linustechtipscom

there's no option to set m.2 slot to gen 3. CrystalDiskInfo, hardware info are still reporting PCIe 4.0

I was thinking it was defaulting to 1x 4.0 lane didn't thought it could be setting to 2.0. How do I manually set PCIe speeds in bios, anyone here familiar with MSI clickbios 5?

had the board for quite a while bought it like November 2019. Had no issue with it what so ever. I've OC 3600x, 5800x, 3600mhz micron E, 3600mhz hynix C and 4400mhz Samsung b with really good results (3800mhz 14-8-16-13-27-288 at 1.53v). Only explanation I have is poor connection between CPU pins responsible for PCIe lanes. OR pins in m.2 slot. Probably not good enough connection for higher frequency PCIe 4.0 link and that's why speeds are all over the place.

I have two 980 pro SSDs 500gb and 1tb. Both of them are extremely slow when connected to top M.2 slot (which is connected via CPU 4x PCIe 4.0 lanes). Both work as expected with very good results when connected to chipset m.2 slot or through cheap amazon PCIe adapter card. And here's 970 Evo plus connected to top M.2 slot (which is connected via CPU 4x PCIe 4.0 lanes). Specs: CPU: Ryzen 5800x motherboard: MSI x570 gaming plus. CrystalDiskInfo, hardware info and Samsung magician report the SSDs connected through 4x PCIe 4.0 Iv'e tested 4 SSDs (980 pro 500gb/1TB, 970 evo and P34A80 1 TB) in both m.2 slots and PCIe adaptor card (PCIE to m.2). The PCIe 3.0 SSDs works fine in both m.2 slots and PCIe adaptor card. PCIe 4.0 SSDs work fine in the bottom (chipset) m.2 slot and PCIe adaptor card. but both 980 pros slow down when plugged into top CPU m.2 slot. Also my bottom 16x PCIe slot is connected to the chipset lanes and top is connected to CPU lanes. when the 980 Pros are plugged into PCIe adaptor card they work fine in either of the PCIE 16x slots. I've updates to SSDs to latest firmware. bios to latest firmware. Updated chip set drivers. forced installed Samsung NVME express drivers. disconnected all SATA devices disabled onboard LAN and audio devices. reseat the CPU twice. please send help I'm trouble shooting this for three days straight.

Up until january 2019 Nvidia didn't support freesync monitors. G-sync was exclusive to G-sync monitors. What you have is a g-sync compatible monitor and these only clock down to 48hz. True, literally makes the difference between getting killed or killing your opponent in almost any FPS game. Input lag is a serious issue which almost every above average players has known about it for years and now Both LTT and Nvidia have confirmed it.

You know that g-sync has latency penalty, beats the purpose of high refresh rate panel. Screen tearing isn't that big of a issue at high FPS so might as well just leave it off. What's most important at high fps is consistency, uncapped FPS will be all over the place and inconsistent. It might also cause stutters. You just want to cap your fps to just above your monitors refresh rate. For games that can't achieve high fps you just want to cap the fps to just above average FPS. Let's say a game gets about 125fps, you wanna cap it to about 130fps. Also G-sync monitor have variable refresh rate. This means if GPU fps drops below the max refresh rate, monitor will dynamically lower the refresh rate all the way down to 1hz. This good for when you max out the settings and game run sub 100 fps. There won't be any tearing even if FPS drops below something like 15 fps. AMD freesync has the same feature but refresh rate doesn't go below 48 hz. So, on AMD freesync is useless below 48fps.

link to store page.

it would hardly be a monopoly because TSMC, Samsung and Qualcomm exist..

what GPU are you using? your score might had went down because you might had something running in background. If you are on windows 10 enable focus assist from notification panel at bottom right and then run the benchmark. I had the exact same problem with my 3600x, lower clocks results were better than higher clock ones. I thinking its a something to do with not enough volts. If i remember correctly CPU was running fine with some extreme low voltages but the results weren't improving. If i remember correctly I was running stable 4.35 GHZ all core with 1.3375 volts but results weren't any better then 4.33 Ghz. If you're overclocking to improve in game Fps I suggest leave it at stock. I tested with many clock speeds 4.3, 4.33, 4.35, Stock gave the best gaming results. I noticed at stock speed it was boosting to around 4.2 max in games. then I overclocked it to 4.3 it was then boosting to 4.1 max in games. And when I clocked it to 3.8 it was boosting to 3.6 in games. whatever I did it was just boosting about 200mhz lower in games than the actual clock speed.

so Intel is dead until 10nm comes out in 2020, but intel 10nm for desktop is cancelled so Intel is dead until 7nm comes out in 2022.

my point is how much better does it preform stock under best conditions and overclocked? not much as intel's overclocked CPU, because the are basically in a way overclocking at stock settings.

all 3 CPUs have 6 core 12 threads. here are the specs for all three cpus Ryzen 5 3600 3.6GHz base, 4.2GHz boost. Ryzen 5 3600x 3.8GHz base, 4.4GHz boost. Core i7 8086k 4.0GHz base, 5.0GHz boost. speeds bins 3/4/4/5/6/10 (with slowest core boosting to 4.3GHz) Clearly 8086k should be faster, but it gets destroyed in multi-threading and just about beats 3600 and 3600x at single threaded workloads. Now of course i7 is manufactured on 14nm++ process and ryzen is made on 7nm process. but intel 14nm++ parts similer to AMD 10 nm parts in density of transistors. so, shouldn't be that much difference in architectural performance . I'm thinking of buying the upcoming i5 10600k which is rumored to have similar specs as i7 8086k but with slightly lower clock speeds. 6 core 12 threads at 3.7 GHz base, 4.7GHz boost and 4.9GHZ single core boost. Now on paper it looks like i5 10600k should destroy 3600x even multi-threading with is 12 threads boosting to 4.9 GHz. but if 10600k preforms similar to Core i7 8086k then there is no point of buying this, and just go AMD.

source? if that's the case then stock 3900x is just an overclock 4.6 GHZ 12 core CPU, given that there is enough cooling

I was looking on CPU benchmarks on the internet and it looks like stock AMD CPU are preforming much better then they should. Could it be possible that AMD CPU are dynamically boosting all core to max boost speed. Intel's mother boards have a setting called Enhanced Turbo which boosts all cores to max boost clock which you can manually enable or disable.

lets say we have the sufficient cooler. What AMD grantees each core will boost to considering no overclocking is applied. So, what are the official speed bins numbers for Rzyen chips?