Last edited:
I got this one pic...Build pics guys, come on. Use some of that tech laying around to take a picture of that tech you are building. I want to see cases and RBG and Ryzen boxes laying around all over the place. Pics of benchmark scores and everything. Lets go!
Yeah, it's pretty terrible... and this is *after* I (sorta) cleaned it when installing the new cooler Saturday.That is some epyc dust in that case.
I found a reason to re-enable SMT on my 2700X, fired up 16 threads at 100% and started to find the thermal limits the Wraith Prism -- and the tolerance threshold of my ears!
So, I ordered a Be Quiet Dark Rock 4. I got it installed today and temps dropped by about 6C. I was a little disappointed by the temp drop, but the noise drop was awesome. It's so quiet that I set my fan curves to be much more aggressive -- and traded some of the quiet for a little extra cooling.
It connects to the stock backplate and is dead easy to install on AM4, as tower coolers go. Joker runs through it here: https://www.youtube.com/watch?v=R5_sRqpk1wA&t=318s . It comes with only one fan though and the brackets won't work with any of my spare 120mm's -- and other brackets won't work either -- so I haven't bothered to do a push-pull test yet.
Edit: I've added a second fan. Be Quiet sells their fans for too much ($25 on newegg for 120mm!) so I used a Fractal Design fan that I already had. However, the BQ fans are a bit non-standard -- their mounting holes are on a centered, thin dogear, not a full fan-width dogear like most. I had to rebend the BQ mounting wire brackets to work, but it's snug and tight and dropped my temps by about 3c. I turned my fan speeds back down so it's *reasonably* quiet again at 100% load.
I was hitting mid 60's prior to the fan curve tweaks and adding the 2nd fan (26c ambient). Now I'm down to mid 50's at full load -- but a proper stress test would produce a few degrees above those.What temps were you getting before? I have a Dark Rock Pro 4 on my 2600 and I feel like my temps aren't accurate, but they might be. This is my first AMD build and overall build in years. At idle I'm at approx 35C, and after 30 mins of aida64 stress it hits right at 49-50C. Does that sound right to you guys? These temps are what is displayed in the Ryzen master software. Fans are set on default curve for Gigabyte Gaming 5 Mobo.
I was hitting mid 60's prior to the fan curve tweaks and adding the 2nd fan (26c ambient). Now I'm down to mid 50's at full load -- but a proper stress test would produce a few degrees above those.
Your temps sound fine to me, though you didn't mention OC, voltages, or ambient temps.
If you want even lower temps, there may be several things you can easily do.
Turn up the CPU fans. The Be Quiet fans are very quiet and you'll gain some that way for very little noise.
You can add a 3rd fan which will not just increase airflow though the radiators, but will also increase the hot airflow out of your case since the exhaust vent will be right next to the increased flow coming off the 3rd fan. I aligned my exhaust fan with my heat sink in hopes of gaining slightly from this effect.
Other than that, there may be an easy way to two to improve your case airflow.
Understood. When I added my 2nd fan I first removed the original fan and rechecked the tension on the screws. It didn't make any difference, of course, but I felt better.My CPU is currently stock. I have a Fractal design meshify C with 2x120mm intake fans and one one exhaust. The two intake fans are aligned with the GPU and the CPU cooler respectively. Ambient temp is probably around 22C. I haven't mounted a CPU cooler in over 10+ years, and was afraid that I had messed up with the thermal paste application or securing the heatsink to bracket tight enough. If those temps sound decent for the setup, then I hope to leave it alone! I couldn't find a real frame of reference for a range of what I should be seeing, and was concerned that I had screwed the pooch on mounting the CPU cooler.
Thanks for the reply! This makes me feel a lot better. I won't be so paranoid over my temperatures now.Understood. When I added my 2nd fan I first removed the original fan and rechecked the tension on the screws. It didn't make any difference, of course, but I felt better.
Anyway, you're running quite a bit cooler than I am, as we would expect considering the different CPU's and your larger/split radiators.
I believe there is a +10C offset on the reported temp in the 2700X only.OK< I have my Taichi up and running. But something does not add up. I think the BIOS is misreporting temps. Ryzen master says between 85 and 88c, 100% load.. Its now running 1.35 vcore and 4.1 ghz, with that 612-2 cooler. Could this be correct ?
Have you tried any other software to see if the same temps are reported? For HWINFO, you're going to look at Tdie.OK< I have my Taichi up and running. But something does not add up. I think the BIOS is misreporting temps. Ryzen master says between 85 and 88c, 100% load.. Its now running 1.35 vcore and 4.1 ghz, with that 612-2 cooler. Could this be correct ?
My build is starting ! I did not realize that the X470 Taichi has a built in HSF for the NVME drive, so I wasted 20 bucks, but its awesome ! And due to what it would take for the stock cooler (I was going to try both and rate the stock one) I am going straight to the Coolermaster 612-2 pictured below. Its out of the case since you won't be able to see squat in the case.
2700X
X470 ASRock Taichi
960 EVO 500 gig NVME
2 x 8 gig PC3600 CL15 Gskill Trident-Z
Coolermaster 612-2
Corsair AX-860 PSU
A 1050TI and a 1060 3 gig fill-in's until I can afford another 1080TI or 2. (and prices go down)
Its an open case with a huge cooler that works great on a 14 core Xeon. https://www.newegg.com/Product/Product.aspx?Item=N82E16835103202Are you putting this together outside?
tdie is 77.5 on hwinfoHave you tried any other software to see if the same temps are reported? For HWINFO, you're going to look at Tdie.
Probably closer to actual temps. Have you tried HWMonitor or aida64 to compare and see if they agree?tdie is 77.5 on hwinfo
So the Team Dark memory I tried out has to be run at 3000 instead of 3200 with The Stilts tweaked CL14 timings. I may try 3066 later.
The only settings changed in the BIOS are cpu voltage offset -.04375 and memory at 1.355.
Running prime95 blend test, 2048-4096, 29GB memory allocation, so far, so good. Hoping it passes the 24 hour test.
I've ran Team Dark 3200 on an ASROCK KILLER SLI/AC for around a year now with a 1700. I've modified so many BIOS settings over the year to get things solid, but one thing I do know is I have to run the memory at 1.41v -- if you take it down to 1.4v I'll have problems. Running Stilt's timings.
I have now gone to a NH-D15, and its 10c cooler.Probably closer to actual temps. Have you tried HWMonitor or aida64 to compare and see if they agree?
That's what I'm using on one of my systems (with 1 fan). The last air cooler I used was on a Phenom 2 system. I couldn't believe the size and weight when it arrived compared to what I used to use.I have now gone to a NH-D15, and its 10c cooler.
Not sure how exactly to answer that. I am pretty sure the x470 has better power delivery for the higher "factory overclocking" and better memory capability for running faster ram and at higher speeds. If you don't change memory, and don't overclock memory or CPU, then its a BIG upgrade even with a 370 mobo IMO.Sorry if it's already been posted, thinking of swapping out my 1700 to 2700X but keeping my AsRock Fata1ity X370 Pro mobo. Do I lose any performance gains by not upgrading to a X470 mobo?