Intel Releases Windows Workaround and Patch for Ethernet Stuttering and Disconnects

[TheDeeGee]

Two examples off the top of my head: disabling TSX in CPUs, disabling DX12 in Haswell. I am sure there's more which isn't really surprising.

Good luck gaming on the Haswell iGPU anyways in 2023, pointless example.

 

[regs]

i211 is not i225. Must be an implementation problem or user issue if you have trouble with i211.

I211 has exactly same problem as I225/I226 with Windows 11 default driver. And once again - almost a dozen of I225-V3 and I226-V working with FreeBSD with no problems.

 

[TheinsanegamerN]

Good luck gaming on the Haswell iGPU anyways in 2023, pointless example.

Keep moving those goalposts, you'll get there eventually. The discussion was Intel abandoning features instead of fixing them, not their feasibility.

How has RPCS3 been "screwed over"? The TSX and AVX-512 code paths are still there and still work, so people with CPUs that have those features will still get the performance boost. You make it sound as if Intel CPUs are designed expressly for console emulation and every change Intel makes is to screw emulator users over... LMAO.

Because you cannot build features around the useful AVX512 set if Intel removes them. They put quite a bit of work into said code path only for the feature to be removed.

 

[Assimilator]

Keep moving those goalposts, you'll get there eventually. The discussion was Intel abandoning features instead of fixing them, not their feasibility.


Because you cannot build features around the useful AVX512 set if Intel removes them. They put quite a bit of work into said code path only for the feature to be removed.

And yet Zen 4 implements AVX-512.

 

[swaaye]

I have seen some disconnect issues with i218 on Win11 as well months back and disabled EEE. Chalked it up to some quirk between the switch and that particular PC. EEE is nice but yeah potential problem.

 

[ValenOne]

And yet Zen 4 implements AVX-512.

Zen 4 implements AVX-512 extensions up to Intel's Icelake level.

Both Icelake's and Zen 4's AVX-512 implementations are done with the multiple 256 bit hardware designs.

Besides SIMD width expansion, AVX-512 introduced 32 registers programming model like on RISC based PowerPC.

 

[Dave65]

Since when? What kinda baseless statement is that?

True tho.

 

[Broken Processor]

As opposed to just, you know... buying a replacement NIC?

Vertical mounted GPU. Besides I'm not in the habit of letting someone away with selling me faulty product's and doing nothing about it.

Took a year but finally got it sorted.

 

[ncrs]

Good luck gaming on the Haswell iGPU anyways in 2023, pointless example.

DX12 isn't only used by games
The "fix" happened in 2020, but it doesn't matter anyway since whether the use cases fit you or not it's still a feature that was removed by Intel.

 

[Assimilator]

Zen 4 implements AVX-512 extensions up to Intel's Icelake level.

Wrong. Zen 4 implements everything that ICL/RKL does, plus BF16.

 

[Chrispy_]

What the hell, Intel?!

This fault is years old and you finally acknowledge and patch the issue, except it's apparently still broken!

Intel NICs used to be the gold standard. Now I actually look for Realtek!

 

[Konceptz]

I226-v
Latest bios
This driver version
EEE disabled since I built the system.

problem still exists, I’ve purchased a 2.5gb pcie Realtek nic. I’ve given up.

 

[regs]

^My problem with I211 gone after installing latest 13.04 driver (I225/I226 version could be different). Pack version is 28.0.
Well, at least not a single disconnect today, but usually 2-4 times per day.

Give it a try

https://downloadmirror.intel.com/772070/Wired_driver_28.0_x64.zip

 

[chrcoluk]

To be honest, disabling a power saving feature on an ethernet port isn't a big deal - but it is a problem that this made it past all their testing in the first place

I wonder if EEE is the issue overall, with many of these companies and 2.5Gbe - mismatched implementations would absolutely explain the various random issues users face

First thing I do on NIC configuration regardless of OS and NIC brand is disable power saving features such as EEE, it has a shaky history.

AVX512
SGX
both of these were within 2 years.

Anything that seems a bother for Intel their strategy seems to be to wash their hands of it, they abandon so much stuff.

 

[fb020997]

Since swapping most of my rig from an X470 to a B550 platform, I was a bit worried that my new mobo had a Realtek 1gbps ethernet chip but my older one had the intel i211. I’ve heard bad things about the realtek…
Well, zero issues here. Both NICs never gave me problems. And now I’m happy that asus cheaped out on my B550 mobo by going with a Realtek 1gbps vs other B550 with the intel 2.5.

 

[chrcoluk]

Keep moving those goalposts, you'll get there eventually. The discussion was Intel abandoning features instead of fixing them, not their feasibility.


Because you cannot build features around the useful AVX512 set if Intel removes them. They put quite a bit of work into said code path only for the feature to be removed.

I see it as a blessing in disguise, RPCS3 was telling people to buy an AVX512 cpu, now its not on new cpus, they will need to improve performance for the majority without it.

I do agree though that it messes up software devs with the abandon mindset.

 

[RJARRRPCGP]

^My problem with I211 gone after installing latest 13.04 driver (I225/I226 version could be different). Pack version is 28.0.
Well, at least not a single disconnect today, but usually 2-4 times per day.

