans were refusing to run at 100%, which was causing my CPU performance to suffer and the chip to run hotter than it should. Fixed the issue and now I can OC to 4.0 on 1.3125 volts and it's stable. Max temps I'm seeing during stress testing is about 70C.

If I don't overclock I'm still not seeing the performance I was seeing from reviews in Cinebench. Multi-core is about right, but single-core is only getting about 160 and I was seeing 166 or so in reviews. I noticed that only the cores that weren't running Cinebench were boosting during the test, while the cores Cinebench was using were staying at 3.7GHz. Not sure why that would be.

Congrats on the Threadripper build!

So jealous of that RAM speed. I've spent the better part of a morning trying to get my ram above 2133 mhz. I've been messing wiht XMP Profiles, over voltage and manually setting the timing, but all I get is crashing at startup. Thankfully, the board dies gracefully and at least 3 resets later we're back to stock. My kit is the F4-3600C17Q-64GTZ, which was on the QVL list, but on that list it looks like Asrock is saying it was a 4 pack of 4GB per stick ram that has been qualified at 3466mhz..which it is anything but. Samsung dual rank B-dies at 16GBs a piece. Don't know if I can really complain at 2133, as the pre AGESA 1.0.0.6. specs say I should be getting 1833 max out of this configuration per regular AM4. I'm wondering if I should return these and get the 128GB G.Skills kit at 2933 when that comes out instead or wait for some AGESA update for Threadrippper. I wonder if an ASUS board would have been better.

As per the processor itself, I'm running it stock normal cash money. Memory has been my focus as for now. I'm giving it a run of Prime right now, but Cores 0, 1, 2 and 8 can hit 4.2 Ghz (4.191.3) all day long it seems, so I guess I lucked out on that aspect.....(Some time later...) After an hour an a half, I'm getting 3.7 Ghz on all cores at 59C. There was a jump to 66.5 on the Tide, and supposedly hitting a max of 93.5C on the Tctl, but looks like all offset +27 non-sense. Perhpas a 240mm radiator is fine for a regular Threadripper operations.

I haven't run Cinebench, nor did I want to, but I'll give it a go later.

Honestly though, nothing that I run at this moment, that isn't synthetic, would stress this system at all, it's ridiculous.

Edit: 3 Hours and 15 mins. later temp hit an average of 62.2C on Prime. Cinebench was at 1454 with Prime running, because why not stress it out more. With just Firefox running etc, got 2806.
 
Congrats on the Threadripper build!

So jealous of that RAM speed. I've spent the better part of a morning trying to get my ram above 2133 mhz. I've been messing wiht XMP Profiles, over voltage and manually setting the timing, but all I get is crashing at startup. Thankfully, the board dies gracefully and at least 3 resets later we're back to stock. My kit is the F4-3600C17Q-64GTZ, which was on the QVL list, but on that list it looks like Asrock is saying it was a 4 pack of 4GB per stick ram that has been qualified at 3466mhz..which it is anything but. Samsung dual rank B-dies at 16GBs a piece. Don't know if I can really complain at 2133, as the pre AGESA 1.0.0.6. specs say I should be getting 1833 max out of this configuration per regular AM4. I'm wondering if I should return these and get the 128GB G.Skills kit at 2933 when that comes out instead or wait for some AGESA update for Threadrippper. I wonder if an ASUS board would have been better.

As per the processor itself, I'm running it stock normal cash money. Memory has been my focus as for now. I'm giving it a run of Prime right now, but Cores 0, 1, 2 and 8 can hit 4.2 Ghz (4.191.3) all day long it seems, so I guess I lucked out on that aspect.....(Some time later...) After an hour an a half, I'm getting 3.7 Ghz on all cores at 59C. There was a jump to 66.5 on the Tide, and supposedly hitting a max of 93.5C on the Tctl, but looks like all offset +27 non-sense. Perhpas a 240mm radiator is fine for a regular Threadripper operations.

I haven't run Cinebench, nor did I want to, but I'll give it a go later.

Honestly though, nothing that I run at this moment, that isn't synthetic, would stress this system at all, it's ridiculous.

