AMD RYZEN Builders Thread

Page 71 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Justinbaileyman

Golden Member
Aug 17, 2013
1,980
249
106
My Crossair is also arriving today. Not going to lie, if this board is DOA or dies from a failed update (that isn't my fault ) then I'll likely be sending everything back to Newegg and wait a couple months for all this nonsense to be sorted out. AMD made a great chip but they are getting screwed over for being the little guy.
I am thinking the very same thing. I am going to keep what I bought and give it a chance,but the second anything goes wrong or my mobo auto bricks, it all goes back asap!!
 

SketchMaster

Diamond Member
Feb 23, 2005
3,100
149
116
A stock 1700 is only going to be clocked at like 3.1GHz on all cores, and you're putting the work on the gpu.

Well you're starting with a 25% less CPU clock and most of the heavy work is being done by the GPU thereby removing the benefit of the extra cores?

Personally I'd think the 1800X would be a better competitor to something like the 5820K with the higher base clock.

Wanted to start with a stock 1700 to see how it scales with tweaking, also didn't feel like dropping extra cash on a 1800X for marginal gains. Using the GPU for the work because that's how I normally render out videos, I remember the days before using GPUs and I don't want to go back

I boosted my RAM up to the proper 2666 and already saw a nice improvement. Render time is already down 20 seconds, so this chip really does love high speed RAM!
 

Justinbaileyman

Golden Member
Aug 17, 2013
1,980
249
106
Sorry, I have a touch of ocd, and my motherboard won't be here until Tuesday due to the weather.
Figures, motherboard didn't show up today... Did a tracking search on Fed-Ex.com and got a all white page with this message.
"The winter storms are causing unavoidable service delays in the midwest and northeast regions."
Now I gotta find time and hope I will get it tomorrow.
 

Malogeek

Golden Member
Mar 5, 2017
1,390
778
136
yaktribe.org
I boosted my RAM up to the proper 2666 and already saw a nice improvement. Render time is already down 20 seconds, so this chip really does love high speed RAM!
Nice! I'm trying to figure out what RAM I want to get to run at least 2933, but been waiting for ASUS to post a BIOS update. Only couple of weeks left that I'm able to return my Corsair LPX for a full refund.
 

Doom2pro

Senior member
Apr 2, 2016
587
619
106

Yeah I posted that last page, it's not just NZXT it's also Corsair (Models using ASETEK Pumps). I have been hearing 3-4 weeks, not sure if that's true and if it is, what date that is relative to (dating back to). There are NZXT forum posts with the same alert you just posted that were posted 5 days ago, so there is that.

I ordered the Corsair AM4 Bracket (sourced from ASETEK also) and they gave me and others an email recently saying shipments will be delayed, due to increased demand. Sounds like it was BS excuse due to same manufacturer issues.
 

SketchMaster

Diamond Member
Feb 23, 2005
3,100
149
116
I boosted my RAM up to the proper 2666 and already saw a nice improvement. Render time is already down 20 seconds, so this chip really does love high speed RAM!

Ok, after going to BIOS 902 and loading the D.O.C.P 4 profile for 2666 RAM I now have my render time down to 7:47! That's massive for just getting the RAM optimized. Going to start working on getting the CPU lightly OCd next.
 

SketchMaster

Diamond Member
Feb 23, 2005
3,100
149
116
I'm getting a really strange issue. The CPU doesn't seem to boost at all, maybe shoot to 3.7 for a moment and then drop back down to 3.1 even if I'm hammering it with OCCT. What's odder is the CPU Temp will report 0* if I change the CPU core ratio from Auto to manual but leaving FID and VID alone (meaning I'm not changing the actual CPU ratio); if I set it back to auto and reboot the temp is "normal" again but I'm noticing it doesn't move much past 29*C even when stressed with OCCT.

I haven't done anything that would fry the CPU. I set it to 3.7 @ 1.35 volts just to get the feel for the new BIOS settings.

Any ideas? Think I have a bad CPU?
 

richierich1212

Platinum Member
Jul 5, 2002
2,741
360
126
Have you tried another program? Cinebench?

I thought TheStilt said Prime is an old program (I assume OCCT is the same) that doesn't stress the new Zen architecture where it will boost enough on it's own.
 

Crono

Lifer
Aug 8, 2001
23,720
1,502
136
I'm getting a really strange issue. The CPU doesn't seem to boost at all, maybe shoot to 3.7 for a moment and then drop back down to 3.1 even if I'm hammering it with OCCT. What's odder is the CPU Temp will report 0* if I change the CPU core ratio from Auto to manual but leaving FID and VID alone (meaning I'm not changing the actual CPU ratio); if I set it back to auto and reboot the temp is "normal" again but I'm noticing it doesn't move much past 29*C even when stressed with OCCT.

