3700x 4.3GHz Voltages

agentdomo

Limp Gawd
Joined
Dec 29, 2011
Messages
298
Ran a few tests on a new build I'm working on for a friend and I managed to push my 3700x to 4.3GHz all core @ 1.238v (1.3v bios w/ vdroop). I haven't seen much binning info on the silicon quality of 3700x's and I want to know if this is normal or if my voltage is particularly low.

Don't mind the toasty temps this is an ITX build with a Scythe Big Shuriken 3 in a high 26-27c ambient temp room

3700x 4.3GHz 1.238v.png
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,617
Yea, that voltage is darn good. The latest silicon is really good... I'm assuming this is a new chip right?
 

sirmonkey1985

[H]ard|DCer of the Month - July 2010
Joined
Sep 13, 2008
Messages
22,230
sounds pretty good for a newer chip. the likely hood of getting a similar die used in the XT models are pretty good these days, the XT chips just guarantee you get one of them.

also core vid is closer to your actual voltage, not what the bios reports. i'd mess around with the LLC and see if you can and tighten it up to maybe get 4.4Ghz stable with the 1.3v set if you have the thermal headroom available for at least stability testing it in your configuration.
 

agentdomo

Limp Gawd
Joined
Dec 29, 2011
Messages
298
Yea, that voltage is darn good. The latest silicon is really good... I'm assuming this is a new chip right?

Yeah bought it a month ago. Looks like AMD has really gotten the manufacturing process together then.

I just hit auto overclock ..


Auto OC for me seems to overvolt the chip. My ITX cooler can't really handle that TDP so I locked the voltage and tried to see how high I could go.
 

agentdomo

Limp Gawd
Joined
Dec 29, 2011
Messages
298
sounds pretty good for a newer chip. the likely hood of getting a similar die used in the XT models are pretty good these days, the XT chips just guarantee you get one of them.

also core vid is closer to your actual voltage, not what the bios reports. i'd mess around with the LLC and see if you can and tighten it up to maybe get 4.4Ghz stable with the 1.3v set if you have the thermal headroom available for at least stability testing it in your configuration.

I can give it a shot right now since the suns down (in CA right now). My understanding was VID is what the processor requests and "VR OUT" and "SVI2 TFN" is what is actually being supplied to the CPU.
 

agentdomo

Limp Gawd
Joined
Dec 29, 2011
Messages
298
4.4Ghz hit on 1.25v (1.3125v bios) hitting a toasty 90c in r15. I made a single pass of r15 on 1.238v (1.3v bios) but crashed subsequently after. Not sure if this is a 24/7 passable and I wont be testing it unfortunately since I'm pretty thermally constrained at this point. Now I kinda don't want to sell this build to my buddy and just take it back to NYC with me to put under water haha.

3700x 4.4GHz 1.25v.png
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,617
Yeah bought it a month ago. Looks like AMD has really gotten the manufacturing process together then.



Auto OC for me seems to overvolt the chip. My ITX cooler can't really handle that TDP so I locked the voltage and tried to see how high I could go.

Did you run the PBO test for the specific target frequency to get your max safe FIT voltage? If not I suggest ya do so asap.
 

crazycrave

[H]ard|Gawd
Joined
Mar 31, 2016
Messages
1,236
Yeah bought it a month ago. Looks like AMD has really gotten the manufacturing process together then.



Auto OC for me seems to overvolt the chip. My ITX cooler can't really handle that TDP so I locked the voltage and tried to see how high I could go.

My 3700x was from Aug 2019 and it's on a x470 board with the stock cooler and have never tried better cooler yet . As you can see it's pulling around 47 watts for that game as my Bios is 6 - 16 - 2020 = very fresh from MSI
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,617
1.26-1.29v fluctuating during a few r20 passes

Fyi, Cinebench in not a very hard test so it won't be accurate for all situations. Remeber voltage scales up or down depending on load. If you really want to find the safest voltage you should run something much harder like P95, etc.
 

agentdomo

Limp Gawd
Joined
Dec 29, 2011
Messages
298
Fyi, Cinebench in not a very hard test so it won't be accurate for all situations. Remeber voltage scales up or down depending on load. If you really want to find the safest voltage you should run something much harder like P95, etc.

I get the same voltages in R20 as I do small FFT P95. R20 uses AVX as well. I just don't want to run any more P95 passes until I get my higher rpm fan for the cpu cooler to see if I can tame the temps.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,617
I get the same voltages in R20 as I do small FFT P95. R20 uses AVX as well. I just don't want to run any more P95 passes until I get my higher rpm fan for the cpu cooler to see if I can tame the temps.

Do what you do, but on the point of R20, it is no where near P95 on a load stress basis. R20 is stupid easy to run and is no where near a reality based load tester.
 

agentdomo

Limp Gawd
Joined
Dec 29, 2011
Messages
298
Do what you do, but on the point of R20, it is no where near P95 on a load stress basis. R20 is stupid easy to run and is no where near a reality based load tester.

Totally aware of that but once again I'm not touting a 24/7 stable OC as I know for a fact that this isn't whether due to thermal limitations or voltages. P95 is great but since I will never see an all thread AVX load of that nature I don't see the point in running a small FFT test for hours and letting my CPU cook.
 
Top