Edit: 3 Hours and 15 mins. later temp hit an average of 62.2C on Prime. Cinebench was at 1454 with Prime running, because why not stress it out more. With just Firefox running etc, got 2806.

I'm running the MSI Gaming Pro Carbon and I'm using an old set of G-Skill Ripjaws 4 3200, so I'm actually surprised it's running at the speed it is considering all the reports I've heard of memory compatibility issues.

I was initially having issues with getting my RAM to run any higher than 2400, but I set the voltage on the RAM to 1.4 and it's been stable ever since.

I think I'm just going to stick with a 4GHz 24/7 OC. I ended up having to raise the voltage to 1.35, but haven't seen any issues since and the temps stay under control with my cooler.
 
Hey gaf, I just put together a Ryzen build and I'm getting 55-60 C in my bios -- is this normal? Should I buy another fan? I'm currently only using the stock fan, but the Ryzen PC guide on gaf says the stock cooler should be fine.
 

Timu

Member
Hey gaf, I just put together a Ryzen build and I'm getting 55-60 C in my bios -- is this normal? Should I buy another fan? I'm currently only using the stock fan, but the Ryzen PC guide on gaf says the stock cooler should be fine.
What's the idle temps when browsing on the desktop?
 
Hey gaf, I just put together a Ryzen build and I'm getting 55-60 C in my bios -- is this normal? Should I buy another fan? I'm currently only using the stock fan, but the Ryzen PC guide on gaf says the stock cooler should be fine.

Ryzen has a temperature offset of 20 degrees. Deduct 20 and you'll have your real temp. Install Ryzen Master and compare it to the motherboard temp and you'll see.
 
Hey gaf, I just put together a Ryzen build and I'm getting 55-60 C in my bios -- is this normal? Should I buy another fan? I'm currently only using the stock fan, but the Ryzen PC guide on gaf says the stock cooler should be fine.

I haven't been able to get my 1700 above 50 C yet with the stock wraith spire.
 
I updated the BIOS on my MSI B350M Gaming Pro to the most recent full release; it promised faster post times, so I figured, sure, why not. Now, I'm getting an issue with my overclock. If I try to manually set VCore to 1.275V and CPU frequency to 3700 MHz (which worked wonderfully and was stable even in extended Prime95 tests on my old BIOS), it just sticks the VCore to 1.275V but my CPU frequency stays at 3.0 GHz in Ryzen Master. However, I can get 3700 MHz if I set the VCore to "Auto" - it seems to idle at 1.1875V, and when I put a small load on it (60% usage or so by messing around in a video editing software), the voltage stays static at 1.1875V. HWInfo, CPU-Z, and Ryzen Master all show the same voltage and clock speed. Throughout all this testing, I reverted my RAM to stock 2133 MHz/1.25V rather than the 2933 Mhz/1.35V I had prior to the BIOS update; I plan to deal with that after I figure out my CPU clock issue.

Anyone have a thought as to what I can do? Is this just a bug in the newest BIOS? Maybe auto voltage is better anyway? I'm just not sure that I trust that a 1.1875V can sustain a 3.7 GHz overclock.

Edit: I disabled the Core C6 State option in the BIOS (it was set at Auto before this BIOS update and after the update) and now I'm at 3.7 GHz and 1.25V no problem. The Core C6 State I think is meant to downclock/downvolt at idle, but it never worked for me anyway, so I see no difference or harm in turning it off.
Edit 2: Got my RAM back to 2933 MHz and 1.35V. Did a few runs through Cinebench without an issue, but now I feel like I need to re-test it all in Prime95. Will do that tonight.
 
Hey gaf, I just put together a Ryzen build and I'm getting 55-60 C in my bios -- is this normal? Should I buy another fan? I'm currently only using the stock fan, but the Ryzen PC guide on gaf says the stock cooler should be fine.

Which Ryzen CPU do you have and what's the temperature idling in Windows? My 1600 is idling at sth. like 45C in Windows. ~75C is the max. I've seen so far during load (Prime95, OCd to 3.6 Ghz).
 
