Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.os.linux.advocacy > #340948 > unrolled thread

Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads

Started byFritz Wuehler <fritz@spamexpire-201601.rodent.frell.theremailer.net>
First post2016-01-19 01:50 +0100
Last post2016-01-30 10:03 +0100
Articles 20 on this page of 106 — 23 participants

Back to article view | Back to comp.os.linux.advocacy


Contents

  Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Fritz Wuehler <fritz@spamexpire-201601.rodent.frell.theremailer.net> - 2016-01-19 01:50 +0100
    Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Mr. Man-wai Chang" <toylet.toylet@gmail.com> - 2016-01-19 16:17 +0800
    Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads JJ <jj4public@vfemail.net> - 2016-01-19 16:04 +0700
      Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Mr. Man-wai Chang" <toylet.toylet@gmail.com> - 2016-01-19 17:11 +0800
        Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Paul <nospam@needed.com> - 2016-01-19 06:27 -0500
          Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Mr. Man-wai Chang" <toylet.toylet@gmail.com> - 2016-01-19 20:55 +0800
        Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Anonymous Remailer (austria)" <mixmaster@remailer.privacy.at> - 2016-01-19 22:55 +0100
          Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads GreyCloud <mist@cumulus.com> - 2016-01-19 15:10 -0700
            Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Nomen Nescio <nobody@dizum.com> - 2016-01-20 03:30 +0100
              Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Slimer <.m@nsn.s> - 2016-01-20 09:32 -0500
                Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Ezekiel" <zeke@nosuchemail.com> - 2016-01-20 15:05 -0500
                  Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Slimer <.m@nsn.s> - 2016-01-20 15:57 -0500
                    Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads G. Morgan <sealteam6@osama-is-dead.net> - 2016-01-31 00:28 -0600
                      Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Paul <nospam@needed.com> - 2016-01-31 02:24 -0500
                  Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Jeremy Bentham <nobody@anemone.mooo.com> - 2016-01-20 23:47 +0100
                    AMD vs. Intel vallor <vallor@cultnix.org> - 2016-01-20 23:21 +0000
                      Re: AMD vs. Intel chrisv <chrisv@nospam.invalid> - 2016-01-21 06:48 -0600
                        Re: AMD vs. Intel Anonymous <anonymous@hoi-polloi.org> - 2016-01-21 19:30 +0100
                          Re: AMD vs. Intel Slimer <.m@nsn.s> - 2016-01-21 14:51 -0500
                            Re: AMD vs. Intel Melzzzzz <mel@zzzzz.com> - 2016-01-21 21:18 +0100
                              Re: AMD vs. Intel Slimer <.m@nsn.s> - 2016-01-21 17:30 -0500
                                Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-21 20:42 -0700
                                  Re: AMD vs. Intel Anonymous <anonymous@hoi-polloi.org> - 2016-01-22 07:15 +0100
                                    Re: AMD vs. Intel Jeremy Bentham <nobody@anemone.mooo.com> - 2016-01-22 08:48 +0100
                                    Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-22 15:24 -0700
                                      Re: AMD vs. Intel Peter Köhlmann <peter-koehlmann@t-online.de> - 2016-01-22 23:34 +0100
                                        Re: AMD vs. Intel Fritz Wuehler <fritz@spamexpire-201601.rodent.frell.theremailer.net> - 2016-01-23 01:58 +0100
                                          Re: AMD vs. Intel DFS <nospam@dfs.com> - 2016-01-22 23:08 -0500
                                            Re: AMD vs. Intel Slimer <.m@nsn.s> - 2016-01-23 09:51 -0500
                                        Re: AMD vs. Intel Anonymous <nobody@remailer.paranoici.org> - 2016-01-23 01:47 +0000
                                          Re: AMD vs. Intel Takuya Saitoh <taka0038@gmail.com> - 2016-01-22 19:50 -0800
                                            Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-23 10:22 -0700
                                            Re: AMD vs. Intel JEDIDIAH <jedi@nomad.mishnet> - 2016-01-23 17:36 -0600
                                              Re: AMD vs. Intel chrisv <chrisv@nospam.invalid> - 2016-01-25 07:55 -0600
                                          Re: AMD vs. Intel Peter Köhlmann <peter-koehlmann@t-online.de> - 2016-01-23 11:24 +0100
                                            Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-23 10:23 -0700
                                              Re: AMD vs. Intel Peter Köhlmann <peter-koehlmann@t-online.de> - 2016-01-23 18:37 +0100
                                                Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-24 10:23 -0700
                                                  Re: AMD vs. Intel Peter Köhlmann <peter-koehlmann@t-online.de> - 2016-01-24 23:23 +0100
                                                    Re: AMD vs. Intel Anonymous <anonymous@hoi-polloi.org> - 2016-01-25 05:04 +0100
                                                    Re: AMD vs. Intel chrisv <chrisv@nospam.invalid> - 2016-01-25 07:49 -0600
                                                      Re: AMD vs. Intel DFS <nospam@dfs.com> - 2016-01-25 09:15 -0500
                                                      Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-25 11:06 -0700
                                                    Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-25 11:05 -0700
                                                    Re: AMD vs. Intel Anonymous <nobody@remailer.paranoici.org> - 2016-01-26 08:02 +0000
                                                Re: AMD vs. Intel Fritz Wuehler <fritz@spamexpire-201601.rodent.frell.theremailer.net> - 2016-01-24 19:06 +0100
                                                  Re: AMD vs. Intel Slimer <.m@nsn.s> - 2016-01-24 16:14 -0500
                                            Re: AMD vs. Intel Anonymous <anonymous@hoi-polloi.org> - 2016-01-24 15:38 +0100
                                              Re: AMD vs. Intel Peter Köhlmann <peter-koehlmann@t-online.de> - 2016-01-24 15:59 +0100
                                                Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-24 10:25 -0700
                                                Re: AMD vs. Intel JEDIDIAH <jedi@nomad.mishnet> - 2016-01-25 08:11 -0600
                                                  Re: AMD vs. Intel Slimer <.m@nsn.s> - 2016-01-25 10:55 -0500
                                                  Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-25 11:07 -0700
                                                  Re: AMD vs. Intel Anonymous <anonymous@hoi-polloi.org> - 2016-01-26 01:58 +0100
                                                    Re: AMD vs. Intel Melzzzzz <mel@zzzzz.com> - 2016-01-26 02:05 +0100
                                                      Re: AMD vs. Intel Peter Köhlmann <peter-koehlmann@t-online.de> - 2016-01-26 11:24 +0100
                                                        Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-27 12:17 -0700
                                                        Re: AMD vs. Intel Anonymous <nobody@remailer.paranoici.org> - 2016-01-27 21:58 +0000
                                                      Re: AMD vs. Intel Jeremy Bentham <nobody@anemone.mooo.com> - 2016-01-26 19:55 +0100
                                        Re: AMD vs. Intel Jeremy Bentham <nobody@anemone.mooo.com> - 2016-01-23 06:20 +0100
                                  Re: AMD vs. Intel Slimer <.m@nsn.s> - 2016-01-22 10:13 -0500
                                    Re: AMD vs. Intel Paul <nospam@needed.com> - 2016-01-22 11:57 -0500
                                    Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-22 15:27 -0700
                                      Re: AMD vs. Intel Slimer <.m@nsn.s> - 2016-01-23 09:37 -0500
                                        Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-23 10:28 -0700
                                          Re: AMD vs. Intel Nomen Nescio <nobody@dizum.com> - 2016-01-24 06:21 +0100
                                            Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-24 10:29 -0700
                                      Re: AMD vs. Intel Paul <nospam@needed.com> - 2016-01-23 13:06 -0500
                                        Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-24 10:36 -0700
                              Re: AMD vs. Intel chrisv <chrisv@nospam.invalid> - 2016-01-22 06:59 -0600
                                Re: AMD vs. Intel GreyCloud <mist@cumulus.com> - 2016-01-22 15:30 -0700
                              Re: AMD vs. Intel chrisv <chrisv@nospam.invalid> - 2016-01-22 07:25 -0600
                      Re: AMD vs. Intel Melzzzzz <mel@zzzzz.com> - 2016-01-21 21:23 +0100
                        Re: AMD vs. Intel Slimer <.m@nsn.s> - 2016-01-21 17:40 -0500
                        Re: AMD vs. Intel DFS <nospam@dfs.com> - 2016-01-21 18:28 -0500
                          Re: AMD vs. Intel Melzzzzz <mel@zzzzz.com> - 2016-01-22 00:59 +0100
                            Re: AMD vs. Intel DFS <nospam@dfs.com> - 2016-01-23 11:20 -0500
                              Re: AMD vs. Intel "Ezekiel" <zeke@nosuchemail.com> - 2016-01-24 19:18 -0500
                    Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Ezekiel" <zeke@nosuchemail.com> - 2016-01-21 10:40 -0500
                      Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads PAS <abc@doremi.net> - 2016-01-21 12:16 -0500
                        Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads chrisv <chrisv@nospam.invalid> - 2016-01-21 11:26 -0600
                          Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads chrisv <chrisv@nospam.invalid> - 2016-01-21 11:31 -0600
                          Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Anonymous <anonymous@hoi-polloi.org> - 2016-01-21 23:42 +0100
                        Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Slimer <.m@nsn.s> - 2016-01-21 13:08 -0500
                      Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Nomen Nescio <nobody@dizum.com> - 2016-01-21 22:31 +0100
                Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Anonymous <nobody@remailer.paranoici.org> - 2016-01-20 20:29 +0000
              Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads GreyCloud <mist@cumulus.com> - 2016-01-20 11:53 -0700
                Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Anonymous <nobody@remailer.paranoici.org> - 2016-01-20 20:59 +0000
                  Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads GreyCloud <mist@cumulus.com> - 2016-01-20 14:29 -0700
                    Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Anonymous <nobody@remailer.paranoici.org> - 2016-01-21 05:44 +0000
                      Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads GreyCloud <mist@cumulus.com> - 2016-01-21 14:17 -0700
                        Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Nomen Nescio <nobody@dizum.com> - 2016-01-22 04:31 +0100
                          Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads GreyCloud <mist@cumulus.com> - 2016-01-21 20:44 -0700
                            Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Anonymous <anonymous@hoi-polloi.org> - 2016-01-22 08:35 +0100
                            Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Nomen Nescio <nobody@dizum.com> - 2016-01-23 06:13 +0100
                            Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Anonymous <anonymous@hoi-polloi.org> - 2016-01-23 10:03 +0100
                            Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Fritz Wuehler <fritz@spamexpire-201601.rodent.frell.theremailer.net> - 2016-01-23 15:39 +0100
                            Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Jeremy Bentham <nobody@anemone.mooo.com> - 2016-01-24 05:06 +0100
                            Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Anonymous Remailer (austria)" <mixmaster@remailer.privacy.at> - 2016-01-24 10:39 +0100
                            Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Anonymous Remailer (austria)" <mixmaster@remailer.privacy.at> - 2016-01-25 00:22 +0100
      Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Paul <nospam@needed.com> - 2016-01-19 05:09 -0500
        Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Diesel <me@privacy.net> - 2016-01-30 08:07 +0000
      Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Nomen Nescio <nobody@dizum.com> - 2016-01-19 22:16 +0100
        Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex  Workloads Euclides Zoto <euclides_zoto@optimum.net> - 2016-01-18 18:20 -0800
      Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads Diesel <me@privacy.net> - 2016-01-30 08:07 +0000
        Re: Intel Skylake Processor Bug Causing PCs To Freeze In Complex Workloads "Anonymous Remailer (austria)" <mixmaster@remailer.privacy.at> - 2016-01-30 10:03 +0100

