Ryzen Owners Zen Garden

[Super XP]

10 years challenge? I still use the same case, keyboard and monitor.

I'm still using the same Classic G15 Gaming Keyboard by Logitech and the Corsair 750W Power Supply.
The rest I gradually updated throughout the years. For me the Case needs to be well designed. That's why I chose Corsair...

 

[biffzinker]

UserBenchmarks: Game 77%, Desk 135%, Work 112%
CPU: AMD Ryzen 5 2600X - 100.4%
GPU: AMD RX 480 - 68.7%
SSD: Samsung 970 Evo NVMe PCIe M.2 500GB - 298.2%
SSD: SanDisk Ultra II 480GB - 109.6%
RAM: Unknown 8GBF1X08QFHH36-135-K 2x8GB - 119.3%
MBD: MSI B350M GAMING PRO (MS-7A39)

 

[Darmok N Jalad]

Been running a 2400G mini-ITX build for about a month, and I’m fairly happy with it. I do end up with my system rebooting instead of waking from sleep occasionally. I can’t tell if it’s my RX 480, which I undervolt under a Wattmann profile. The thing is rock stable in gaming, even when overclocked to 4.1ghz. I don’t have much time to fiddle with it, as what little time I’m on it, I’m playing games. It’s basically my “Xbox” for my 50” UHD TV.

 

[Super XP]

UserBenchmarks: Game 81%, Desk 124%, Work 121%
CPU: AMD Ryzen 7 1700X - 98.1%
GPU: AMD RX 580 - 74.3%
SSD: Corsair Force MP500 NVMe PCIe M.2 480GB - 145.3%
SSD: Corsair Force NVMe PCIe M.2 480GB - 244.8%
RAM: Unknown=> Trident Z DDR4 3200 C16 2x16GB - 108.2%
MBD: Asrock X370 Gaming X

AM4, 1 CPU, 8 cores, 16 threads
Base clock 3.6 GHz, turbo 3.85 GHz (avg)
Performing way above expectations (93rd percentile)

 

[GoldenX]

Testing the 45w cTDP, plus setting -50% power on the 270X. Just for the lulz.

 

[Super XP]

Testing the 45w cTDP, plus setting -50% power on the 270X. Just for the lulz.

Let us know how it goes.

 

[GoldenX]

Let us know how it goes.

I would love to have an APC/Power Meter, the only thing I can see is the estimated power consumption from the CPU.
24W on the CPU while gaming seems nice. Performance is still good enough.

 

[Darmok N Jalad]

I think it comes down to what you get in boost on a few cores at the lower TDP. Care to post your findings?

It’s one reason I picked the 2400G—65W and it seems to have no problem hanging around 3.9ghz on all cores in games. Would love it if AMD would make a cheap 4C/8T 3000 series that clocks to 4.5ghz or more.

 

[GoldenX]

I think it comes down to what you get in boost on a few cores at the lower TDP. Care to post your findings?

It’s one reason I picked the 2400G—65W and it seems to have no problem hanging around 3.9ghz on all cores in games. Would love it if AMD would make a cheap 4C/8T 3000 series that clocks to 4.5ghz or more.

Is there any program that can register the use of the different steps, a la Android?

 

[Darmok N Jalad]

Is there any program that can register the use of the different steps, a la Android?

You can use something like HWiNFO. It will show you your max clocks and also average clocks for each core. Might even be able to do some logging, but I’ve never tried it.

 

[GoldenX]

You can use something like HWiNFO. It will show you your max clocks and also average clocks for each core. Might even be able to do some logging, but I’ve never tried it.

I seem to get no turbo at all, but a fixed normal nominal clock (3100 on my case).

 

[Darmok N Jalad]

Makes sense, I guess. But it still holds its own in games?

 

[GoldenX]

Makes sense, I guess. But it still holds its own in games?

4 cores at 3100 is like an old lower model i5. It's more than enough for normal/light gaming, plus I play at 1440x900, so the load is a bit lower than on FullHD.
Now, setting the GPU at -50%, that kills performance like overheating.

 

[YautjaLord]

Just testing the cam, sorry for such a delay, just bought the cam exactly a week ago. Panasonic Lumix DC-ZS70, good in bunch of uses, be it vlogging, concerts/fests/etc.... Also learning how to record in low light, indoors.

Will be used this year for vlogging, Bloodstock & other rock/punk/metal events & more.

*EDIT*

Just found out vlogging is not so ideal with this thing (Lumix DC-ZS70) due to average to poor mic audio, but it's according to one article about this cam. I haven't tried it for myself yet, will do some mic testing myself to hear the results. But otherwise, i love this baby so far, also has 4k photo & video shooting - not in my list of top priorities, but it's fun to be aware of it's existence. Photos & videos of the rig come out crisp with the help of the plethora of features. Cheers.

