Is voltage typical for TR owners?

G

gerbiaNem

2[H]4U
Joined
Mar 6, 2005
Messages
2,169
I'm looking at the EK review and seeing 1.43 vcore. I have mine set at 1.275 (shows 1.284 cpu-z) for the same overclock 100% stable. I haven't even bothered trying to push it lower (just set it and forgot it). Any owners want to chime in if this is normal? Does [H] just have an early fab unit?

https://image.************/cGP75w/Capture.jpg
 
Looks like you got a good chip. Have you run any stress tests to fully load it at that overclock? What are you cooling it with?
 
Prime95 stable for almost a day. I'm using an h115i.

If I had more time, I would push it to the limit.
 
I run my TR at stock. It seems that there is no benefit to single threaded performance to have an OC all the time. I could be wrong. OP can you bench using cinebench at stock and then OC 4ghz single threaded?
 
When I get home I will. You're probably right about the single-threaded performance. The boost clock hits 4 on single threaded. I assume the OC will mainly benefit the multi-threaded applications.
 
serbiaNem said:
Prime95 stable for almost a day. I'm using an h115i.

If I had more time, I would push it to the limit.
Click to expand...
Are you matching our 4GHz benchmark scores? I have not seen one go close to that low on vCore at 4GHz.
 
tangoseal said:
I run my TR at stock. It seems that there is no benefit to single threaded performance to have an OC all the time. I could be wrong. OP can you bench using cinebench at stock and then OC 4ghz single threaded?
Click to expand...

Benched over RDP, so the numbers might be a little skewed.

Stock 2900, forgot to screenshot it.

OC at nearly 3400, so all cores are pumping at 4GHz.
KJT9U3r.jpg


And for reference. Single threaded Hyper Pi seems fine too.
bm15HcK.jpg


I'll try lower the voltage as low as it can go first chance I get.
 
Here it is chugging away like a champ at 1.26v. 1.25 ran for about 5 minutes before it crashed, I'm going to let this run for a bit, but looks good so far. I'm thinking about pushing it past 4GHz, but maybe later.
bIQRLRD.png
 
I just did a 1 HR Prime95 on my 1950x @ 4.00GHz with 1.35v. Here are results:

upload_2017-11-6_16-48-52.png


This is using a XSPC block and quad 120mm Koolance radiator, 4x 120mm EKWB Vadar fans, with a D5 pump. Ambient temps are 83F (~28C-29C). I will see if I can get any lower voltage stable and let you guys know what happens. I will say that if you can do 4GHz on an AIO, then you really have an awesome chip!

Update:

The lowest stable voltage I was able to use was 1.325v. I have run Prim95 for over 2 hours now; any lower voltage and I get lockups within ~15 - 30 mins..

Whats rather weird though is I just now noticed that HWINFO is still reporting me as running at 1.35v whereas Ryzen Master is reporting correctly at 1.325v. I have the voltage set manually via UEFI. Can anyone else confirm whether their voltages are being reported correctly? My gut is telling me that CPU-Z and HWIF are correct and that Ryzen Master is wrong. I wonder if there are certain voltage steps that while are technically available, are not actually enforced?

You would think I should be able to somewhat tell based off temp readings, but my ambient temps are about 12F lower now vs. the height of the day, so not an accurate comparison. Here is a screen shot with all the conflicting info:

upload_2017-11-6_20-9-10.png
 
Last edited:
I used the same settings as in my previous post and am now Prime stable at 4050mhz. I did a Gooseberry run for kicks and got a 21:27.20:

upload_2017-11-6_21-34-7.png
 
H2R2P2 said:
Whats rather weird though is I just now noticed that HWINFO is still reporting me as running at 1.35v whereas Ryzen Master is reporting correctly at 1.325v. I have the voltage set manually via UEFI. Can anyone else confirm whether their voltages are being reported correctly? My gut is telling me that CPU-Z and HWIF are correct and that Ryzen Master is wrong. I wonder if there are certain voltage steps that while are technically available, are not actually enforced?

You would think I should be able to somewhat tell based off temp readings, but my ambient temps are about 12F lower now vs. the height of the day, so not an accurate comparison.
Click to expand...

I can confirm that the voltages are off on mine too. Ryzen Master doesn't read the half steps from the looks of things.


Great job on the 4.05GHz. I tried 4.1GHz at the 1.275v but it wasn't prime stable when disk access was added to the mix. Maybe a chipset voltage adjustment would produce better results if it's an IO issue? I know Intel voltages had to be within a certain range while changing them.

I'm extremely impressed with this chip. It's been the easiest 24/7 daily usage overclock experience I've ever had.
 
serbiaNem said:
I can confirm that the voltages are off on mine too. Ryzen Master doesn't read the half steps from the looks of things.


Great job on the 4.05GHz. I tried 4.1GHz at the 1.275v but it wasn't prime stable when disk access was added to the mix. Maybe a chipset voltage adjustment would produce better results if it's an IO issue? I know Intel voltages had to be within a certain range while changing them.

I'm extremely impressed with this chip. It's been the easiest 24/7 daily usage overclock experience I've ever had.
Click to expand...

What motherboard do you have? Also- what type of storage are you using (ex: NVME, U2, traditional SATA)? Reason I am asking about the storage is if you have NVME, you could disable your SATA ports and try your overclock again. That way you could be certain whether its your CPU or your IO that is responsible. What temps were you getting at 4100MHz @ 1.275 before you crash? I have a couple 1950x systems and while I can get 4200+ with extreme measures, its hard to tell if its motherboard or CPU limits.
 
You must log in or register to reply here.
Back
Top