Bought Ryzen 3900x...
Hey all, decided to share the first impression of this CPU, as I was requested by a couple people.I ordered this processor to OEM version.
But I came BOX version with a cooler, nakleechek all as expected.
Spoiler
The processor to insert in the socket is even easier than intel
Bought fee with a processor the msi x470 gaming pro carbon
When buying asked to update the BIOS, I requested for this service of$5.
In General, he assembled all the system unit, included and everything worked.
RAM (G Skill RipjawsV F4-3200C16-8GVKB) just picked up the XMP profile and without problems turned on.
Cooling castle Deepcool rgb 240
However BIOS is still buggy, it profiles.
The processor is not dispersed, works on stock frequencies
PBO I feel actually nothing, only to eat more starts up to 150w.
Spoiler
In the benchmark CPU Z gives the following results (set for comparison 18 core from Intel)
Spoiler
In Cinebench r15:
Spoiler
Today I will test in games, the theme will be supplemented.
Tests will be run with SMT without SMT
Write what games/programs would you like to see in the tests.
Watch dogs 2 SMT
Spoiler
AC: Origins SMT
Spoiler
Crysis SMT
Spoiler
Gta 4 SMT
Spoiler
Battlefield 1 SMT
Spoiler
Gta 5 SMT
Spoiler
Witcher 3 SMT
Spoiler
Total war Warhammer 2 SMT
Spoiler
PBO seems to be shut off. Not high voltage 1.464 v? it's in easy ? it to load rolls up to 1.5 volt )? The process of degradation of the stone run. So you games it took something) All is nice, but just for gaming it is pointless to take it. It is clear that there will soon be upgrading consoles, but not be limited to 3600, max 3700x. The difference in games is not so big, especially at resolutions above 1080p, the difference will generally minuscule) between 3700x and 3900x. Do a test BF1.
Aloska
Well... Yes, will always run the electrons in its silicon crystal, pleasing high performance in games. Congratulations!
To queue comparisons and tests?
Aloska
Oooo, and here's the first reviews I see) Laid out the description of the RAM and keep that spirit in the games I asked for or any their add) And here's another: Increase a bit the resolution, and vidyahu a shame as the dogs I see she is working at half power. Don't forget about the Assassins!!!! I want to personally compare with his performance in Alexandria. Very interesting! Do the review VIDOS, like our mutual friend chela.
Spoiler
So that would have to ask...)))
IT IS VERY IMPORTANT!!!!!!!!!!!! Do ka test memory cache in AIDA.
AIDA Cache & Memory Benchmark
STANDS of Milivoje wrote:
IT IS VERY IMPORTANT!!!!!!!!!!!! Do ka test memory cache in AIDA.
For Matisov (zen2), These values generally have little influence on that, unless latency (delay), and very little (slightly more error).
The thing in Horse size L3 cache. Even the low value of the entry In memory on 6-8 core, has virtually no effect on actual performance (2-3% FPS from 3900x if the difference in speed of record 2 times!).
Here the cache CPU (speed), YES. There is something to see.
Ryazancev wrote:
Here the cache CPU (speed),
Oh, yeah) Here, by the way will be interesting to see the results!
STANDS of Milivoje
Spoiler
Source
Spoiler
Tell you what for a man the hemorrhagic (TESTS), when everything is all tested. Another thing is GAMES. They are different+different graphics settings and resolution.
Aloska
The most important thing I forgot to ask)
Damn it, or asked.... In-ear dolblyus, don't remember(
What's your PERC was before you decided to take as many as 12 cores?????????????????
And is there any big differences in performance between the new and the old PERC?
Ryazancev
This man is familiar with piebald. As we are all here. I believe he is much more than all sorts of paid tests online. Moreover, he has a good chart of indicators on the left side of the screen. It's pretty obvious what I need! Wasn't bad if he announced everything, that is voice recorded and concluded. Then all a fairy tale!)
STANDS of Milivoje
STANDS of Milivoje wrote:
And is there any big differences in performance between the new and the old PERC?
Was 6700, Professor tasks, the capacity increased 3 times, in games like there is no difference, as the games until more than 8 threads and not loaded.
And more on the future than the needs.
test, please, for these games to show me odnopoliy:
GTA 4
Saints row 2
these games are the most krovoobraschenie projects of the past decade (unsuccessful pitch-ports from consoles) and work in odnopolie. use full HD and maximum graphics.
test even the first crysis on the highest graphics in DX10 and fullHD. the first part processor oppression more harshly than the next two. reach at least to the first checkpoint Koreans on the shore, where in the story you come to a hill and greeted the dawn. run there, hit the checkpoint and show what fps is.
and test the first assassins creed on max graphics in fullhd.
this I why I ask: I want to compare the FX 8xxx, to trace the evolution of processor performance. I have now i9-7900x, and was formerly FX 8320, where the games normally are not exported even with gtx 1080. on utube such perversion is now not exactly find the so that you have a chance to be a pioneer.
and I would love to see the rendering in 3d studio max.
Aloska
And acceleration is what? promised 5GHz. And performance per core is generally the first test was above what Intel. I like the overclocking performance per core in CPU-Z 600 points.
Aloska wrote:
In the load voltage decreases down to 1.1 v
It's just a hell from 1.467 to SAG to 1.1. I have for example 1.146 1.137 to sags. Just ditch the degrade the percent easy because in addition to drawdown at the time of normalization is the opposite of a short leap naprugi higher than idle. And 1.457, not stress? I'm on my overclocking all cores at 5GHz 1.3 above, do not put.
KillKammbest
KillKammbest wrote:
And 1.457, not stress? I'm on my overclocking all cores at 5GHz 1.3 above, do not put.
This is the standard frequency of the AMD, do you have any idea what AMD claims waiting if the percent starts to degrade?
When the temperature of the CPU increases, the voltage decreases in order to prevent degradation of the crystal.
And Yes, I threw the above test cpu z
Processors ryzen 3xxx series there is no sense to drive, and so they snipe from the factory overclocked.
Aloska
It's just what GIMP to percent because at the time of alignment naprugi after load is a short-term jump higher naprugi in a simple equivalent drawdown under load.
Aloska wrote:
Processors ryzen 3xxx series there is no sense to drive, and so they snipe from the factory overclocked.
How came out the first article about 5GHz on all cores + performance per core than Intel but in reality it turned out to be fake. I personally really believed 3600 just tear in down and ashes 8700К due to higher performance per core and no problem overclocking up to 5GHz on all cores.
KillKammbest
On the same frequencies AMD wins intel.
About 5 GHz was, a lie.
But my 12 cores for$ 500, Intel would never allow it.
KillKammbest wrote:
personally really believed 3600 just tear in down and ashes 8700К due to higher performance per core and no problem overclocking up to 5GHz on all cores.
You know miracles do not happen, here's a couple of generations and I'll hang a portrait of Lisa su on your wall!)
Aloska
Yeah, I looked. Performance in AIDA Cache & Memory Benchmark memory is too small of course. Opera your I don't know. But the indicators with regards to percent on three levels of cache are fine)