Just an update: When I use Ryzen Master Utility to OC my 1950x, I can run 4.0GHz with no issues at 1.35v. I got tired of having to open the utility and apply the OC every time though so I tried with the BIOS and I couldn't manage to get a stable OC with the same settings. I even tried all the way up to 1.375v but the temperature was too hard to keep under control in stress tests. I haven't been able to test as long as I'd like, but has anyone else seen something similar??
 

kotodama

Member
Just an update: When I use Ryzen Master Utility to OC my 1950x, I can run 4.0GHz with no issues at 1.35v. I got tired of having to open the utility and apply the OC every time though so I tried with the BIOS and I couldn't manage to get a stable OC with the same settings. I even tried all the way up to 1.375v but the temperature was too hard to keep under control in stress tests. I haven't been able to test as long as I'd like, but has anyone else seen something similar??

Did you try upping the SOC voltage to say 1.1?

--

Update myself: Got mine to run at 4GHz at 1.3v, but I didn't do any real stress testing on it yet, because I found that HWiNFO wasn't reporting one of my sticks of Ram's stats. So, I went back to messing with that. I checked the system dialog and Windows was detecting 64GB, but only using 48GB. I reseated the ram and I was able to finally get my ram to run at 3600MHz.. for 5 mins before it started freaking out. Went back to 2133Mhz and it was stable again, but HWiNFO was still not reporting the info on that last stick. However Windows claimed to be using the full 64GB. Memtest at 2133Mhz was fine. At 3600Mhz, it started spitting errors during the test run. Right now, giving it a 3466Mhz run.
 

cyen

Member
Did you try upping the SOC voltage to say 1.1?

--

Update myself: Got mine to run at 4GHz at 1.3v, but I didn't do any real stress testing on it yet, because I found that HWiNFO wasn't reporting one of my sticks of Ram's stats. So, I went back to messing with that. I checked the system dialog and Windows was detecting 64GB, but only using 48GB. I reseated the ram and I was able to finally get my ram to run at 3600MHz.. for 5 mins before it started freaking out. Went back to 2133Mhz and it was stable again, but HWiNFO was still not reporting the info on that last stick. However Windows claimed to be using the full 64GB. Memtest at 2133Mhz was fine. At 3600Mhz, it started spitting errors during the test run. Right now, giving it a 3466Mhz run.

So jelous even though i would not have pratical use for TR, would love to just toy around with the cpu.
 
Did you try upping the SOC voltage to say 1.1?

--

Update myself: Got mine to run at 4GHz at 1.3v, but I didn't do any real stress testing on it yet, because I found that HWiNFO wasn't reporting one of my sticks of Ram's stats. So, I went back to messing with that. I checked the system dialog and Windows was detecting 64GB, but only using 48GB. I reseated the ram and I was able to finally get my ram to run at 3600MHz.. for 5 mins before it started freaking out. Went back to 2133Mhz and it was stable again, but HWiNFO was still not reporting the info on that last stick. However Windows claimed to be using the full 64GB. Memtest at 2133Mhz was fine. At 3600Mhz, it started spitting errors during the test run. Right now, giving it a 3466Mhz run.

Not yet, like I said I didn't get much of a chance to play around with voltages other than CPU and DRAM. I'll try changing the SOC voltage and see what happens.

I also still need to run memtest, I've been focused on trying to get my CPU OC'd first lol... I probably should have done that first but oh well
 

kotodama

Member
So jelous even though i would not have pratical use for TR, would love to just toy around with the cpu.

Yeah, I definitely would have been fine with a regular Ryzen 1700.

Not yet, like I said I didn't get much of a chance to play around with voltages other than CPU and DRAM. I'll try changing the SOC voltage and see what happens.

I also still need to run memtest, I've been focused on trying to get my CPU OC'd first lol... I probably should have done that first but oh well

Indeed. We're attacking this thing from different angles lol. Anyways, I stabilized my ram at 3333Mhz so far. Almost got 3466Mhz, and possible could with a voltage of 1.4, but 3333Mhz is fine for Samsung dual rank I reckon atm. Especially, since it seems that all reviews were doing 3200Mhz with single rank.

As for OC'ing my 1950x, I'm going to keep it at stock for now, since my 240mm radiator isn't sufficient for all core stress at 4Ghz. Prime95 crashed it in 10 minutes. I could probably do 3.7Ghz, but it's really touching that 68C recommended max. Anyways, stock speed is fine, but it'll be fun to get one of those Enermax liqTech TR4 AIO LCS when they come out and give this another go.
 