*EDIT #2*

And it's official: buying a studio lamp. The footage on my rig is not so good compared to how it looks on cam without a lamp, so studio lamp is a necessity. Will order one on ebay this Monday, found one for ~5GBP (+~5GBP shipping) vintage one. Flash defaults to disabled when recording video, setting ISO & other stuff manually doesn't seem to cut it. The long & arduous road to get things done the way i want, makes one (me) wanna brake some shit. Will post short clip on YT once the lamp is in my grasp & i actually recorded vid with these two (cam & lamp). Dimly lit indoor areas are not strong aspect of this cam without a lamp.

 

[Super XP]

UserBenchmarks: Game 79%, Desk 115%, Work 117%
CPU: AMD Ryzen 7 1700X - 93.6%
GPU: AMD RX 580 - 74.3%
SSD: Corsair Force MP500 NVMe PCIe M.2 480GB - 141.6%
SSD: Corsair Force NVMe PCIe M.2 480GB - 227.1%
RAM: Unknown=> Trident Z DDR4 3200 C16 2x16GB - 106.4%
MBD: Asrock X370 Gaming X

UserBenchmarks: Game 79%, Desk 115%, Work 117%
CPU: AMD Ryzen 7 1700X - 94.3%
GPU: AMD RX 580 - 74.3%
SSD: Corsair Force MP500 NVMe PCIe M.2 480GB - 143.9%
SSD: Corsair Force NVMe PCIe M.2 480GB - 226.6%
RAM: Unknown=> Trident Z DDR4 3200 C16 2x16GB - 107.8%
MBD: Asrock X370 Gaming X

AMD Ryzen 7 1700X-$348
49,849 User benchmarks, average bench 89%
AM4, 1 CPU, 8 cores, 16 threads
Base clock 3.6 GHz, turbo 3.85 GHz (avg)

Performing way above expectations (93rd percentile)

UserBenchmarks: Game 82%, Desk 121%, Work 119%
CPU: AMD Ryzen 7 1700X - 97.9%
GPU: AMD RX 580 - 75.7%
SSD: Corsair Force MP500 NVMe PCIe M.2 480GB - 145.6%
SSD: Corsair Force NVMe PCIe M.2 480GB - 229.6%
RAM: Unknown=> Trident Z DDR4 3200 C16 2x16GB - 108.3%
MBD: Asrock X370 Gaming X

 

[YautjaLord]

Here's mine:
UserBenchmarks: Game 111%, Desk 150%, Work 138%
CPU: AMD Ryzen 7 2700X - 105.2%
GPU: Nvidia GTX 1070 - 104%
SSD: Samsung 960 Pro NVMe PCIe M.2 512GB - 330.5%
USB: WD Elements 10B8 500GB - 34.1%
RAM: G.SKILL Flare X DDR4 3200 C14 2x8GB - 107.6%
MBD: Gigabyte GA-AX370-Gaming K7
Probably for epeen, OC'ing any component basically is legit with this app/thingy. In my case it's CPU, haven't OC'd graphics card/RAM/etc.... at all. Gaming is DX9/10 only, no DX11, let alone DX12 in this suite. They have separate DX11 & 12 test at the site, though, if my memory serves me well.

P.S. Strange, it lists me as GB (Great Britain) user, wtf?

 

[Super XP]

Your Ram is C14 and Mine is C16. Shouldn't you have better performing over mine?

I slightly OCed the CPU a bit. GPU is stock.
Wow and your M.2 Samsung is FASTer than my Corsair I stalled in the mobo's Ultra M.2 slot. Nice.

 

[YautjaLord]

Probably not, TridentZ is a bit faster than FlareX, i also see you have 32 gigs (2x16) as opposed to "just" 16 (2x8) that i have, might it be the case? Nevermind, does it's job just fine, opens large files (MP4, 1080p, 3 or 4GB in size) really fast from get go & finishes merging them in Blender Video Editor under 10mins (3 or more files, each "weighs" few good GBs).

Mine is sitting comfortably in M.2, not U.2 (or Ultra M.2, also a band jk), i really love Samsung's SSDs now, ain't i? Win 10 cloned* from my prev SATA III 850 Pro to this one in less than 15mins in it's 150GB+ entirety.

*With the help by Samsung OS Migration Tool thingy.