Give it a try

https://downloadmirror.intel.com/772070/Wired_driver_28.0_x64.zip

I thought I saw that the ASRock X570 Steel Legend has the '211, but need to check again.

Edit: The ASRock X570 Steel Legend indeed has the i211.

 

[Mussels]

What do you mean? Those exist. You mean in the consumer space?

None of these have anything to do with an ethernet chip what?

Makes sense to me, they disabled these technologies in various products after launch due to issues
Intel BIOS Updates Officially Disable Alder Lake's AVX-512 Capabilities for example

They tend to bring them back in later products, but it does come off as quite a risk to buy intel hardware without knowing what supported feature is going to get suddenly cut away 6 months later


Now, do i think disabling EEE if 2.5Gb is detected to be a major issue? Not really. This ones trivial, as a hardware revision will permanently fix the issue soon enough and percentage wise very few ethernet adaptors globally will be using extra power without it - and even then i'm pretty sure the wattage on ethernet chips is quite low

 

[ValenOne]

Wrong. Zen 4 implements everything that ICL/RKL does, plus BF16.

Wrong. Read https://en.wikipedia.org/wiki/Advanced_Vector_Extensions

Zen 4 is missing Tiger Lake's VP2INTERSECT.

Zen 4 is missing Alder Lake's AVX-VNNI VEX encoded for AVX2. Zen 4 supports VNNI at AVX-512 EVEX encoded a.k.a AVX512-VNNI.

Intel made a mess after AVX2 extensions.

AM5's CPU support can evolve longer when compared to the dead-end LGA1700.

 

[RJARRRPCGP]

Wrong. Read https://en.wikipedia.org/wiki/Advanced_Vector_Extensions

Zen 4 is missing Tiger Lake's VP2INTERSECT.

Zen 4 is missing Alder Lake's AVX-VNNI VEX encoded for AVX2. Zen 4 supports VNNI at AVX-512 EVEX encoded a.k.a AVX512-VNNI.

Reminds me that Athlon XPs don't have SSE2, but they have the first SSE, like Pentium III. Athlons before Palomino, didn't even have SSE at all.


And before that, K6-2 didn't have an instruction that Pentium IIs had!

Uggh, looks like history repeating itself again!

 

[ValenOne]

Reminds me that Athlon XPs don't have SSE2, but they have the first SSE, like Pentium III. Athlons before Palomino, didn't even have SSE at all.


And before that, K6-2 didn't have an instruction that Pentium IIs had!

Uggh, looks like history repeating itself again!

Alder Lake/Raptor Lake's AVX-VNNI VEX encoded instructions have a 16-registers programming model and 256-bit register size.

AVX512-VNNI EVEX encoded has a 32-registers programming model and up to 512-bit register size. AVX512's EVEX encoding has support for AVX-128 EVEX, AVX-256 EVEX, and AVX-512 EVEX register sizes. Intel disabled the entire AVX-512 EVEX encoding feature on Raptor Lake while AMD's Zen 4 supports IceLake AVX-512 + BF16 extension support.

Both are Intel-originated AVX extensions that are incompatible with each other.

 

[Mussels]

I know I joined in on that discussion myself (just to say i understood that users viewpoint) but we should try and keep this about the ethernet devices, even in general rather than specifics of other intel products

 

[regs]

AM5's CPU support can evolve longer when compared to the dead-end LGA1700.

For AM4 user it's still a new socket

 

[ValenOne]

For AM4 user it's still a new socket

AM4 has at least three microarchitecture changes, not just refinement.

Zen 1.0 has quad 128-bit SIMD units.
Zen 1.5 has 12 nm refinements from 14 nm.
Zen 2.0 has quad 256-bit SIMD units.
Zen 3.0 has quad 256-bit SIMD and two extra FPU units expansion. Load-Store unit count expansion. Some AVX-512 instruction support e.g. AVX-512 Vector AES, AVX-512 Carry-Less Multiplication Quadword.
Zen 3 X3D cache. L3 cache expansion.

Reminds me that Athlon XPs don't have SSE2, but they have the first SSE, like Pentium III. Athlons before Palomino, didn't even have SSE at all.


And before that, K6-2 didn't have an instruction that Pentium IIs had!

Uggh, looks like history repeating itself again!

K6-2 has 64-bit 3DNow SIMD (64-bit hardware implementation) which includes Pentium II's MMX SIMD. K6-2 doesn't support Pentium III's SSE (64-bit hardware implementation).

Pentium III's 128-bit SSE is implemented on 64-bit SSE hardware until Core 2's full 128-bit bit SSE hardware.
K8 has 128-bit SSE FADD and 64-bit FMUL SIMD hardware. Pentium IV has 64-bit SSE hardware implementation.
K10 has 128-bit SSE FADD and 128-bit FMUL SIMD hardware with K8's three instructions per cycle decoder limitations. Core 2's full 128-bit bit SSE hardware has four instructions per cycle decoders.
The instruction set may not follow actual hardware implementation.

PowerPC's Altivec 128-bit SIMD competition has 128-bit hardware implementation at PowerPC G4's introduction i.e. no bulldust 64-bit SIMD hardware with fake 128-bit SIMD instruction set.