FEATURED AMD ZEN Discussion (Previous Rumor Thread)

Nagork, beautiful data! This will be very helpful for all of us and understanding the needs of the CPU as we increase frequency.

Bluefalcon13 said:

No, but a minor increase in ambient would've driven it to instability if folding or some other high intensity task. It's kinda hard to burn a block of aluminum without burning everything else

Mind you this is all cores overclocked to 3.7ghz. Not turbo when thermal headroom exists.

Click to expand...

Was not thinking the block of aluminum would burn up, it was the fan motor that I thought you were talking about.

Microcenter has put back the $30 combo discount with Ryzen CPUs.

Also, I noticed that ASRock finally pulled the B350 1.43 BIOS I told them was faulty last week. No replacement, just gone. Since that one added the core voltage selection, there's no way to set it on either of the ASRock B350 motherboards. While I know AMD screwed all the motherboard companies by not sharing info or providing CPU samples early enough, some of what they're doing is just plain stupid.

wingman99 said:

Was not thinking the block of aluminum would burn up, it was the fan motor that I thought you were talking about.

Click to expand...

I figured you weren't talking about the aluminum [emoji14]

I'm not worried about fans. Those are cheap, in comparison to a $330 CPU.

IDK if any one else as seen them ( i will look for them ) but i watched to vids where they benched ryzen and 7700k with lower gfx cards . the ryzen was much closer with say a 1060 to intel ( beating it in a few ) and most of the time the avg was lower but the min frames were higher. with a 1070 things were much closer than with the 1080ti IDK this is but to me this just points out more and more that a R5 will make a amazing budget gamer.

http://wccftech.com/retailers-selling-amd-ryzen-5-cpus-3-weeks-official-launch/


Welp, Ryzen 5 benchmarks are incoming sooner rather than later

||Console|| said:

IDK if any one else as seen them ( i will look for them ) but i watched to vids where they benched ryzen and 7700k with lower gfx cards . the ryzen was much closer with say a 1060 to intel ( beating it in a few ) and most of the time the avg was lower but the min frames were higher. with a 1070 things were much closer than with the 1080ti IDK this is but to me this just points out more and more that a R5 will make a amazing budget gamer.

Click to expand...

It was techspot. http://www.techspot.com/review/1360-amd-ryzen-5-1600x-1500x-gaming/

Shows that the (6?) games they tested are GPU bottlenecked with a 1060 and even 1070.

Okay, after ASRock took down the B350 1.43 BIOS with the faulty core voltage setting, they now released a new 1.44 BIOS with, you guessed it, a core voltage setting. Since I returned my board, as ASRock suggested I do, I can't say whether it now works or not.

A little something for shrimpbrine ....... if you have not already seen this .

Ryzen deLidding
http://wccftech.com/amd-ryzen-delid...p-tim/?utm_source=wccftech&utm_medium=related

And for you Bluefalcon ...... Ryzen 4 benchmarks.......

http://wccftech.com/amd-ryzen-4-cor...7700k/?utm_source=wccftech&utm_medium=related

I got sick of the issues with Asus not cold booting properly with faster memory settings. I updated to the latest BIOS, adjusted SOC voltage, and it still didn't work. I'm done with that board. It's going back. Unfortunately, now AI Suite 3 won't uninstall, go figure.

I'm also done with the Asrock B350 board that wouldn't boot properly. I could try it again, but I just don't have it in me to waste anymore time on it. It's just going back.

I went back to the Gigabyte Gaming 3. Same issue with shutting down when running at 3.9 GHz. I noticed that there's no LLC setting in the BIOS, and that voltage runs lower than the requested voltage under load, so I think that could be the issue. The Asus was always running above requested voltage, so the auto LLC setting must be fairly high on that board. For about a 3.7 GHz overclock it seems fairly stable, but higher than that is a no go. So, for anyone who wants to do only light overclocking it's probably an OK board. At least my memory actually runs properly on it.

