AGESA 1.0.0.6 is really good. I built my system last week and did some experiments. R7 1700X, Taichi (BIOS 2.36 now), 2x16 GB TridentZ 3200 CL15.AGESA 1.0.0.6 continues to bring interesting surprises.
After wasting a lot of time trying to get DDR4-3600 stable (not happening yet; I can get it to run almost everything except y-cruncher), I went back to the drawing board and decided to see what kind of subtimings AMD gave me using Ryzen Master to set DDR4-3200 14-14-14-32 (actually I went for 28 this time) as I had in the past. I went through my old routine, booted it up, confirmed that it was working, and then took a look in the UEFI to see what was going on.
Curiously the UEFI was showing the current timings/subtimings that had been set by Ryzen Master, but it was showing the DIMM's XMP settings as what was currently set in the UEFI. For example, it showed CAS/CL as 14 but the config field was set to 17. Booting into Windows asserted the Ryzen Master settings provided I changed nothing in the UEFI. Changing settings in the UEFI apparently "cancelled out" whatever Ryzen Master was trying to do. It did not overwrite settings changed by hand.
I can go back and record all those settings and post them here if you want to see exactly what Ryzen Master and the older pre-1.0.0.6 AGESA revisions were doing when setting the DDR4-3200 strap.
So anyway, I got 3200 stable (preditably) without changing anything. Using 1.4v vDIMM and 1.175v SoC voltage (LLC level 3) I got DDR4-3333 working with the exact same timings as what Ryzen Master set for DDR4-3200 14-14-14-28. Cool! AIDA64 reports ~52.8 GB/s read and 69.2 ns latency. Then I ran Firestrike to compare it to my Firestrike score from when I was running DDR4-3600 18-19-19-36 (1T and 2T; it made no difference at all).
The main thing that jumped out at me was the combined score. When running DDR4-3600, I only had the CPU at 3.6 GHz, and I got a combined score of over 5200 using a stock Powercolor 390. When running the CPU at 4 GHz with RAM at DDR4-3333 14-14-14-28, I got a combined score of ~4900 with the same Powercolor 390.
It seems to me that fabric speed is a major limiter of combined score performance in Firestrike.
You guys think there will be new steppings of the Ryzen chip line? I would love to see a 1500x hit 4.5GHz or more. This 3.9Ghz stuff is a real bummer right now.
It's the process limits, not the architecture. If they move Ryzen 2 to LPU, it should clock higher than the LPP process they use currently.You guys think there will be new steppings of the Ryzen chip line? I would love to see a 1500x hit 4.5GHz or more. This 3.9Ghz stuff is a real bummer right now.
Wraith Max is only for OEMs and SIs, not for retail or bundled with CPUs.1600X, 1700X and 1800X are "NOFAN" SKUs. The rest either come with the Wraith Spire (RGB for 1700, non for 1600, 1500X) or Wraith Stealth (1400 and probably the entire R3 line).Where are all the Wraith Max coolers? when I purchased my 1800x it didn't come with any cooling and I guess they still don't.. The whole reason I purchased the 1800x over the 1700 was to get that darn RGB Wraith Max cooler.
Is 78C in Ryzen Master under load, too hot for a Ryzen 5 1600 CPU?
Edit: Make that 83C in Ryzen Master, after using 10 cores to crunch WCG in BOINC, and leaving two cores "free", but Task Manager shows nearly every core fully utilized.
Since at least one person asked me for timing/subtiming info, I've done my best while short on time to provide information about my subtimings for DDR4-3466:
https://www.dropbox.com/s/cjf7f314m2ne2m4/3466.BIN?dl=0
That's a .bin file of my UEFI settings on an X370 Taichi running UEFI rev 2.34 beta. Good luck parsing that, though if you have a Taichi or other ASRock board that uses the same file format, you might be able to read my subtimings using that file. User beware, my settings might not be good for your system if you actually try using them. For me it gets me full AVX stability with DDR4-3466 with the CPU @ 4 GHz. Also notable is that my system seems to like me using 1.4v vcore + Level 3 llc instead of 1.35v vcore + Level 1 llc when running certain applications like SuperPi on higher memory clocks, since the llc doesn't really seem to kick in when running single-threaded workloads. The lower vcore kills memory stability. Interesting.