I haven't done anything that would fry the CPU. I set it to 3.7 @ 1.35 volts just to get the feel for the new BIOS settings.

Any ideas? Think I have a bad CPU?

Try feeding it more voltage, at least temporarily.
 

SketchMaster

Diamond Member
Feb 23, 2005
3,100
149
116
Have you tried another program? Cinebench?

I thought TheStilt said Prime is an old program (I assume OCCT is the same) that doesn't stress the new Zen architecture where it will boost enough on it's own.

Try feeding it more voltage, at least temporarily.

I set it to 1.35V and ran Cinebench, still held firm at 3.1Ghz though it did finally show it was getting warm.
 
Last edited:

SketchMaster

Diamond Member
Feb 23, 2005
3,100
149
116
Also even with the Vcore set to 1.35 I'm seeing it acting hellishly strange... It'll drop to 1.31 when under load but spike up to 1.38 when idle!!
 

DrMrLordX

Lifer
Apr 27, 2000
22,035
11,620
136
So! The machine is up and running. I had a brief scare where one DIMM wasn't detecting, but it turned out all that I had to do was reseat it. Here's the stuff that went into it:

https://www.dropbox.com/s/eo0jc5wlkcxhsix/CAM00172.jpg?dl=0

Pic ain't much but it shows everything except the new keyboard. The PSU, monitor, case, and SATA SSD are all from my old machine. I'm using one of my old mining cards (Powercolor 390) for display.

Right now I'm trying to get the latest Preview build back up and running on this machine. It's a bit of a headache since the Intel AC wireless NIC is currently running at 802.11b speeds, and that's irritating. I might have to disable it and go back to my Rosewill AC NIC, or use my n USB NIC in the meantime. This thing is really slow.

Once I get the right Win10 on here I'll move on to UEFI updates, and then some more crap.

Initial impressions of hardware:

The Taichi is a pretty board. All that "board armor" that you see on it is at least partially metalized. I still thing the heatsink surfaces on my A88x-Pro were more practical, but the Taichi does look better. It looks kind of stupid from the back with the backplate not installed.

BIOS reset button on the back is a very nice touch.

The first PCIe 1x slot is disabled when using m.2_1 slot (first m.2) which is silly. The chip has a total of 32 PCIe lanes, and they want to disable that slot? Also why disable PCIe-5 (3rd full slot) when using m.2_2? You can run PCIe-2 and 3 at 8x and run 5 at 4x and still have enough lanes left for both 1x slots AND a 4x NVMe. Maybe they're worried about USB/SATA?

Installing m.2 SSDs is a huge pita. Seriously, the tiny screw is horrible, and woe betide anyone who dares to move the standoff. Get tweezers. That may not be particular to the Taichi but still . . .

I am happy to report that the D15S does fit in the Thor v2. Some folks seemed to think that it didn't/wouldn't. Fits like a charm. Vengeance LPX works beautifully even with a 3-fan configuration (2x NF-A14 + stock fan). Using 3x NF-A14 might not work though, since the smaller stock fan rests right on top of the DIMMs. I've noticed no overheating of the NVMe gumstick drive . . . yet. Gotta keep an eye on that.

The "box"-type fan clips for 140x140x25mm fans (such as Noctua NF-A14 Industrials) are . . . interesting. It took me about an hour to sort out how to use them correctly. Maybe I'm just special.

TG Conductonaut is really runny. It sort of beads up like old-school CLP, but not quite so badly. It is runny enough that I do not recommend coating your heatsink base with it. Or maybe I just used too much. It's really hard to control the syringe. One little bump on the plunger and away you go! It's easy to spread with the enclosed swabs. It also looks like the swabs should be reusable several times.

NF-A14 IndustrialPPC 3000 RPM fans are pretty loud. They are awesome, too. I'll do more temp reporting once I get software + UEFI set up on this thing. But so far temps are very low at stock according to the UEFI. I'm not seeing the 1800X/1700X 20C offset . . . yet.

More impressions later.
 

SketchMaster

Diamond Member
Feb 23, 2005
3,100
149
116
Hmmm... Not impressed with the stock 1700 vs my i7 5820K @ 4GHz. Rendered out a video in Adobe Premiere that has a good mix of color correction, warp stabilize, downscaling of 4k to 1080p. Both systems are using the same 32GB 2666 RAM and GTX 1070, I used CUDA rendering and a 2 pass with 25 Target / 35 Max bitrate.
  • i7 5820K @ 4Ghz render time: 6:51
  • Stock 1700 render time: 8:20
Going to play around with making sure I have all the right drivers and see what the RAM is running at. Not a good start though...