Freshmaker

I am Korean.
Why does MSI's BIOS cripple OC'ing? Set any kind of OC, and it's instantly locked to half speed.

Only way to get the 1700 back to 3.0ghz is to reflash the BIOS.
 
What's the idle temps when browsing on the desktop?

Ryzen has a temperature offset of 20 degrees. Deduct 20 and you'll have your real temp. Install Ryzen Master and compare it to the motherboard temp and you'll see.

I haven't been able to get my 1700 above 50 C yet with the stock wraith spire.

Which Ryzen CPU do you have and what's the temperature idling in Windows? My 1600 is idling at sth. like 45C in Windows. ~75C is the max. I've seen so far during load (Prime95, OCd to 3.6 Ghz).

Thanks for the responses! It's a Ryzen 5 1600. That temp was in the BIOS before I had Windows installed. It looks good now from 30-45 C when idling in Windows.

Just worried about my first build going ok :p
 
*As a general heads up to those unaware or less experienced, you should always clear your BIOS/UEFI settings before flashing to a newer version. This isn't always needed and some motherboard update methods make this redundant, but not all boards or update processes are created equal so it can potentially save from future headaches in the event your particular board is susceptible to issues stemming from this step.*



Why does MSI's BIOS cripple OC'ing? Set any kind of OC, and it's instantly locked to half speed.

Only way to get the 1700 back to 3.0ghz is to reflash the BIOS.
Have you updated to the latest BIOS/UEFI for your motherboard and checked whether other related settings are on auto or manual?

If the issue persists despite an update and after closely looking through your settings, you may need to revert to a previous stable BIOS.


A few search results:

https://forum-en.msi.com/index.php?topic=286410.0
http://www.tomshardware.co.uk/answers/id-3399319/ryzen-1600x-stuck-2ghz-unable-reach-6ghz.html
http://www.tomshardware.com/answers/id-3473892/amd-ryzen-1500x-running-low-core-clock-speed.html
https://linustechtips.com/main/topi...-to-37-ghz-results-as-underclock-to-1550-mhz/
https://www.reddit.com/r/overclocking/comments/6hp3el/cpu_clocks_half_speed_inside_windows_10/

Slash_DK

Check your Windows Power Plan. There is an option to restrict the clockspeed to a percentage of the maximum. Also, restore BIOS to defaults and check if the issue persists.
junglePanther_gb72:

Thanks for the reply. Last night I restored the bios to complete default and it booted at the proper speed. The power settings inside windows is at High Performance with the Minimum processor state set to 100%. I had turned off the computer overnight and this morning it seems to still be holding the settings as the computer booted inside windows with the proper CPU clock speed (not overclocked, since I restored the bios to default). I am gonna turn it off again for another 4 hours and turn it back on to see if it still holds the proper clock speed, and if it does I will try to push the processor only to 3.7 and if that holds I will try the memory to 2600 again to see if thats the real cause of the issue.

Again, thanks for the replay and I will keep you posted on the issue
GreenHornet188

I have the exact same problem. I found that you need to make sure that the vcore voltage is not on auto.

When I updated my bios it reset it and I had the 1599mhz issue.

Btw I have the same CPU, motherboard and ram


Anyways, stock speed is fine, but it'll be fun to get one of those Enermax liqTech TR4 AIO LCS when they come out and give this another go.
There have been several new TR4 adapters and full-coverage water blocks released or nearing release. I'll post when I have a moment.
 

nubbe

Member
I replaced my 3930k system with a Ryzen

Ryzen 1700X
Asus Strix B350-F
Corsair CMK32GX4M2D3200C16 (Bought it before the launch, SAD!)
Noctua NH-D15

The 1700x is running at 3925MHz with VCore offset +0.02500, if I try to go for 4GHz I need to pump in a ridicules amount of VCore power.

The CPU goes just a few degrees Celsius over 50C (-20C) when under full Prime95 load. So it is way cooler than the 3930k.

I get the RAM to run solid at 2666MHz 14-13-13-28, 2933MHz is not happening no matter the latencies.