According to Aida64, I get 68.7ns latency and 53167 MB/s read bandwidth with these settings.
I've tried basically the same timings at 3600 and have found it quite bootable, but not stable. Um yet.
Anyway if the .bin file isn't sufficient, I'll have to find some other way to record the info. It's a shame I can't export to raw text. Hey ASRock, there's a new feature for the UEFI! Get right on that.
Wow. Definitely better-binned. Or I'm doing something wrong (entirely possible).and my 1600X @ 4 GHz with 1.3v no problem.
Wow. Definitely better-binned. Or I'm doing something wrong (entirely possible).
The best I can seemingly get my 1600 to, is 3.70Ghz @ 1.300V in UEFI.
Can't do 3.80Ghz @ 1.3500V, nor even 1.400V. Temps are crazy, too.
Using the stock cooler than came with my 1600.
Fan speed setting in UEFI is set to "standard".
This guy got to 3.85Ghz with 1.400V, max temp 72C.
https://www.youtube.com/watch?v=cPWMIpVrVQ0
Hmm, yeah, my plan is to, eventually, get a Phantek P400 tempered glass case, and then a 240 or 280mm Rad CLC AIO kit to cool the CPU.
It's just kind of weird that I can't even stabilize 3.80Ghz at all, not even at 1.400V. Still wondering about my video card, or my mobo VRMs, if it's just completely blacking out, and not BSOD'ing.
It may not be the CPU (cores) itself.
I've moved on to overclocking my RAM.
I've got my DDR4-2400 Team Dark 2x8GB kit, running at 2667 @ 1.300V VDIMM.
Running some DC WUs, hopefully they don't crash. I figure if my RAM's bad, though, I'll get a BSOD, rather than just a black-out.
https://www.youtube.com/watch?v=H3zg-LfxkN4
See 21:40.
"Spire will limit you. It's just not meant for overclocking."
Hmm. Looks like cooling is my issue.
So far, so good, running my Team Dark DDR4-2400 at 2666.
Wow. Definitely better-binned. Or I'm doing something wrong (entirely possible).
The best I can seemingly get my 1600 to, is 3.70Ghz @ 1.300V in UEFI.
Can't do 3.80Ghz @ 1.3500V, nor even 1.400V. Temps are crazy, too.
Using the stock cooler than came with my 1600.
Fan speed setting in UEFI is set to "standard".
This guy got to 3.85Ghz with 1.400V, max temp 72C.
https://www.youtube.com/watch?v=cPWMIpVrVQ0
AGESA 1.0.0.6 is really good. I built my system last week and did some experiments. R7 1700X, Taichi (BIOS 2.36 now), 2x16 GB TridentZ 3200 CL15.
That dual ranked RAM is running nicely at 3200 14-13-13-13-32-1T and 1.35V now, no VDD SOC increase was necessary. I should try 28, too.
Higher clocks didn't work so far. I tried ASRock settings (OC Tab), AMD CBS DRAM settings, and Ryzen Master.
Thank you. I will take a look at that over the weekend when I have a bit more time. I'll see if I can find you some timings that can get you stable as well. I'm able to run my DDR4-3200 at the suggested timings and my 1600X @ 4 GHz with 1.3v no problem. So, we can probably loosen your stuff up a bit more, and tighten the subtimings. What are your flavors of stress programs if you are able to boot?
So I tried the AGESA 1.0.0.6 Beta BIOS that just came out for the Prime X370.
So far the results are a little disappointing and not as good with the AsusPeasantPrime X370 as the C6H and other high-end boards from what I've read. This is one instance where it would have paid off to get a more expensive board (normally that just gets you more bling and possibly better VRM cooling).