After a lot of cussing, I got a stable 3.9Ghz @ 1.35V (LLC brings it up to ~1.37v under load) and got a render time of 6:39. Still worried about the Vcore Jumping up and down 0.02V, the CPU not Boosting when stock, and the strange thermal issue. I'll try swinging by Frys or another shop to pick up a 1700 (maybe X, if they give me a good price) to make sure it's not the CPU. SXSW is in full swing so traffic is a mess!

I'm pretty impressed with the results for all the pain I just went through. While I did go through a lot of hassle for a ~2% gain over the i7 5820k I have to take into consideration the maturity, price, and scaling of the Ryzen. Out of the box I was able to get > 9% jump in performance from just getting the RAM tuned and the OC got me another +17% boost, and that's with most of the work being offloaded to the GPU.
 

krumme

Diamond Member
Oct 9, 2009
5,956
1,595
136
After a lot of cussing, I got a stable 3.9Ghz @ 1.35V (LLC brings it up to ~1.37v under load) and got a render time of 6:39. Still worried about the Vcore Jumping up and down 0.02V, the CPU not Boosting when stock, and the strange thermal issue. I'll try swinging by Frys or another shop to pick up a 1700 (maybe X, if they give me a good price) to make sure it's not the CPU. SXSW is in full swing so traffic is a mess!

I'm pretty impressed with the results for all the pain I just went through. While I did go through a lot of hassle for a ~2% gain over the i7 5820k I have to take into consideration the maturity, price, and scaling of the Ryzen. Out of the box I was able to get > 9% jump in performance from just getting the RAM tuned and the OC got me another +17% boost, and that's with most of the work being offloaded to the GPU.

As standard the cpu voltage goes to 1.3625 xfr so you are perfectly safe. Your cpu can perhaps run 4.0 at 1.39V? Seems very fine to me.

Your workload probably have a lot of branching that is not favorable to Ryzen and therefore depend on ram latency and ccx shoufling. Its nearly a worst case scenario for zen you just need some avx2 in the loop to make it worse. 3200 ram would probably get it 8% faster than your hsw 4 vs 4 ghz. Some workloads is just like that.
 

Valantar

Golden Member
Aug 26, 2014
1,792
508
136
I set it to 1.35V and ran Cinebench, still held firm at 3.1Ghz though it did finally show it was getting warm.
If I've understood the peculiarities of Ryzen correctly, any adjustment to the CPU settings that even resembles overclocking (such as adjusting voltage) will disable boost completely and set the CPU into "OC mode" (P0 power state). Are you sure it was completely stock when you tested it, with no adjustments at all?
 

JimmiG

Platinum Member
Feb 24, 2005
2,024
112
106
I set it to 1.35V and ran Cinebench, still held firm at 3.1Ghz though it did finally show it was getting warm.

Well multi-threaded Cinebench will use the all-core boost of +100 MHz.

For single-threaded Cinebench, you may need to switch to the Balanced power plan for Precision Boost to work. Without core parking, it doesn't seem to detect idle cores properly and you're limited to the "all cores" boost speed, which is just 100 MHz above the base clock.

Also keep in mind that the single-threaded Cinebench thread will jump around between cores, so you'll have to use something that monitors all cores to keep track of which core is boosting.
 
Last edited:

dacostafilipe

Senior member
Oct 10, 2013
772
244
116
Update on my PRIME B350M-A build.

I've now got the RAM running at 2666Mhz without DCOP, will shoot for more this weekend.

Will also try the older 0405 BIOS again this weekend and if I can run the RAM at the same speed, I will stick to it. The 0502 does seems to read the CPU temperature wrong (20C offset?) and messes up with the fan speed.

With the 0502 I also get higher Vcore compared to 0405. At stock, it reads 1.415 Vcore in the BIOS and CPU-Z shows me up to 1.52V ... kinda scary :/
 

Valantar

Golden Member
Aug 26, 2014
1,792
508
136
I have to say the R5 1600X is looking just as attractive as the rumors made it out to be. Considering that I still don't have a motherboard, and probably won't until way past the R5 launch, I'm seriously considering returning my 1700X. Although I can just move that to my girlfriend's PC when we get around to building it (she has more use for 8 cores than I do). Hm.
 
Reactions: blure007
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |