The Game Feels Sluggish, Enemies Are Too Fast, and You Die in a Millisecond

[agendarsky]

pretty much wrong topic yes, but your theory is bad , everyone who ever experienced this problem can tell this is also happening on offline server even with lan disabled in bios. Its very old knowledge my dude. https://www.youtube.com/watch?v=U_PZnAEF4kw

[JimCarry]

well after thousands of possible fixes , usb 2.0 controller mode - fullspeed works but desync, sluggish feel is tied to svid power delivery / bad vdroop for real, im experimenting atm with 14700k and sluggish feel / off timing is gone by adjusting ia vr config and system agent vr config- to be more specific iam running ia ac loadline at value 10 and ia dc loadline at 15, and same for sa vr config. Also my motherboard (z790 gigabyte aorus -w ) have option to bypass fully integrated voltage regulator by motherboard and its very promising 4now.
The settings are Setup Question = RichTek VccIA Control Overrides
Help String = VccIA Voltage can be overrides from Richtek SMBUS/PMBUS VR. Use the next option for exact voltage configuration
Token =664 // Do NOT change this line
Offset =339
Width =01
BIOS Default =[00]Disabled
Options =[00]Disabled // Move "*" to the desired Option
*[01]Enabled

Setup Question = Fixed VID Mode
Help String = Fixed VID Mode Enabled for the VccIA/VccGT Voltage rail Overrides from Richtek SMBUS/PMBUS VR. When Disabled, voltage rail will use IMVP from SVID BUS.
Token =665 // Do NOT change this line
Offset =33B
Width =01
BIOS Default =[00]Disabled
Options =[00]Disabled // Move "*" to the desired Option
*[01]Enabled

Setup Question = VccIA (mV)
Help String = Voltage Range of RichTek VccIA from 250mV - 2170mV; Default is 1200 mV
Token =666 // Do NOT change this line
Offset =33D
Width =02
BIOS Default =<1200>
Value =<1450> <- this one is value for my vcore , maybe i will go little lower after time , im keeping good temps atm.

Setup Question = RichTek VccGT Control Overrides
Help String = VccGT Voltage can be overrides from Richtek SMBUS/PMBUS VR. Use the next option for exact voltage configuration
Token =667 // Do NOT change this line
Offset =33A
Width =01
BIOS Default =[00]Disabled
Options =[00]Disabled // Move "*" to the desired Option
*[01]Enabled

Setup Question = Fixed VID Mode
Help String = Fixed VID Mode Enabled for the VccIA/VccGT Voltage rail Overrides from Richtek SMBUS/PMBUS VR. When Disabled, voltage rail will use IMVP from SVID BUS.
Token =668 // Do NOT change this line
Offset =33C
Width =01
BIOS Default =[00]Disabled
Options =[00]Disabled // Move "*" to the desired Option
*[01]Enabled

what about amd ?

[rusihhh]

pretty much wrong topic yes, but your theory is bad , everyone who ever experienced this problem can tell this is also happening on offline server even with lan disabled in bios. Its very old knowledge my dude. https://www.youtube.com/watch?v=U_PZnAEF4kw

Bro, with all due respect — you didn’t actually address anything from the theory.

You said nothing about connection routing, packet shaping, netcode logic, or client-server dynamics.
All you did was drop a YouTube link and start talking about some mystical BIOS config for a specific CPU…
With what? Magical amperage values?
What exactly is that supposed to prove?
Who even asked for that?

Well, maybe your tweaks just improved your PC performance, and that’s why the game feels better.
That’s great — but it’s not an argument against the network-layer theory.

Also, about your “offline” take:
Funny thing — for a lot of players, it’s the exact opposite.
They train perfectly offline — smooth aim, clean movement —
then go online and suddenly the game feels like sludge. Same PC. Same FPS.
What changed? Just the connection path to the server.

That points to selective data delivery and tick sync issues — not your onboard LAN chip.

Please stop derailing a thread about netcode with hardware superstition.
We’re discussing server behavior, not motherboard volt-mod voodoo.

And honestly — I’m tired of chasing magic numbers in BIOS and registry hacks,
especially when the feel of the game literally changes from one match to the next.

[drosku]

https://www.peeringdb.com/net/5918
Riots peeringdb as they provided me with it, check with isp, request highest tech support possible, tier3, make them check it dont even bother with basic support, some ppl may get some new information and “gifts”
Funnly enough I can request better peering directly from my ISP directly to them as i dont have peering directly with as6507 lol
I will let you guys now.


Update 1-ish?Its actually possible how come i never came across changing peering? I sent them an email requesting different peering and per their words “this is something unusual as they receive peering issues rarely mainly for companies and will make it work regardless”

[ahead]

This is just a quick research about the spaghetti code. Sadly, the original Twitter message was deleted, but the idea is still clear and easy to understand:
https://www.reddit.com/r/GlobalOffensiv ... etti_code/

Please ignore all the stupid trolls.

[agendarsky]

pretty much wrong topic yes, but your theory is bad , everyone who ever experienced this problem can tell this is also happening on offline server even with lan disabled in bios. Its very old knowledge my dude. https://www.youtube.com/watch?v=U_PZnAEF4kw

Bro, with all due respect — you didn’t actually address anything from the theory.

You said nothing about connection routing, packet shaping, netcode logic, or client-server dynamics.
All you did was drop a YouTube link and start talking about some mystical BIOS config for a specific CPU…
With what? Magical amperage values?
What exactly is that supposed to prove?
Who even asked for that?

Well, maybe your tweaks just improved your PC performance, and that’s why the game feels better.
That’s great — but it’s not an argument against the network-layer theory.

Also, about your “offline” take:
Funny thing — for a lot of players, it’s the exact opposite.
They train perfectly offline — smooth aim, clean movement —
then go online and suddenly the game feels like sludge. Same PC. Same FPS.
What changed? Just the connection path to the server.

That points to selective data delivery and tick sync issues — not your onboard LAN chip.

Please stop derailing a thread about netcode with hardware superstition.
We’re discussing server behavior, not motherboard volt-mod voodoo.

And honestly — I’m tired of chasing magic numbers in BIOS and registry hacks,
especially when the feel of the game literally changes from one match to the next.

Because the whole problem is unaddresable theory , since all benchs showing no issue, yet we know it exists. You're probably first who is saying he doesnt have problem offline and thats good for you, its actually reason im strict when it comes to technical terms like desync, while you may mean its some sort of isp latency issue, i adress that as frame to frame and i/o synchronization problem, that is present in games as terrible input latency, terrible recoil problem and instruction slowdown that appears like your whole game runs literally slower ( your character, all animations). Its also present in desktop, its subtle but its there, stuff like windows opening little bit slower, youtube and websites generally takes more time. If u read any of this threads and what people tryed i bet u wont even create this thread, people have this with minimum isp jitter / low bufferbloat. Also it doesnt make any sense to server influence game engine at that level, you may have some weird peaks by other players , missed shots but slugishness, stuttering or even input lag? I dont think so . As long as i know this problem is very very old and persist on every modern pc so dont take it personally but i dont believe that you dont have an offline problem, you just probably never experienced pc without this problem and u dont know how high the bar can be. Its not about chasing magic numbers in bios, its about understanding the what triggers this problem. One thing i am really sure about , its motherboard based because for me on 100 % it started after mobo change, immediately right after change. And that was 14 years ago, and trust me i had almost every architecture since then, im actually scared to count how much money i spendt on differente hardware, so if its present on socket 478 aswell on coffe lake, am4 or raptor lake then its must be a system that have hierarchy and can be tracked, u guys that hate bios related stuff, really need some patience. Peace

[rusihhh]

(…long post about “it happens on every PC”, “it’s the motherboard”, “I changed BIOS settings”, “YouTube opens slower”, “this started 14 years ago”, etc…)

I honestly don’t understand why you keep mixing hardware upgrades — which obviously improve overall performance — with a thread specifically focused on server-side netcode behavior.

Yes, it’s normal that changing your motherboard, CPU, or socket will make your PC feel different.
But what does that have to do with how a game server decides what data to send you?