I'm on the fence as to whether to keep the Gaming 3 or not. If it's stable at 3.7 GHz that's probably enough for my R7 1700 chip, since that's the most efficient overclock setting anyway. I still have the second Gaming 3, which I'll probably test to see if I have the same rebooting problem at 3.9 GHz. I'm just a bit fed up with this whole process right now.

Yesterday I did manage to pick up two Asrock Fatal1ty Gaming Pro motherboards from Micro Center, so I'll probably focus on those next. It was a three hour round trip to get them, so I hope they work a bit better than the other boards I've had so far. It was crazy to drive that far, but I was desperate with the poor motherboard stock that is available.

I had no issues at 3.9GHz on my 1700X and I was testing it on ASUS X370-Pro and Gigabyte AB350-Gaming 3. If it shuts down, screen goes black or something like that then CPU isn't stable. In most cases voltage is too low. SOC voltage is not really helping in memory clock. On ASUS DDR VTT is for memory clock but on Gigabyte there is no voltage for that. Still on Gigabyte I could run memory at 3200 only by setting XMP or manually 3200. On ASUS 3200 couldn't work stable without bumping VTT ( and VDIMM as VTT had to be higher than the limit at 1.35V VDIMM ).
I just don't think it's motherboard's fault but some settings, cooling or just bad luck to CPU.

The board definitely plays some sort of role because it's stable on the one but not the other. I'm also using the same memory and heatsink on both. It's kind of a pick your poison situation because the Gigabyte works better with higher memory settings, but the Asus is more stable at higher CPU frequency. I will try increasing the voltage even further on the Gaming 3 to make up for it not having an LLC setting. With the Asus it was really running at around 1.411 volts, so I'll see what I need to set to get that under load. It looks like at a setting of 1.425 it actually provides 1.416 under load, so I'll try running at that setting for a bit.

Update: That didn't work at all, it just shut down in less than 2 minutes. How much voltage were you giving your chip? It seems like the problem occurs as I ramp up the voltage. It could be a QC issue even. I'll have to compare to the other Gaming 3 board. First I'll try dropping back to stock on the memory and see if that makes a difference.

I don't know. I have my ram running 2933 (only 3000 ram) and my CPU only cranked to 3.7ghz cause it's rock solid stable. I need to use my computer, not fight with temperature or stability issues. This weekend I'll be writing a program to generate initial stages of a keystream generator (similar to A5/2 gsm keystream generator) for a project in one of my classes. I know the code will work, cause I'm going to modify an open source implementation of it, and make it multi-threading. Then feed the beast for days on end. My only concern is power stability (from the wall outlet).

Either it's a voltage setting that needs to be set in bios, or it's just straight up a ram compatibility issue with the Asus board IMO.

It's possibly a compatibility issue. From my standpoint it really doesn't matter though. I changed all the different voltages, it didn't make any difference. SOC, VTT, DDR voltage were all boosted. The issue wasn't that it wasn't stable, it was simply that it wouldn't do a cold boot. I bet a couple months from now it will be fixed with a BIOS update, but I don't want to deal with it in the interim.

No change with stock memory on the Gaming 3. Now I'll try Flare X and see if it makes a difference (I don't think it's going to.) After that I'll try the other Gaming 3.

Nagorak said:

It's possibly a compatibility issue. From my standpoint it really doesn't matter though. I changed all the different voltages, it didn't make any difference. SOC, VTT, DDR voltage were all boosted. The issue wasn't that it wasn't stable, it was simply that it wouldn't do a cold boot. I bet a couple months from now it will be fixed with a BIOS update, but I don't want to deal with it in the interim.

No change with stock memory on the Gaming 3. Now I'll try Flare X and see if it makes a difference (I don't think it's going to.) After that I'll try the other Gaming 3.

Click to expand...

I was eyeballin those flarex sticks. Which one did you get?