On one hand it is annoying I got the RAM before launch, on the other hand it is €100 more expensive now and I could sell for a tiny profit :p
 

DieH@rd

Banned
My Ryzen 1600 and Asus Strix B350-F have just arrived. :)

Are there any promotions for free games for these products? [they were both purcahsed in Hungary]
 

Freshmaker

I am Korean.
·feist·;246398252 said:
Have you updated to the latest BIOS/UEFI for your motherboard and checked whether other related settings are on auto or manual?

Yep. MSI said they're aware of the problem and it's AMD's fault.
 

DieH@rd

Banned
Hooooooly shit, PUBG is now a totally different game. The game FLIES on Ryzen, while on FX8350 and 8gb of ram was a bloody stuttering mess. Before I though that my lowly RX460 was limiting me, but now, I see that I was mistaken.

/hugz Ryzen

Now... if I could only get a good deal on GPU. But I don't have a lot of hope, miners have snatched everything in my territory.
 

gt86

Member
I replaced my 3930k system with a Ryzen

Ryzen 1700X
Asus Strix B350-F
Corsair CMK32GX4M2D3200C16 (Bought it before the launch, SAD!)
Noctua NH-D15

The 1700x is running at 3925MHz with VCore offset +0.02500, if I try to go for 4GHz I need to pump in a ridicules amount of VCore power.

The CPU goes just a few degrees Celsius over 50C (-20C) when under full Prime95 load. So it is way cooler than the 3930k.

I get the RAM to run solid at 2666MHz 14-13-13-28, 2933MHz is not happening no matter the latencies.

On one hand it is annoying I got the RAM before launch, on the other hand it is €100 more expensive now and I could sell for a tiny profit :p

Interesting, I have the 1700 and was disappointed that I can only get to 3.8ghz stable, 3.9ghz works for the most part with a rare shutdown. Oh well plenty fast enough. Luckily my ram does run at 2933
 
Yep. MSI said they're aware of the problem and it's AMD's fault.
Which board do you have? I haven't experienced this with my MSI X370 Gaming Pro Carbon.


Have you tried using Ryzen Master to overclock instead of going through BIOS as a workaround?
Additionally, if it is in fact due to AMD shouldn't this issue be more prevalent across a wide range of motherboards from each manufacturer?



If anyone is aware of this being the case (downclock issue being more widespread), please feel free to share your experiences or post sources from other users and manufacturers covering this.

This thread is not a 24/7 Ryzen lovefest. I've posted positive and negative items alike, including first-hand experiences of my own, as people should be as informed of as many positives and negatives as there may be in relation to these CPUs and associated platforms.



Freshmaker,
Have you gone through the troubleshooting mentioned in the links I posted or any others you yourself may have come across? What has MSI suggested as a solution?





Hooooooly shit, PUBG is now a totally different game. The game FLIES on Ryzen, while on FX8350 and 8gb of ram was a bloody stuttering mess. Before I though that my lowly RX460 was limiting me, but now, I see that I was mistaken.

/hugz Ryzen

Now... if I could only get a good deal on GPU. But I don't have a lot of hope, miners have snatched everything in my territory.
Congrats on the build. PUBG already performed fairly well one Ryzen despite the state of optimisation from the dev. and other issues.

They've recently improved CPU usage further.


PlayerUnknown's Battlegrounds gets 6 core-plus CPU optimization
https://venturebeat.com/2017/08/09/playerunknowns-battlegrounds-gets-6-core-plus-cpu-optimization/
https://steamcommunity.com/games/578080/announcements/detail/1451701826007887564

Getting support for multicore CPUs could make a difference for a game like PlayerUnknown's Battlegrounds due to its persistent nature. It has up to 100 players all running around a map at the same time, and that puts a lot of processing on the CPU as opposed to the graphics card. We'll have to see if this new optimization makes a difference, but it is something that theoretically could improve performance for anyone rocking a chip with more than 4 cores.

Players,

Today we look at what we have changed and fixed in our week 20 patch.
This patch will be deployed to the Test Servers on Wednesday, August 9th, and if everything is stable, to the Live Servers on Thursday.

Client Optimization

  • Optimized game performance for 6-core or higher CPU
  • Optimized UI
 

Freshmaker

I am Korean.
Which board do you have? I haven't experienced this with my MSI X370 Gaming Pro Carbon.


Have you tried using Ryzen Master to overclock instead of going through BIOS as a workaround?

It's a MSI B350M Bazooka.

Ryzen Master reacts exactly the same. Only the MSI commander software will work, but I have to reapply that every time I boot up if I want to use that which is kinda annoying.

No problems getting my RAM to play nice at 3200 ironically.

feist, up til I gave up, I had updated the BIOS to the latest, and I had manually set the voltage. Instant lock to half speed. Only thing I didn't do was adjust the windows power profile. So I'll try that next.

MSI has suggested reflashing the BIOS from base settings 3x in a row. Which sounds like witchcraft so I've been content to just leave it at stock. My searches on the issue tend to all revolve around MSI motherboards too so if it actually is an AMD issue, it's remarkably constrained. My main plan is just to wait for a new BIOS update and go from there.

Edit: After tinkering for an hour it appears that it'll lock to half speed if you input a manual voltage value it thinks is too high. Managed to get 3.6ghz. Anything higher and finding the perfect voltage target seems super finicky. Couldn't find a value that 3.7 liked at all, and I got it to boot into 3.8, but it immediately crashed when I tried to run Cinebench.
 
Why does MSI's BIOS cripple OC'ing? Set any kind of OC, and it's instantly locked to half speed.

Only way to get the 1700 back to 3.0ghz is to reflash the BIOS.

I had the same problem on my Mortar Arctic. I searched for answer and tried everything but the problem never fully went away (some boots would be OC most would be UC). I ended just returning the board for an Asrock itx board and couldn't be happier. First mobo I can hit 3200mhz on my 3000mhz lpx kit.
 
I had the same problem on my Mortar Arctic. I searched for answer and tried everything but the problem never fully went away (some boots would be OC most would be UC). I ended just returning the board for an Asrock it board and couldn't be happier. First mobo I can hit 3200mhz on my why 3000mhz lpx kit.

Yeah, I'm convinced MSI shit the bed with its AM4 B350 motherboards. I've scoured the internet and found numerous obscure problems with the boards related to overclocking CPU and/or RAM that I don't see with other brands of mobos. If I didn't get my board for $16 (<3 Microcenter open box) I'd be pissed.
 
Hooooooly shit, PUBG is now a totally different game. The game FLIES on Ryzen, while on FX8350 and 8gb of ram was a bloody stuttering mess. Before I though that my lowly RX460 was limiting me, but now, I see that I was mistaken.

define "flying", 1080p60? I've only run PUBG on my 1700 @ stock, figure I should try it OC'd but my 1080's still at gigabyte's RMA center, so stuck with my 390 for now :(
 

saunderez

Member
define "flying", 1080p60? I've only run PUBG on my 1700 @ stock, figure I should try it OC'd but my 1080's still at gigabyte's RMA center, so stuck with my 390 for now :(
Don't think I had a frame counter up last I played and the multicore optimisations hadn't been done but 1080p60 wasn't too far off then and I only have a 290. I was surprised by the games performance to be honest.Im running my 1700 at 3.6. It's an effortless OC that low you don't even have to mess with voltages.
 
Don't think I had a frame counter up last I played and the multicore optimisations hadn't been done but 1080p60 wasn't too far off then and I only have a 290. I was surprised by the games performance to be honest.Im running my 1700 at 3.6. It's an effortless OC that low you don't even have to mess with voltages.

Yeah I finally got to play some PUBG with the 3.7 OC and its definitely an improved experience. My 390 can drive ~1080p but boy does it get HOT (93C lol)
 

tuxfool

Banned
Yeah, I'm convinced MSI shit the bed with its AM4 B350 motherboards. I've scoured the internet and found numerous obscure problems with the boards related to overclocking CPU and/or RAM that I don't see with other brands of mobos. If I didn't get my board for $16 (<3 Microcenter open box) I'd be pissed.

wow, that is quite the deal.
 

Mrbob

Member
Sorry if this has been posted already but will future Ryzen 2, 3, and 4 cpus be compatible with current x370 or B350 motherboards?
 
Top Bottom