 

[Super XP]

Wow 15min of 150+ GB lol FAST.
That Samsung looks like its significantly faster over my already speedy Corsair Force MP510 480GB M.2 chip. Installed in my Ultra M.2 slot it says I can achieve up to 30 GB/s and the normal M.2 gets about 20 GB/s. I tested them both out and the Ultra M.2 gains a lot faster performance.

Not much of a performance increase between the 2.

Samsung 960 PRO M.2 512GB Internal SSD (MZ-V6P512BW) (This is what you have I assume?)
Max. Sequential Read Speed Up to 3,500 MB/s
Max. Sequential Write Speed Up to 2,100 MB/s

Corsair CSSD-F480GBMP510 Force Series MP510 480GB NVMe PCIe Gen3 x4 M.2 SSD
Sequential Read 3,480 MB/s
Sequential Write 2,000 MB/s

 

[YautjaLord]

Don't exactly remember the numbers, but yeah it is more or less the same. And i said less than 15mins cause while it was cloning OS from 850 Pro to this one, i did something to keep myself occupied & it took something like 12+ mins to finish, came back, the entirety of Steam games library, drivers, apps & stuff copied/cloned. This SSD is fast as all f*ck, yeah. Might run CrystalDiskMark again to be certain. And yeah, that's the name of SSD, picked it up cause it also appears in mobo's M.2 SSDs compatibility list.

 

[Super XP]

Nice.
Maybe I'll also try CrystalDiskMar. Just need to remember if I have it. Lol

 

[Hardi]

and here's mine, strange that my 850evo got better score then pro, but the evo's completely empty so maybe its that.

UserBenchmarks: Game 143%, Desk 155%, Work 149%
CPU: AMD Ryzen 7 2700X - 107.7%
GPU: Nvidia GTX 1080-Ti - 173.5%
SSD: Samsung 970 EVO Plus 500GB - 327%
SSD: Samsung 850 Pro 1TB - 124.5%
SSD: Samsung 850 Evo 500GB - 131.6%
RAM: Corsair Vengeance LPX DDR4 3200 C16 2x8GB - 110.3%
MBD: Gigabyte X470 AORUS GAMING 7 WIFI

 

[EntropyZ]

I wonder if I have posted mine, can't find it in the posting history, so here are updated results after some tweaking around clock speeds, timings and power states.

UserBenchmarks: Game 103%, Desk 89%, Work 73%
CPU: AMD Ryzen 5 2600 - 92.5%
GPU: Nvidia GeForce GTX 1070 - 111%
SSD: Crucial MX300 525GB - 96.2%
SSD: ADATA Premier SP550 240GB - 87.1%
RAM: Corsair Vengeance LED DDR4 3200 C16 2x8GB - 118%
MBD: ASUSTek ROG Strix X470-I Gaming

I locked the CPU at 4GHz via PStates, it could go higher since it is on an AIO (and technically it already matches or surpasses an average 2600X at 200MHz less, if I am correct), personally I prefer using as low voltage as possible, I kept 1.288v for the highest state and that seems to be OK, not sure if it will go lower but I got it stable at that on the first try and didn't bother with any testing after that. SMT is disabled, since the games I play don't take advantage of extra threads anyway. Higher clocks are a must for poorly optimized stuff.

I don't think there is much point in re-testing with SMT enabled, plenty of results out there. I've just thrown a curve-ball here a little.

Squeezing some more speed out of the system would be easy, but the power savings start dropping off, obviously.

I wanted to sell that memory kit for some G.Skill Tridents, but it seems it's not really worth switching TBH. Who would have thought Samsung E-dies were worth something. They were trouble to run, well... Until AGESA updates dropped for almost perfect DDR4 support.

 

[Hardi]

nice ram score, what timings are you using?

 

[EntropyZ]

nice ram score, what timings are you using?

The latency is still what drives me absolutely nuts because Intel platforms boast it to be that much better. It's all numbers though... In real usage cases it probably doesn't make any difference when I game, but alas can't test that without a i5-8400/i5-8600K somewhere lying around.

I'm happy with the numbers I dialed in for the kit. But I am wondering if it can go further. It is also helped by having an ITX motherboard, because the DDR slot traces are closer to the socket. Which in a nutshell, usually nets better stability when trying to dial the lowest timings possible.

I had an X470 MSI Gaming Plus motherboard which couldn't do such timings at the time, surprisingly. Maybe the newer BIOS updates tuned something later, but I have sold the board on basis that I wanted a portable system capable of smashing games... So, no luck re-testing that. I wish I had disposable income to just hoard hardware, but some people just wouldn't be happy with that.

The memory kit has been with me since Zen launch, and it's been getting better with age.

Watching @buildzoid really made me start caring about memory timing control in the first place. His detailed explanations and insight were invaluable to me.