Skybuck Flying <skybuckflying@gmail.com> writes:
>AMD HAD 3 YEARS TO FIX THIS SHIT ?!

Spectre V1 has been reported to them in June 2017, so it's now over
5.5 years. Even the public was informed more than five years ago.

>WHY IS NOT FIXED YET ?!

Apparently they think that customers don't care enough to make it a
competetive advantage to fix it (or a competetive disadvantage if the
competitor fixes it and they don't).

Alternative theory: A three-letter agency pressures all major CPU
manufacturers not to fix it. Counterargument: They could just as well
pressure them to provide backdoors to the ME/PSP or the like, which
would probably be more convenient for them.

- anton
--
'Anyone trying for "industrial quality" ISA should avoid undefined behavior.'
Mitch Alsup, <c17fcd89-f024-40e7-a594-88a85ac10d20o@googlegroups.com>

On Wednesday, February 22, 2023 at 10:45:52 AM UTC-6, Anton Ertl wrote:
> Skybuck Flying <skybuc...@gmail.com> writes:
> >AMD HAD 3 YEARS TO FIX THIS SHIT ?!
> Spectre V1 has been reported to them in June 2017, so it's now over
> 5.5 years. Even the public was informed more than five years ago.
<
> >WHY IS NOT FIXED YET ?!
<
> Apparently they think that customers don't care enough to make it a
> competetive advantage to fix it (or a competetive disadvantage if the
> competitor fixes it and they don't).
<
And how would they know their customers level of care ??
Did AMD lose market share to Intel or ARM in the last 5 years or not ??
{none of AMD, Intel, or ARM have addressed the underlying problem}.
RISC-V has not addressed the problems, either.
<
So, there is no evidence that consumers in the market gives a flying frick !!
>
> Alternative theory: A three-letter agency pressures all major CPU
> manufacturers not to fix it. Counterargument: They could just as well
> pressure them to provide backdoors to the ME/PSP or the like, which
> would probably be more convenient for them.
>
> - anton
> --
> 'Anyone trying for "industrial quality" ISA should avoid undefined behavior.'
> Mitch Alsup, <c17fcd89-f024-40e7...@googlegroups.com>

MitchAlsup <MitchAlsup@aol.com> writes:
>On Wednesday, February 22, 2023 at 10:45:52 AM UTC-6, Anton Ertl wrote:
>> Skybuck Flying <skybuc...@gmail.com> writes:
>> >AMD HAD 3 YEARS TO FIX THIS SHIT ?!
>> Spectre V1 has been reported to them in June 2017, so it's now over
>> 5.5 years. Even the public was informed more than five years ago.
><
>> >WHY IS NOT FIXED YET ?!
><
>> Apparently they think that customers don't care enough to make it a
>> competetive advantage to fix it (or a competetive disadvantage if the
>> competitor fixes it and they don't).
><
>And how would they know their customers level of care ??
>Did AMD lose market share to Intel or ARM in the last 5 years or not ??
>{none of AMD, Intel, or ARM have addressed the underlying problem}.
>RISC-V has not addressed the problems, either.
><
>So, there is no evidence that consumers in the market gives a flying frick !!g

The 'consumers in the market' have accepted the software
workarounds, albeit reluctantly due to the performance impacts.

On Wednesday, February 22, 2023 at 1:25:11 PM UTC-6, Scott Lurndal wrote:
> MitchAlsup <Mitch...@aol.com> writes:
> >On Wednesday, February 22, 2023 at 10:45:52 AM UTC-6, Anton Ertl wrote:
> >> Skybuck Flying <skybuc...@gmail.com> writes:
> >> >AMD HAD 3 YEARS TO FIX THIS SHIT ?!
> >> Spectre V1 has been reported to them in June 2017, so it's now over
> >> 5.5 years. Even the public was informed more than five years ago.
> ><
> >> >WHY IS NOT FIXED YET ?!
> ><
> >> Apparently they think that customers don't care enough to make it a
> >> competetive advantage to fix it (or a competetive disadvantage if the
> >> competitor fixes it and they don't).
> ><
> >And how would they know their customers level of care ??
> >Did AMD lose market share to Intel or ARM in the last 5 years or not ??
> >{none of AMD, Intel, or ARM have addressed the underlying problem}.
> >RISC-V has not addressed the problems, either.
> ><
> >So, there is no evidence that consumers in the market gives a flying frick !!g
>
> The 'consumers in the market' have accepted the software
> workarounds, albeit reluctantly due to the performance impacts.
<
Perhaps there is simply no alternative at the consumer level, and that the
giants see no particular gain for the effort it would take.