You’re talking about tweaking BIOS voltage offsets and claiming you “fixed input lag” —
but maybe all you did was get your PC to finally run properly in the first place?

Boosting raw performance ≠ solving network-level behavior.

Meanwhile, in this thread I’ve provided concrete data — Wireshark packet traces — and asked others to check if they observe similar patterns.
And your reply is... that you swapped a motherboard?

What’s next — you’ll say the more expensive the CPU, the better the netcode works?

By that logic, we should assume players competing on entry-level laptops or 8-year-old rigs should be unplayable —
yet they’re not. Some of them still frag just fine.

So respectfully:
If you want to contribute to this topic, bring packet-level data, not motherboard nostalgia.

Upgrading your PC ≠ disproving netcode theory.
Flashing BIOS ≠ packet structure analysis.
And no, we’re not secretly editing BIOS settings mid-match to explain why one round feels amazing and the next is unplayable.

IMHO, you’re sounding more like another “golden power socket” hunter at this point.

[agendarsky]

Im not telling im fixed something, videos are just visual proof of correct synch and its more like im seeking for attention to raise attention to internal power delivery. Your thread contains some nice data but trust me , even if you ever got perfect minimal packet jitter, things wont be much better because its just only little part ot wider image. Thats all i wanted to say

[paddyhhh]

Game CS2
-=WarmupServer=-™™ Mirage 21 Riffe&NoAWP FFA DM
185.242.115.38:22222
~20 players
Ping 10ms 0 jitter 0 loss
Faceit 3k elo
Feeling: medium, good hitreg but very fast reacting enemies

Playtime Duration:
419.81 seconds (~7 minutes) trimmed start and end packets

Packet Sizes:

Average: 907.38 bytes

Median: 1299 bytes

Minimum: 198 bytes

Maximum: 1299 bytes

Time Delta Between Packets:

Average: 0.00793 s (7.93 ms)

Median: 0.01226 s (12.26 ms)

Minimum: 0.00000 s

Maximum: 0.02154 s

[rusihhh]

Im not telling im fixed something, videos are just visual proof of correct synch and its more like im seeking for attention to raise attention to internal power delivery. Your thread contains some nice data but trust me , even if you ever got perfect minimal packet jitter, things wont be much better because its just only little part ot wider image. Thats all i wanted to say

Thank you — I appreciate the respectful tone, and I get what you’re saying about hardware-level concerns.

But just to clarify again — this theory isn’t about jitter, loss, or delay in the traditional sense.
In fact, the game reports everything as fine:
→ 0% packet loss
→ Ping is stable
→ No out-of-order packets
→ Smooth net graph

And yet — the gameplay feels off, delayed, or “dead before you see anything.”

Let’s illustrate why.

Imagine the game engine is designed to send you a full update per tick (say, 60 ticks/sec).
Each update should include multiple elements: player positions, actions, animations, visibility, shot fired, etc.

Now suppose:

- Player A receives a full packet in Tick T1 that contains:
→ Enemy appears, peeks corner, starts firing = ACTION_1, ACTION_2, ACTION_3
→ All in one coherent update
- Player B receives a degraded stream:
→ Tick T1: enemy moves
→ Tick T2: enemy peeks
→ Tick T3: enemy shoots
→ Tick T4: hit registered
(Note: numbers are exaggerated for clarity — in reality, multiple packets may land between, but the logic remains)

So what happens?

→ You should have reacted on Tick T1
→ But you only receive fragmented motion and action spread across several ticks
→ By the time you “see” the enemy shooting — you’re already hit
→ You’re not reacting to the enemy — you’re reacting to your own death animation

This is why it feels like:

- No time to fight back
- Enemies “snap” or appear too fast
- You die instantly
- And nothing helps — not FPS, not ping, not hardware upgrades

The client is forced to interpolate missing data — you’re watching a guess, not a reality.

So yes, power delivery or motherboard stability may affect system responsiveness,
but this problem is at a different level:
the server is deciding how much data you’re allowed to receive per tick — and that affects your fundamental ability to play reactively.