News AMD Posts First Official Fix for Windows 11 Bugs

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
My point is that this is exactly what swan said.
He didn't say screw benchmarks, he said during pandemic almost nobody will care about benchmarks.
I don't know why the "pandemic" bit is relevant at all in that statement... It's always been the case in my eyes.

Also, that's a self-fulfilled prophecy. Intel has engaged in dubious practices that don't really allow AMD to gain mind-share. Enthusiasts are the only ones that actually recognize AMD. At least, people that has a minimum interest in hardware and they're not just looking for a utilitarian thing. Kind of like in the car enthusiast world. You have people that only cares about commute and then you have people that buys based on their own specific brand/specs preference. The rest just buys whatever is cheaper and known-to-be reliable. This is to say, "regular" people recognizes Intel over AMD because of TV ads, found in more laptop offerings all over the place, better placement in shelves, etc. Sure, you can say AMD should've pushed harder to be on par with Intel, but their budgets are not even in the same league. That's why that statement is so daft, it's kind of dumb. Well, that's why that guy is no longer at Intel.

Again, I don't know what your point is.

Regards.
 

technofox

Distinguished
Nov 5, 2010
10
0
18,510
0
I just patched both of my Ryzen systems and they seem far more responsive than prior to patching - this includes updating drivers. I have to do more testing but even on Reddit, people are showing benchmarks that this issue has been addressed except for the L3 bandwidth issue but I cannot confirm the veracity of their claims.
 
I don't know why the "pandemic" bit is relevant at all in that statement... It's always been the case in my eyes.

Also, that's a self-fulfilled prophecy. Intel has engaged in dubious practices that don't really allow AMD to gain mind-share. Enthusiasts are the only ones that actually recognize AMD. At least, people that has a minimum interest in hardware and they're not just looking for a utilitarian thing. Kind of like in the car enthusiast world. You have people that only cares about commute and then you have people that buys based on their own specific brand/specs preference. The rest just buys whatever is cheaper and known-to-be reliable. This is to say, "regular" people recognizes Intel over AMD because of TV ads, found in more laptop offerings all over the place, better placement in shelves, etc. Sure, you can say AMD should've pushed harder to be on par with Intel, but their budgets are not even in the same league. That's why that statement is so daft, it's kind of dumb. Well, that's why that guy is no longer at Intel.

Again, I don't know what your point is.

Regards.
[/QUOTE]
No, Intel marketing discredits the benchmarks they were praising previously and the sheep follow suit :D
Was this statement you made for something else and not about swans comments that I linked to?
If so please provide a link because that interview with swan is the only thing that comes to mind.
If it was about swan then you just spend your last two post supporting his comments.
 
Was this statement you made for something else and not about swans comments that I linked to?
If so please provide a link because that interview with swan is the only thing that comes to mind.
If it was about swan then you just spend your last two post supporting his comments.
It is not only that bit you quoted, which I was also referring to.

When AMD started using Cinebench to show their improvements in their Ryzen line, Intel took a rather embarrassing step back by saying "Cinebench is not relevant". There's also the kerfuffle with the other benchmark suit I can't remember the name of, from which it was either AMD or Intel that bailed from them. Was it PassMark? I'm too lazy to Google it. This all happened when Ryzen started making noise and started to annoy Intel.

So, while they show numbers and whatnot, on the other hand they discredit numbers when they don't suit them. That's just scummy and stupid.

EDIT: I Googled it anyway, lol
https://www.techspot.com/news/44386-amd-via-nvidia-quit-benchmark-group-due-to-intel-bias.html
EDIT2: Ah, this one was AMD whining about GPUs not getting representation in the benchmark. Good riddance, I guess.

Regards.
 
Last edited:
It is not only that bit you quoted, which I was also referring to.

When AMD started using Cinebench to show their improvements in their Ryzen line, Intel took a rather embarrassing step back by saying "Cinebench is not relevant". There's also the kerfuffle with the other benchmark suit I can't remember the name of, from which it was either AMD or Intel that bailed from them. Was it PassMark? I'm too lazy to Google it. This all happened when Ryzen started making noise and started to annoy Intel.
While intel has used cinebench to show performance, the same way as tons of other software, they never went out to say that it is especially important so how is it a step back?
You are joining this in you head with the fact that people on forums were riding around on that cinebench was compiled with the intel compiler. So everybody though that intel had their hands in that so that must mean that intel considers it important.
This is fan fiction and not what happened.

Was this the benchmark thing you mentioned?
This benchmark is bad because it is measuring raw power...¯\(ツ)
AMD says Intel's use of the SYSmark benchmarking tool unfairly favours Intel products by putting too much emphasis on raw CPU performance.
 
While intel has used cinebench to show performance, the same way as tons of other software, they never went out to say that it is especially important so how is it a step back?
You are joining this in you head with the fact that people on forums were riding around on that cinebench was compiled with the intel compiler. So everybody though that intel had their hands in that so that must mean that intel considers it important.
This is fan fiction and not what happened.

Was this the benchmark thing you mentioned?
This benchmark is bad because it is measuring raw power...¯\(ツ)


Sigh...

This is an expansion to your original link: https://www.techradar.com/news/intel-doesnt-want-its-cpus-to-be-judged-just-on-benchmarks-but-also-broader-benefits

I'll stop here.

Regards.
 
Sigh...
Other than the authors personal opinions this article doesn't have a shred more info then the other one.
Although the other main thrust of Swan’s YouTube speech, as we mentioned at the outset, is to stop focusing on benchmarks, and given the current coronavirus situation, instead look more to broader benefits
 

Dr3ams

Great
Sep 29, 2021
29
31
60
0
Windows 11 is a piece of crap OS and not just because of the hardware instabilities. This is Microsoft's version of a Mac ecosystem.

I tried it one day and the next day rolled back my system to Windows 10. I don't think Microsoft is going to pull their head out of their a$$ before 2025, so here's hoping that Linux will get full gaming support by then.
 
My point is that this is exactly what swan said.
He didn't say screw benchmarks, he said during pandemic almost nobody will care about benchmarks.
True. With everybody selling out of everything, finding any decent high end laptop is near impossible. Even middle of the road ones are hard to find.. (I'm not sure about entry level)

My brother n law had 5 orders canceled on him because they were sold out. He needed it for his son's college. He finally had to pay $2400 for a trash gigabyte model that was way overpriced for the specs. It's going to be loaded down with bloatware, have a short battery life and cost way too much.
 
That's just Windows 11, shows worst latency on Intel 10,11th gen
https://ibb.co/ZcWwQh0

For my Zen2, 3900x with 3800cl14 bdie, I don't see a huge difference...
win10 62.9 ns- win11 63.5ns
This is due to mitigations to prevent adjacent side line cache poisoning similar to Spectre like attacks. (Although most attacks focused on the internals of branch predictors. But other similar methods can be employed)

I believe 13th gen and Zen 4 Pro models are supposed to use encrypted thread enclaves. That means each Code segment will get it's own encryption key. So if there is cache poisoning, it will decode to garbage once "decryption" is engaged. Think of it this way:

Code A (valid code): I'm speaking French. I'll use the CPU later to translate me to English
Code B (Malicious code): I'm speaking German. I'll inject my German code into your French Code.

CPU: Why is there german code inside a french program.? Abort with security exception.
 
Reactions: wifiburger

ASK THE COMMUNITY

TRENDING THREADS