Nagorak said:

The board definitely plays some sort of role because it's stable on the one but not the other. I'm also using the same memory and heatsink on both. It's kind of a pick your poison situation because the Gigabyte works better with higher memory settings, but the Asus is more stable at higher CPU frequency. I will try increasing the voltage even further on the Gaming 3 to make up for it not having an LLC setting. With the Asus it was really running at around 1.411 volts, so I'll see what I need to set to get that under load. It looks like at a setting of 1.425 it actually provides 1.416 under load, so I'll try running at that setting for a bit.

Update: That didn't work at all, it just shut down in less than 2 minutes. How much voltage were you giving your chip? It seems like the problem occurs as I ramp up the voltage. It could be a QC issue even. I'll have to compare to the other Gaming 3 board. First I'll try dropping back to stock on the memory and see if that makes a difference.

Click to expand...

I'm curious if it is a faulty motherboard with the Gaming 3? Can't wait to see how the other board works. Statistically from puget systems 1 out of 18 motherboards sold from the factory are faulty for one reason or another. Most motherboard are not tested from the factory, though Processors and memory are tested before the customer receives them.

I didn't have much more success with the second board. I got it to run maybe a few minutes longer, but the same thing still happened.

However, I did happen to notice that the heat sink on the board was burning hot to the touch (actually burned myself a little bit, LOL). So, I opened the System Information Viewer and watched what happened under load on Prime 95. The VRM MOS temperature just keeps increasing and eventually gets as high as 131 C, before the system shuts down (interestingly, at 128 C the temperature actually wraps around to -128 C, so I guess the programmers never expected such a high value). Anyway, I suspect this is probably the source of the problem: the VRMs are slowly overheating. It would also explain why running at higher voltages causes the problem to happen sooner, since the VRMs have more trouble keeping up.

The discrepancy between Woomack's experience and mine could be because I've been running on an open bench with no air flow to save time. In a well ventilated case with better air flow, and with a better CPU heat sink so that less heat bleeds over from the CPU socket, the situation might look better. Actually, even the Wraith Spire might do better since it blows down on the board, whereas I am using a tower cooler that doesn't cool the board at all.

Bluefalcon13 said:

I was eyeballin those flarex sticks. Which one did you get?

Click to expand...

I got the 3200 MHz version. To be honest, I haven't noticed any real difference between them and the Trident Z so far.

Nagorak said:

I didn't have much more success with the second board. I got it to run maybe a few minutes longer, but the same thing still happened.

However, I did happen to notice that the heat sink on the board was burning hot to the touch (actually burned myself a little bit, LOL). So, I opened the System Information Viewer and watched what happened under load on Prime 95. The VRM MOS temperature just keeps increasing and eventually gets as high as 131 C, before the system shuts down (interestingly, at 128 C the temperature actually wraps around to -128 C, so I guess the programmers never expected such a high value). Anyway, I suspect this is probably the source of the problem: the VRMs are slowly overheating. It would also explain why running at higher voltages causes the problem to happen sooner, since the VRMs have more trouble keeping up.

The discrepancy between Woomack's experience and mine could be because I've been running on an open bench with no air flow to save time. In a well ventilated case with better air flow, and with a better CPU heat sink so that less heat bleeds over from the CPU socket, the situation might look better. Actually, even the Wraith Spire might do better since it blows down on the board, whereas I am using a tower cooler that doesn't cool the board at all.

Click to expand...

Sounds like thermal shutdown for the VRM instead of thermal throttling of the CPU like on the old AM3+ motherboards. To view the VRM temp what is System Information Viewer?

Can you point a house fan on the VRM heatsink to see if it will cool enough to keep going?

Unfortunately, now AI Suite 3 won't uninstall, go figure.

Click to expand...

Revo Uninstaller in Safe Mode.

AI Suite digs in like a Missouri tick. There used to be an official AI Suite uninstaller in the ASUS support section. I don't know if it is still there or even applies to AI Suite 3. I used it dig the AI Suite 1 hooks out of my Windows installation.