It seems 3333 C14 1T is reasonably stable at 1.4V DRAM, 60 ohms and 1.1V SOC. That's a pretty modest increase from the 3200 C14 I was running with AGESA 1.0.0.4, and that only required 1.35V DRAM and 1.0V SOC. I'll have to do some more testing, because I didn't have time to leave Memtest running for very long last night.
3600 and 3466 seem to be out of reach with my combo of CPU, board and RAM. 3600 won't even load Windows (complains about corrupted dll's etc.) and 3466 is memtest unstable even at 1.4V DRAM and 1.15V SOC.
So, a few notes for prospective first-time (or first time in ten years, as me) water cooling builders:
- Don't go anywhere near a Fractal Define R4. I would like to send a sincere f*ck you to the design team behind that case. "Front 240mm radiator support" my rear end. Sure, if you're willing and able to mangle the front fan mount entirely, and spend 20 minutes fiddling to make the screws connect. And don't get me started with the top mount that I guess could fit either a radiator or a set of fans (even though it is offset!) Great job, Fractal. Yes, I know it's an old design. I timed my purchase perfectly - just a couple of months before the R5 launched. *pats self on back* (Oh, did you say mangled fan mount? Yes. Watch your fingers.
And that was me being extra careful.- If your radiator mount is fiddly, give yourself enough slack in the tubing to move stuff around, for example to set the fans to pull rather than push.
- Check that the fans spin freely before connecting everything and filling the loop. Saves you a lot of annoying rat-tat-tat noises and a couple of hours of involuntary case modding.
Other than that, I'm very happy with my build. The increase in general speed from the previous CPU is rather remarkable. Haven't been able to run many benchmarks yet, but I expect everything to go smoothly.
I am considering an R4 for air cooling. Water cooling makes my skin crawl. So, aside from water cooling difficulties, does it have any other issues?I can confirm the R4 is complete crap for water setups. I sold it for a Corsair Carbide Air 540. A little less quiet, but an absolute joy to work with because there is so much space everywhere.
I am considering an R4 for air cooling. Water cooling makes my skin crawl. So, aside from water cooling difficulties, does it have any other issues?
Thanks, I'll do that. I can't stand a noisy computer.Why get the older R4 when the R5 is usually the same price, and fixes most of the complaints people had about the R4?
The R5 reviews point out what changed, so you might want to read a couple of them before buying.
I think it's a shame that it lacks a basement, as cable management space is very limited. I always struggle to get my right side panel on properly. Unless you need optical drives, it wastes a lot of space on that bracket. The dust filters are utterly useless. I mean, they stop some dust, but not nearly enough. The SSD mounts on the back of the motherboard tray are utterly useless unless you use an ITX motherboard, as they screw in from the front of the tray (and are thus covered by the motherboard when it's installed). Still, it's not a bad case. It's reportedly more quiet than the R5, has more than enough fan mounts, decent airflow for a silent case, removable HDD cages, and the insulated covers for the top and side fan mounts are nice. The included fans are okay too. I've grown to think it looks a bit dated and boring (the shiny black plastic front fascia doesn't help), but it's stylishly minimalist.I am considering an R4 for air cooling. Water cooling makes my skin crawl. So, aside from water cooling difficulties, does it have any other issues?
I would probably recommend that too. Or the Define S (ATX) or C (mATX) if you don't need 5.25" bays. As I've said, the R4 is reportedly quieter. But not by much, and everything else that I have pointed out should be drastically better.Thanks, I'll do that. I can't stand a noisy computer.
For the near future anyway, I need a Blu-ray burner. I'd like to get rid of needing that at some point.I would probably recommend that too. Or the Define S (ATX) or C (mATX) if you don't need 5.25" bays. As I've said, the R4 is reportedly quieter. But not by much, and everything else that I have pointed out should be drastically better.