Page 4 of 6 — ← Prev page 1 2 3 [4] 5 6  Next page →


#341370 — Re: AMD vs. Intel

FromSlimer <.m@nsn.s>
Date2016-01-22 10:13 -0500
SubjectRe: AMD vs. Intel
Message-ID<n7tgpn$ljv$1@dont-email.me>
In reply to#341306
On 2016-01-21 10:42 PM, GreyCloud wrote:
> On 01/21/16 15:30, Slimer wrote:
>> On 2016-01-21 3:18 PM, Melzzzzz wrote:
>>> On Thu, 21 Jan 2016 14:51:14 -0500
>>> Slimer<.m@nsn.s>  wrote:
>>>
>>>> However, we're ignoring another possibility: if x86-64 becomes too
>>>> expensive to buy into, PowerPC platforms might return to prominence
>>>> and computers such as the AmigaOne might actually be noticed by the
>>>> marketplace.
>>>
>>> ARM is big contender not PowerPC...
>>
>> Dude, it doesn't matter. Just because ARM is doing well right now
>> doesn't mean that PowerPC ceased to exist.
>>
> The PowerPC series is at G7 level now and IBM uses them soley for servers.

I can imagine. However, a response to x64 becoming much more expensive
COULD be to settle for other platforms. Intel making its chips
inaccessible to regular people because of high prices could result in a
company like IBM slashing the prices of their RISC processors and the
people behind AmigaOne and MorphOS taking advantage to sell machines
which can do everything that you would want to do at a fraction of the
price a Windows x64 set-up would cost you. I'm not saying that it WILL
happen, I'm saying that it COULD happen.


-- 
Slimer
EFF & OpenMedia member / IFAW, Mozilla & PETA supporter
"Everything seems to work fine, except that occasionally everything
freezes." - Another one of GNU/Linux's many victims
"Sound wasn't even a gauranteed thing on a PC in the 90s." - Linux
apologist JEDIDIAH explaining why sound never worked/works right in Linux

[toc] | [prev] | [next] | [standalone]


#341376 — Re: AMD vs. Intel

FromPaul <nospam@needed.com>
Date2016-01-22 11:57 -0500
SubjectRe: AMD vs. Intel
Message-ID<n7tmt7$ftk$1@dont-email.me>
In reply to#341370
Slimer wrote:
> On 2016-01-21 10:42 PM, GreyCloud wrote:
>> On 01/21/16 15:30, Slimer wrote:
>>> On 2016-01-21 3:18 PM, Melzzzzz wrote:
>>>> On Thu, 21 Jan 2016 14:51:14 -0500
>>>> Slimer<.m@nsn.s>  wrote:
>>>>
>>>>> However, we're ignoring another possibility: if x86-64 becomes too
>>>>> expensive to buy into, PowerPC platforms might return to prominence
>>>>> and computers such as the AmigaOne might actually be noticed by the
>>>>> marketplace.
>>>> ARM is big contender not PowerPC...
>>> Dude, it doesn't matter. Just because ARM is doing well right now
>>> doesn't mean that PowerPC ceased to exist.
>>>
>> The PowerPC series is at G7 level now and IBM uses them soley for servers.
> 
> I can imagine. However, a response to x64 becoming much more expensive
> COULD be to settle for other platforms. Intel making its chips
> inaccessible to regular people because of high prices could result in a
> company like IBM slashing the prices of their RISC processors and the
> people behind AmigaOne and MorphOS taking advantage to sell machines
> which can do everything that you would want to do at a fraction of the
> price a Windows x64 set-up would cost you. I'm not saying that it WILL
> happen, I'm saying that it COULD happen.

I don't think you have any idea at all, what's going on.

IBM got rid of its fab. Now, do you think that will change
what kind of a company they are ? Where are your cheap
RISC processors going to come from now ? If they (eventually)
use an outside fab, their chips, ounce for ounce, will cost
the same as everyone else.

http://money.cnn.com/2014/10/20/investing/ibm-sale-earnings/index.html

You may not believe it, but IBM likes "high-margin" projects,
and selling you cheap RISC chips is simply not in their genes.
If you want *anything* from IBM, you will pay dearly for it.

*******

Owning a fab (as Intel does) is like "drug addiction".

Each time the fab needs a $2 billion to $5 billion overhaul,
something has to pay for that. So rather than Intel
producing "whizzy technology because its fun", instead
they're slaves to a balance sheet, struggling to pay
for their next drug fix. If there is even a slight
disturbance in sales volume, it can spell disaster
for their "whole machine". These businesses are
precarious, and getting more so as time passes.

With the Dramurai, the game was "fixed", as some
memory manufacturing companies (companies running their
own 12" wafer lines) were being propped up by money provided
by national governments. However, during the last price dip,
one of the companies got out of the business. So even national
governments don't have deep enough pockets to interfere in the
market forever.

For me, the biggest fear from a strategic point of view,
is not having fabs on our own soil. I think you can understand
why that might be important. But no matter how
important it might be, someone has to pay for
those. And as long as business controls them,
they'll "move freely over the surface of the
planet". Economic forces will determine winners
and losers. TSMC, providing services for multiple
fabless companies, seeks to even out its chip load,
by dealing with multiple designs at the same time.
Unlike an Intel fab that (as far as I know) handles
only Intel business at the moment.

IBM entertained supplying chips to third parties.
They made chips for us. We had their silicon library on
disk, and I could look at the performance characteristics
of their chip tech. But, they couldn't even out their
supply bumps, even by trying "to be their own TSMC".
So if anything, you end up suspecting that there is
just no good business in owning a fab. Being a drug
addict, sucks.

In some cases, the chip fab movement is literal.
Like the chip manufacturing plant that they
put on a barge and moved from North America
to Ireland. I checked one day with a supplier,
to see if I could get a certain chip, and
I was told "those are out of stock for the
next 46 weeks - the chip plant is on a barge somewhere,
out on the ocean". So in some cases, chip plants
actually do move, lock, stock and barrel. For the
life of me, I couldn't figure out how a "barge"
was the best way to do that. Shurely they could have
Fedexed the thing, and had it overnight :-)

    Paul

[toc] | [prev] | [next] | [standalone]


#341396 — Re: AMD vs. Intel

FromGreyCloud <mist@cumulus.com>
Date2016-01-22 15:27 -0700
SubjectRe: AMD vs. Intel
Message-ID<n7ua94$t4$1@dont-email.me>
In reply to#341370
On 01/22/16 08:13, Slimer wrote:
> On 2016-01-21 10:42 PM, GreyCloud wrote:
>> On 01/21/16 15:30, Slimer wrote:
>>> On 2016-01-21 3:18 PM, Melzzzzz wrote:
>>>> On Thu, 21 Jan 2016 14:51:14 -0500
>>>> Slimer<.m@nsn.s>   wrote:
>>>>
>>>>> However, we're ignoring another possibility: if x86-64 becomes too
>>>>> expensive to buy into, PowerPC platforms might return to prominence
>>>>> and computers such as the AmigaOne might actually be noticed by the
>>>>> marketplace.
>>>>
>>>> ARM is big contender not PowerPC...
>>>
>>> Dude, it doesn't matter. Just because ARM is doing well right now
>>> doesn't mean that PowerPC ceased to exist.
>>>
>> The PowerPC series is at G7 level now and IBM uses them soley for servers.
>
> I can imagine. However, a response to x64 becoming much more expensive
> COULD be to settle for other platforms. Intel making its chips
> inaccessible to regular people because of high prices could result in a
> company like IBM slashing the prices of their RISC processors and the
> people behind AmigaOne and MorphOS taking advantage to sell machines
> which can do everything that you would want to do at a fraction of the
> price a Windows x64 set-up would cost you. I'm not saying that it WILL
> happen, I'm saying that it COULD happen.
>
>
They should.  The markets, at least to me, are getting rather boring.
If I was the chairman of the board at IBM, I'd have fired the idiot that 
didn't try to cut the power consumption on the PowerPC line.  That 
would've kept IBM in contention with Intel and kept their profits up, 
which as I understand it now is falling.  Their single core G5 at the 
time did pretty good as one core, and I could only imagine what an 8 
core G5 with low power consumption could've done for the market.

[toc] | [prev] | [next] | [standalone]


#341435 — Re: AMD vs. Intel

FromSlimer <.m@nsn.s>
Date2016-01-23 09:37 -0500
SubjectRe: AMD vs. Intel
Message-ID<n8032p$b4a$1@dont-email.me>
In reply to#341396
On 2016-01-22 5:27 PM, GreyCloud wrote:
> On 01/22/16 08:13, Slimer wrote:
>> On 2016-01-21 10:42 PM, GreyCloud wrote:
>>> On 01/21/16 15:30, Slimer wrote:
>>>> On 2016-01-21 3:18 PM, Melzzzzz wrote:
>>>>> On Thu, 21 Jan 2016 14:51:14 -0500
>>>>> Slimer<.m@nsn.s>   wrote:
>>>>>
>>>>>> However, we're ignoring another possibility: if x86-64 becomes too
>>>>>> expensive to buy into, PowerPC platforms might return to prominence
>>>>>> and computers such as the AmigaOne might actually be noticed by the
>>>>>> marketplace.
>>>>>
>>>>> ARM is big contender not PowerPC...
>>>>
>>>> Dude, it doesn't matter. Just because ARM is doing well right now
>>>> doesn't mean that PowerPC ceased to exist.
>>>>
>>> The PowerPC series is at G7 level now and IBM uses them soley for
>>> servers.
>>
>> I can imagine. However, a response to x64 becoming much more expensive
>> COULD be to settle for other platforms. Intel making its chips
>> inaccessible to regular people because of high prices could result in a
>> company like IBM slashing the prices of their RISC processors and the
>> people behind AmigaOne and MorphOS taking advantage to sell machines
>> which can do everything that you would want to do at a fraction of the
>> price a Windows x64 set-up would cost you. I'm not saying that it WILL
>> happen, I'm saying that it COULD happen.
>>
>>
> They should.  The markets, at least to me, are getting rather boring.
> If I was the chairman of the board at IBM, I'd have fired the idiot that
> didn't try to cut the power consumption on the PowerPC line.  That
> would've kept IBM in contention with Intel and kept their profits up,
> which as I understand it now is falling.  Their single core G5 at the
> time did pretty good as one core, and I could only imagine what an 8
> core G5 with low power consumption could've done for the market.

It all depends on whether IBM is interested in getting back in the game
which made them a household name or whether they want to continue to
deny that they are essentially dying and that it is just a matter of
time before they are forgotten entirely. Intel WILL get arrogant and
WILL make mistakes which WILL alienate customers. Those customers MIGHT
go to AMD and pay for less expensive processors (I know that's what *I*
did for a time) but they might also be ready for an entirely different
experience without GNU/Linux or Windows. It would just be nice to see
IBM try to take advantage of the arrogance to sell a new platform
(debatably a better platform) with an entirely different operating
environment. It's unlikely, but it's fairly exciting to consider.

-- 
Slimer
EFF & OpenMedia member / IFAW, Conservative Party of Canada, Mozilla &
PETA supporter

[toc] | [prev] | [next] | [standalone]


#341455 — Re: AMD vs. Intel

FromGreyCloud <mist@cumulus.com>
Date2016-01-23 10:28 -0700
SubjectRe: AMD vs. Intel
Message-ID<n80d4f$ia1$1@dont-email.me>
In reply to#341435
On 01/23/16 07:37, Slimer wrote:
> On 2016-01-22 5:27 PM, GreyCloud wrote:
>> On 01/22/16 08:13, Slimer wrote:
>>> On 2016-01-21 10:42 PM, GreyCloud wrote:
>>>> On 01/21/16 15:30, Slimer wrote:
>>>>> On 2016-01-21 3:18 PM, Melzzzzz wrote:
>>>>>> On Thu, 21 Jan 2016 14:51:14 -0500
>>>>>> Slimer<.m@nsn.s>    wrote:
>>>>>>
>>>>>>> However, we're ignoring another possibility: if x86-64 becomes too
>>>>>>> expensive to buy into, PowerPC platforms might return to prominence
>>>>>>> and computers such as the AmigaOne might actually be noticed by the
>>>>>>> marketplace.
>>>>>>
>>>>>> ARM is big contender not PowerPC...
>>>>>
>>>>> Dude, it doesn't matter. Just because ARM is doing well right now
>>>>> doesn't mean that PowerPC ceased to exist.
>>>>>
>>>> The PowerPC series is at G7 level now and IBM uses them soley for
>>>> servers.
>>>
>>> I can imagine. However, a response to x64 becoming much more expensive
>>> COULD be to settle for other platforms. Intel making its chips
>>> inaccessible to regular people because of high prices could result in a
>>> company like IBM slashing the prices of their RISC processors and the
>>> people behind AmigaOne and MorphOS taking advantage to sell machines
>>> which can do everything that you would want to do at a fraction of the
>>> price a Windows x64 set-up would cost you. I'm not saying that it WILL
>>> happen, I'm saying that it COULD happen.
>>>
>>>
>> They should.  The markets, at least to me, are getting rather boring.
>> If I was the chairman of the board at IBM, I'd have fired the idiot that
>> didn't try to cut the power consumption on the PowerPC line.  That
>> would've kept IBM in contention with Intel and kept their profits up,
>> which as I understand it now is falling.  Their single core G5 at the
>> time did pretty good as one core, and I could only imagine what an 8
>> core G5 with low power consumption could've done for the market.
>
> It all depends on whether IBM is interested in getting back in the game
> which made them a household name or whether they want to continue to
> deny that they are essentially dying and that it is just a matter of
> time before they are forgotten entirely. Intel WILL get arrogant and
> WILL make mistakes which WILL alienate customers. Those customers MIGHT
> go to AMD and pay for less expensive processors (I know that's what *I*
> did for a time) but they might also be ready for an entirely different
> experience without GNU/Linux or Windows. It would just be nice to see
> IBM try to take advantage of the arrogance to sell a new platform
> (debatably a better platform) with an entirely different operating
> environment. It's unlikely, but it's fairly exciting to consider.
>
If IBM doesn't wake up and realize just how close to the precipice they 
are, then it will be only a matter of time for their demise. They're 
already losing money.  Their first attempt was to ditch Linux.  As I've 
said before, Linux branding is the kiss of death to a product line.
I don't know what IBM thinks they can do with going to Apple and buying 
their machines, after they arrogantly ignored Steve Jobs recommendation 
to lower the power consumption of their cpus.

AMD can get out of the doldrums if they utilized some of the design 
paradigms of the DEC Alpha team used.  After all, when DEC was around 
the Alphas were on top.

[toc] | [prev] | [next] | [standalone]


#341514 — Re: AMD vs. Intel

FromNomen Nescio <nobody@dizum.com>
Date2016-01-24 06:21 +0100
SubjectRe: AMD vs. Intel
Message-ID<dc952f5d02586c5fc15b196074fc4e16@dizum.com>
In reply to#341455
In article <n80d4f$ia1$1@dont-email.me>
GreyCloud <mist@cumulus.com> wrote:
>
> On 01/23/16 07:37, Slimer wrote:
> > On 2016-01-22 5:27 PM, GreyCloud wrote:
> >> On 01/22/16 08:13, Slimer wrote:
> >>> On 2016-01-21 10:42 PM, GreyCloud wrote:
> >>>> On 01/21/16 15:30, Slimer wrote:
> >>>>> On 2016-01-21 3:18 PM, Melzzzzz wrote:
> >>>>>> On Thu, 21 Jan 2016 14:51:14 -0500
> >>>>>> Slimer<.m@nsn.s>    wrote:
> >>>>>>
> >>>>>>> However, we're ignoring another possibility: if x86-64 becomes too
> >>>>>>> expensive to buy into, PowerPC platforms might return to prominence
> >>>>>>> and computers such as the AmigaOne might actually be noticed by the
> >>>>>>> marketplace.
> >>>>>>
> >>>>>> ARM is big contender not PowerPC...
> >>>>>
> >>>>> Dude, it doesn't matter. Just because ARM is doing well right now
> >>>>> doesn't mean that PowerPC ceased to exist.
> >>>>>
> >>>> The PowerPC series is at G7 level now and IBM uses them soley for
> >>>> servers.
> >>>
> >>> I can imagine. However, a response to x64 becoming much more expensive
> >>> COULD be to settle for other platforms. Intel making its chips
> >>> inaccessible to regular people because of high prices could result in a
> >>> company like IBM slashing the prices of their RISC processors and the
> >>> people behind AmigaOne and MorphOS taking advantage to sell machines
> >>> which can do everything that you would want to do at a fraction of the
> >>> price a Windows x64 set-up would cost you. I'm not saying that it WILL
> >>> happen, I'm saying that it COULD happen.
> >>>
> >>>
> >> They should.  The markets, at least to me, are getting rather boring.
> >> If I was the chairman of the board at IBM, I'd have fired the idiot that
> >> didn't try to cut the power consumption on the PowerPC line.  That
> >> would've kept IBM in contention with Intel and kept their profits up,
> >> which as I understand it now is falling.  Their single core G5 at the
> >> time did pretty good as one core, and I could only imagine what an 8
> >> core G5 with low power consumption could've done for the market.
> >
> > It all depends on whether IBM is interested in getting back in the game
> > which made them a household name or whether they want to continue to
> > deny that they are essentially dying and that it is just a matter of
> > time before they are forgotten entirely. Intel WILL get arrogant and
> > WILL make mistakes which WILL alienate customers. Those customers MIGHT
> > go to AMD and pay for less expensive processors (I know that's what *I*
> > did for a time) but they might also be ready for an entirely different
> > experience without GNU/Linux or Windows. It would just be nice to see
> > IBM try to take advantage of the arrogance to sell a new platform
> > (debatably a better platform) with an entirely different operating
> > environment. It's unlikely, but it's fairly exciting to consider.
> >
> If IBM doesn't wake up and realize just how close to the precipice they
> are, then it will be only a matter of time for their demise. They're
> already losing money.  Their first attempt was to ditch Linux.  As I've
> said before, Linux branding is the kiss of death to a product line.

IBM's arrogance is a heavy cross to drag around.  The major 
stupid thing they did was engage India to do development and 
support as a cost cutting measure.  More like a throat cutting 
measure.  India had to turn around and engage Americans to 
support IBM's India folly - on IBM's dime.

Customers HATE India support with a passion.  Really hate them.  
Won't even talk to them, let the company computing resources 
take a total shit and lose data kind of hate.

> I don't know what IBM thinks they can do with going to Apple and buying
> their machines, after they arrogantly ignored Steve Jobs recommendation
> to lower the power consumption of their cpus.

IBM is still inventing water cooled technology to compete in 
markets where a generic Intel box has 48 cores.  We can easily 
put ten of those in a rack and have 480 cores at our disposal - 
without water involved.

> AMD can get out of the doldrums if they utilized some of the design
> paradigms of the DEC Alpha team used.  After all, when DEC was around
> the Alphas were on top.

AMD did.  They used EV6, then in a stroke of genius, dumped it 
for NUMA.

[toc] | [prev] | [next] | [standalone]


#341543 — Re: AMD vs. Intel

FromGreyCloud <mist@cumulus.com>
Date2016-01-24 10:29 -0700
SubjectRe: AMD vs. Intel
Message-ID<n831gr$4mt$1@dont-email.me>
In reply to#341514
On 01/23/16 22:21, Nomen Nescio wrote:
> In article<n80d4f$ia1$1@dont-email.me>
> GreyCloud<mist@cumulus.com>  wrote:
>>
>> On 01/23/16 07:37, Slimer wrote:
>>> On 2016-01-22 5:27 PM, GreyCloud wrote:
>>>> On 01/22/16 08:13, Slimer wrote:
>>>>> On 2016-01-21 10:42 PM, GreyCloud wrote:
>>>>>> On 01/21/16 15:30, Slimer wrote:
>>>>>>> On 2016-01-21 3:18 PM, Melzzzzz wrote:
>>>>>>>> On Thu, 21 Jan 2016 14:51:14 -0500
>>>>>>>> Slimer<.m@nsn.s>     wrote:
>>>>>>>>
>>>>>>>>> However, we're ignoring another possibility: if x86-64 becomes too
>>>>>>>>> expensive to buy into, PowerPC platforms might return to prominence
>>>>>>>>> and computers such as the AmigaOne might actually be noticed by the
>>>>>>>>> marketplace.
>>>>>>>>
>>>>>>>> ARM is big contender not PowerPC...
>>>>>>>
>>>>>>> Dude, it doesn't matter. Just because ARM is doing well right now
>>>>>>> doesn't mean that PowerPC ceased to exist.
>>>>>>>
>>>>>> The PowerPC series is at G7 level now and IBM uses them soley for
>>>>>> servers.
>>>>>
>>>>> I can imagine. However, a response to x64 becoming much more expensive
>>>>> COULD be to settle for other platforms. Intel making its chips
>>>>> inaccessible to regular people because of high prices could result in a
>>>>> company like IBM slashing the prices of their RISC processors and the
>>>>> people behind AmigaOne and MorphOS taking advantage to sell machines
>>>>> which can do everything that you would want to do at a fraction of the
>>>>> price a Windows x64 set-up would cost you. I'm not saying that it WILL
>>>>> happen, I'm saying that it COULD happen.
>>>>>
>>>>>
>>>> They should.  The markets, at least to me, are getting rather boring.
>>>> If I was the chairman of the board at IBM, I'd have fired the idiot that
>>>> didn't try to cut the power consumption on the PowerPC line.  That
>>>> would've kept IBM in contention with Intel and kept their profits up,
>>>> which as I understand it now is falling.  Their single core G5 at the
>>>> time did pretty good as one core, and I could only imagine what an 8
>>>> core G5 with low power consumption could've done for the market.
>>>
>>> It all depends on whether IBM is interested in getting back in the game
>>> which made them a household name or whether they want to continue to
>>> deny that they are essentially dying and that it is just a matter of
>>> time before they are forgotten entirely. Intel WILL get arrogant and
>>> WILL make mistakes which WILL alienate customers. Those customers MIGHT
>>> go to AMD and pay for less expensive processors (I know that's what *I*
>>> did for a time) but they might also be ready for an entirely different
>>> experience without GNU/Linux or Windows. It would just be nice to see
>>> IBM try to take advantage of the arrogance to sell a new platform
>>> (debatably a better platform) with an entirely different operating
>>> environment. It's unlikely, but it's fairly exciting to consider.
>>>
>> If IBM doesn't wake up and realize just how close to the precipice they
>> are, then it will be only a matter of time for their demise. They're
>> already losing money.  Their first attempt was to ditch Linux.  As I've
>> said before, Linux branding is the kiss of death to a product line.
>
> IBM's arrogance is a heavy cross to drag around.  The major
> stupid thing they did was engage India to do development and
> support as a cost cutting measure.  More like a throat cutting
> measure.  India had to turn around and engage Americans to
> support IBM's India folly - on IBM's dime.
>

As I've said before, if I were the CEO I'd fire some really stupid 
people and go looking for some smart people and pay the price.
It would be worth it to them.

> Customers HATE India support with a passion.  Really hate them.
> Won't even talk to them, let the company computing resources
> take a total shit and lose data kind of hate.
>

I got some really stupid support from HP from some indian over there 
over an HP printer.  All he thought the solution was is to clean the 
print heads, when the problem was very dark color photo prints.
Eventually HP fixed their software, but it took a good year for them to 
wake up to the fact that it was their software that was messed up.

>> I don't know what IBM thinks they can do with going to Apple and buying
>> their machines, after they arrogantly ignored Steve Jobs recommendation
>> to lower the power consumption of their cpus.
>
> IBM is still inventing water cooled technology to compete in
> markets where a generic Intel box has 48 cores.  We can easily
> put ten of those in a rack and have 480 cores at our disposal -
> without water involved.
>
>> AMD can get out of the doldrums if they utilized some of the design
>> paradigms of the DEC Alpha team used.  After all, when DEC was around
>> the Alphas were on top.
>
> AMD did.  They used EV6, then in a stroke of genius, dumped it
> for NUMA.
>
I haven't heard that part.

[toc] | [prev] | [next] | [standalone]


#341461 — Re: AMD vs. Intel

FromPaul <nospam@needed.com>
Date2016-01-23 13:06 -0500
SubjectRe: AMD vs. Intel
Message-ID<n80fab$qrn$1@dont-email.me>
In reply to#341396
GreyCloud wrote:

> They should.  The markets, at least to me, are getting rather boring.
> If I was the chairman of the board at IBM, I'd have fired the idiot that 
> didn't try to cut the power consumption on the PowerPC line.  That 
> would've kept IBM in contention with Intel and kept their profits up, 
> which as I understand it now is falling.  Their single core G5 at the 
> time did pretty good as one core, and I could only imagine what an 8 
> core G5 with low power consumption could've done for the market.

Low power consumption is a side effect of the development process.

Even Intel, owning its own fab and having teams capable
of redefining transistor stacks multiple times over the years,
they still can't hit the power target they want, to compete with ARM.
They can only get "close". (Cheating with scenario design power
versus TDP, throttling when the chip gets hot and so on.)

Of course they "try to cut the power consumption",
every day when they come into work. But the work is
"hard". You can't wave a magic wand, and change your
position in the industry.

The architecture you select, defines "the gorilla you
drag around on your back". And as good as Intel is,
they'd can't squeeze enough advantage out of their fab,
to compensate for "X86/X64".

The fabs use equipment made by other companies. You
need lithography equipment. No silicon company tries
to make all the equipment in their production facility.
They need "nice machines" fully automated, ready for
unattended operation. All they can make in their own
labs, is spaghettio machines with wires and cables all
over the place, vacuum pumps, glass christmas trees,
stuff all over the place, and a team of PhDs adjusting
the knobs every day to keep it running. External companies
make the real production equipment. And the industry works
together to set the next process node, so you can actually
build and purchase the items you need to build a fab.

http://www.ibtimes.com/ibm-major-breakthrough-computer-chips-featuring-tiny-7-nanometer-transistors-2002277

   "IBM said Thursday said it had successfully created
    computer chips with microscopic 7-nanometer transistors,
    a major breakthrough for the tech industry that should
    allow hardware manufacturers to continue making faster
    computers and smartphones over the next few years.

    By creating tiny transistors of only 7 nm (1 nm equals
    1 billionth of a meter) IBM is able to fit more transistors
    onto its chips than possible with the current 22-nm and 14-nm
    transistor chips.

    "blah blah blah scaling to 7 nm and beyond," said Arvind Krishna,
    senior vice president and director of IBM Research, in a statement.
    "That’s why IBM has remained committed to an aggressive basic
    research agenda that continually pushes the limits of semiconductor
    technology."

    For decades, transistors have gotten smaller, allowing for more
    powerful computer chips to be created. This is known as Moore's Law,
    but blah blah blah for at least a few more years.

    To create the 7 nm transistors, IBM teamed up with other chipmakers,
    including Samsung, and used a silicon-germanium mixture instead
    of just silicon. It also had to develop new manufacturing processes
    to achieve working transistors.

    For IBM, the next challenge will be to figure out   <----
    a way to create these chips at a mass scale so
    they can be commercially viable.
   "

So the last paragraph is key. Yes, they made a breakthrough.
These don't happen every day. And someone needs to make
the equipment, refine the process, so that they can
build even a single football field sized fab to do it.

Even with the example photo of a test chip (perhaps done
with an E-beam), the rest of the industry will work to
make the production method cheap enough to actually
build a fab with. So if the chairman of IBM said
"make me a low power PowerPC", the fab guy would
say "not until we have production facilities
to do it". Making single wafers in your basement
with the PhDs, is not a production facility.

So IBM solved the problem of what to make substrates
or wafers out of. But how the transistors get transferred
to masks, for all those process steps that take 12 weeks
to complete, others are still working on this. It appears
they might still be using optical techniques in the
ultraviolet.

http://semiengineering.com/eda-races-to-7nm-despite-litho-uncertainties/

See, fun stuff. Still using optical methods
after all these years.

https://en.wikipedia.org/wiki/Multiple_patterning

IBM can use much more expensive (lab prototype)
methods, like an E-beam writer, to make their test
chip. But whatever method they used, might be
too slow or too expensive for production purposes.

A lot of teamwork goes into this stuff. Even though
the companies compete, they can only keep the lithography
machine company in business, if he makes the same
equipment for everybody.

    Paul

[toc] | [prev] | [next] | [standalone]


#341544 — Re: AMD vs. Intel

FromGreyCloud <mist@cumulus.com>
Date2016-01-24 10:36 -0700
SubjectRe: AMD vs. Intel
Message-ID<n831tg$796$1@dont-email.me>
In reply to#341461
On 01/23/16 11:06, Paul wrote:
> GreyCloud wrote:
>
>> They should. The markets, at least to me, are getting rather boring.
>> If I was the chairman of the board at IBM, I'd have fired the idiot
>> that didn't try to cut the power consumption on the PowerPC line. That
>> would've kept IBM in contention with Intel and kept their profits up,
>> which as I understand it now is falling. Their single core G5 at the
>> time did pretty good as one core, and I could only imagine what an 8
>> core G5 with low power consumption could've done for the market.
>
> Low power consumption is a side effect of the development process.
>
> Even Intel, owning its own fab and having teams capable
> of redefining transistor stacks multiple times over the years,
> they still can't hit the power target they want, to compete with ARM.
> They can only get "close". (Cheating with scenario design power
> versus TDP, throttling when the chip gets hot and so on.)
>
> Of course they "try to cut the power consumption",
> every day when they come into work. But the work is
> "hard". You can't wave a magic wand, and change your
> position in the industry.
>
> The architecture you select, defines "the gorilla you
> drag around on your back". And as good as Intel is,
> they'd can't squeeze enough advantage out of their fab,
> to compensate for "X86/X64".
>
> The fabs use equipment made by other companies. You
> need lithography equipment. No silicon company tries
> to make all the equipment in their production facility.
> They need "nice machines" fully automated, ready for
> unattended operation. All they can make in their own
> labs, is spaghettio machines with wires and cables all
> over the place, vacuum pumps, glass christmas trees,
> stuff all over the place, and a team of PhDs adjusting
> the knobs every day to keep it running. External companies
> make the real production equipment. And the industry works
> together to set the next process node, so you can actually
> build and purchase the items you need to build a fab.

I know, I live down the street from a company that was bought out for 
the expertise that the locals had brought to the company.
It is a business unto itself and is very successful.

>
> http://www.ibtimes.com/ibm-major-breakthrough-computer-chips-featuring-tiny-7-nanometer-transistors-2002277
>
>
> "IBM said Thursday said it had successfully created
> computer chips with microscopic 7-nanometer transistors,
> a major breakthrough for the tech industry that should
> allow hardware manufacturers to continue making faster
> computers and smartphones over the next few years.
>
> By creating tiny transistors of only 7 nm (1 nm equals
> 1 billionth of a meter) IBM is able to fit more transistors
> onto its chips than possible with the current 22-nm and 14-nm
> transistor chips.
>
> "blah blah blah scaling to 7 nm and beyond," said Arvind Krishna,
> senior vice president and director of IBM Research, in a statement.
> "That’s why IBM has remained committed to an aggressive basic
> research agenda that continually pushes the limits of semiconductor
> technology."
>
> For decades, transistors have gotten smaller, allowing for more
> powerful computer chips to be created. This is known as Moore's Law,
> but blah blah blah for at least a few more years.
>
> To create the 7 nm transistors, IBM teamed up with other chipmakers,
> including Samsung, and used a silicon-germanium mixture instead
> of just silicon. It also had to develop new manufacturing processes
> to achieve working transistors.
>
> For IBM, the next challenge will be to figure out <----
> a way to create these chips at a mass scale so
> they can be commercially viable.
> "

It always boils down to production problems.  Once those are solved, 
they may be able to lift their profits off the deck, as long as someone 
else doesn't beat them to the punch.

>
> So the last paragraph is key. Yes, they made a breakthrough.
> These don't happen every day. And someone needs to make
> the equipment, refine the process, so that they can
> build even a single football field sized fab to do it.
>
> Even with the example photo of a test chip (perhaps done
> with an E-beam), the rest of the industry will work to
> make the production method cheap enough to actually
> build a fab with. So if the chairman of IBM said
> "make me a low power PowerPC", the fab guy would
> say "not until we have production facilities
> to do it". Making single wafers in your basement
> with the PhDs, is not a production facility.
>

That is true, but IBM should have told Apple what they were asking for.
Intel beat IBM to the punch and most likely was the straw that broke the 
camels back.

> So IBM solved the problem of what to make substrates
> or wafers out of. But how the transistors get transferred
> to masks, for all those process steps that take 12 weeks
> to complete, others are still working on this. It appears
> they might still be using optical techniques in the
> ultraviolet.
>
> http://semiengineering.com/eda-races-to-7nm-despite-litho-uncertainties/
>
> See, fun stuff. Still using optical methods
> after all these years.

Isn't X-ray lithography a bit better?

>
> https://en.wikipedia.org/wiki/Multiple_patterning
>
> IBM can use much more expensive (lab prototype)
> methods, like an E-beam writer, to make their test
> chip. But whatever method they used, might be
> too slow or too expensive for production purposes.
>
> A lot of teamwork goes into this stuff. Even though
> the companies compete, they can only keep the lithography
> machine company in business, if he makes the same
> equipment for everybody.

Grant you, it is a very big interdependent industry.

>
> Paul

[toc] | [prev] | [next] | [standalone]


#341341 — Re: AMD vs. Intel

Fromchrisv <chrisv@nospam.invalid>
Date2016-01-22 06:59 -0600
SubjectRe: AMD vs. Intel
Message-ID<od94abl87kah5depbdsfrh3oek7lgnmffd@4ax.com>
In reply to#341267
Melzzzzz wrote:

> slime wrote:
>>
>> However, we're ignoring another possibility: if x86-64 becomes too
>> expensive to buy into, PowerPC platforms might return to prominence
>> and computers such as the AmigaOne might actually be noticed by the
>> marketplace.

Another possibility is that you'll someday buy a fscking clue, and
also stop being a lying shit.

But I doubt it.

> ARM is big contender not PowerPC...

Leave it that worthless troll to spew such nonsense.

-- 
"Vinyl is technically better (than CD)"  -  "Slimer"

[toc] | [prev] | [next] | [standalone]


#341397 — Re: AMD vs. Intel

FromGreyCloud <mist@cumulus.com>
Date2016-01-22 15:30 -0700
SubjectRe: AMD vs. Intel
Message-ID<n7uaec$t4$2@dont-email.me>
In reply to#341341
On 01/22/16 05:59, chrisv wrote:
> Melzzzzz wrote:
>
>> slime wrote:
>>>
>>> However, we're ignoring another possibility: if x86-64 becomes too
>>> expensive to buy into, PowerPC platforms might return to prominence
>>> and computers such as the AmigaOne might actually be noticed by the
>>> marketplace.
>
> Another possibility is that you'll someday buy a fscking clue, and
> also stop being a lying shit.
>
> But I doubt it.
>
>> ARM is big contender not PowerPC...
>
> Leave it that worthless troll to spew such nonsense.
>
You have no clue, shit-fer-brains.

[toc] | [prev] | [next] | [standalone]


#341345 — Re: AMD vs. Intel

Fromchrisv <chrisv@nospam.invalid>
Date2016-01-22 07:25 -0600
SubjectRe: AMD vs. Intel
Message-ID<9ma4ab5deo27jho74kiemv0rs0ljgeobtv@4ax.com>
In reply to#341267
> "Slimer" wrote:
>>
>> However, we're ignoring another possibility: if x86-64 becomes too
>> expensive to buy into, PowerPC platforms might return to prominence
>> and computers such as the AmigaOne might actually be noticed by the
>> marketplace.

x86-64 has been quite affordable for years already.  Even if Intel
gains a monopoly and improvements price/performance slow, it will
never will never be "too expensive to buy into".

Also the desktop PC market is in decine.  At this point there is no
room for another competitor.

-- 
"Have you ever met an advocate who wasn't repulsive? Look at pictures
of the people in here. (Advocate's name redacted) looks like a serial
rapist and (advocate's name redacted) looks like he has a collection
of child pornography on his computer. No sane person would approach
these people."  -  "Slimer"

[toc] | [prev] | [next] | [standalone]


#341268 — Re: AMD vs. Intel

FromMelzzzzz <mel@zzzzz.com>
Date2016-01-21 21:23 +0100
SubjectRe: AMD vs. Intel
Message-ID<20160121212329.5340eac1@maxa-pc>
In reply to#341202
On 20 Jan 2016 23:21:49 GMT
vallor <vallor@cultnix.org> wrote:

> 
> Then you add in the fact that though AMD's Bulldozer architecture 
> allegedly had "8 real cores", there was the recent scandal where we 
> discovered that they have 4 FPU's per processor.  

This was no secret. Problem is that shops sell these CPU's as 8 core,
while they are really not. Then again question is : what is core?

> 
> Unless AMD comes up with something amazing, their goose may be 
> cooked...possibly roasted over one of their processors.

Zen is around the corner I hear.

> 
> (Not sure what that means for their GPU business, I use NVidia.)
> 

AMD has open sourced their GPU hardware which means much better drivers
for their cards. I might shift to AMD when such drivers appear, I hope
very soon. AMD could be preferred choice on Linux....

[toc] | [prev] | [next] | [standalone]


#341280 — Re: AMD vs. Intel

FromSlimer <.m@nsn.s>
Date2016-01-21 17:40 -0500
SubjectRe: AMD vs. Intel
Message-ID<n7rmk2$2un$1@dont-email.me>
In reply to#341268
On 2016-01-21 3:23 PM, Melzzzzz wrote:
> On 20 Jan 2016 23:21:49 GMT
> vallor <vallor@cultnix.org> wrote:
> 
>>
>> Then you add in the fact that though AMD's Bulldozer architecture 
>> allegedly had "8 real cores", there was the recent scandal where we 
>> discovered that they have 4 FPU's per processor.  
> 
> This was no secret. Problem is that shops sell these CPU's as 8 core,
> while they are really not. Then again question is : what is core?
> 
>>
>> Unless AMD comes up with something amazing, their goose may be 
>> cooked...possibly roasted over one of their processors.
> 
> Zen is around the corner I hear.
> 
>>
>> (Not sure what that means for their GPU business, I use NVidia.)
>>
> 
> AMD has open sourced their GPU hardware which means much better drivers
> for their cards. I might shift to AMD when such drivers appear, I hope
> very soon. AMD could be preferred choice on Linux....

Open-sourcing the drivers won't mean shit. They'll still underperform
and cause crashes because the people who will be working on "improving"
them are so incompetent that they might as well be cleaning Joseph "7"
Michael out of toilets alongside Wideload Porkster. Five years from now,
even afer this change, AMD's GNU/Linux drivers will still be awful and
games will STILL require NVIDIA first.


-- 
Slimer
EFF & OpenMedia member / IFAW, Mozilla & PETA supporter
"Everything seems to work fine, except that occasionally everything
freezes." - Another one of GNU/Linux's many victims
"Sound wasn't even a gauranteed thing on a PC in the 90s." - Linux
apologist JEDIDIAH explaining why sound never worked/works right in Linux

[toc] | [prev] | [next] | [standalone]


#341289 — Re: AMD vs. Intel

FromDFS <nospam@dfs.com>
Date2016-01-21 18:28 -0500
SubjectRe: AMD vs. Intel
Message-ID<n7rpd7$fqh$2@dont-email.me>
In reply to#341268
On 1/21/2016 3:23 PM, Melzzzzz wrote:


> AMD has open sourced their GPU hardware which means much better drivers
> for their cards. I might shift to AMD when such drivers appear, I hope
> very soon. AMD could be preferred choice on Linux....


Where's a link to AMD talking about open sourcing their GPUs?

I don't see a mention of it here:

AMD Open Source Zone
http://developer.amd.com/tools-and-sdks/open-source/


[toc] | [prev] | [next] | [standalone]


#341295 — Re: AMD vs. Intel

FromMelzzzzz <mel@zzzzz.com>
Date2016-01-22 00:59 +0100
SubjectRe: AMD vs. Intel
Message-ID<20160122005904.575cb3cb@maxa-pc>
In reply to#341289
On Thu, 21 Jan 2016 18:28:04 -0500
DFS <nospam@dfs.com> wrote:

> On 1/21/2016 3:23 PM, Melzzzzz wrote:
> 
> 
> > AMD has open sourced their GPU hardware which means much better
> > drivers for their cards. I might shift to AMD when such drivers
> > appear, I hope very soon. AMD could be preferred choice on
> > Linux....  
> 
> 
> Where's a link to AMD talking about open sourcing their GPUs?
> 
> I don't see a mention of it here:
> 
> AMD Open Source Zone
> http://developer.amd.com/tools-and-sdks/open-source/
> 
> 
> 

I posted link to article that described all in detail here in group
some time ago. 

[toc] | [prev] | [next] | [standalone]


#341447 — Re: AMD vs. Intel

FromDFS <nospam@dfs.com>
Date2016-01-23 11:20 -0500
SubjectRe: AMD vs. Intel
Message-ID<n8094e$vkh$5@dont-email.me>
In reply to#341295
On 1/21/2016 6:59 PM, Melzzzzz wrote:
> On Thu, 21 Jan 2016 18:28:04 -0500
> DFS <nospam@dfs.com> wrote:
>
>> On 1/21/2016 3:23 PM, Melzzzzz wrote:
>>
>>
>>> AMD has open sourced their GPU hardware which means much better
>>> drivers for their cards. I might shift to AMD when such drivers
>>> appear, I hope very soon. AMD could be preferred choice on
>>> Linux....
>>
>>
>> Where's a link to AMD talking about open sourcing their GPUs?
>>
>> I don't see a mention of it here:
>>
>> AMD Open Source Zone
>> http://developer.amd.com/tools-and-sdks/open-source/
>>
>>
>>
>
> I posted link to article that described all in detail here in group
> some time ago.


So again you post wild-ass claims with zero evidence.

You can do better, SMelzzzzz.  We already have idiots like 7 and Dumb 
Willie for that.

[toc] | [prev] | [next] | [standalone]


#341588 — Re: AMD vs. Intel

From"Ezekiel" <zeke@nosuchemail.com>
Date2016-01-24 19:18 -0500
SubjectRe: AMD vs. Intel
Message-ID<n83pfd$a3f$1@dont-email.me>
In reply to#341447
"DFS" <nospam@dfs.com> wrote in message news:n8094e$vkh$5@dont-email.me...
> On 1/21/2016 6:59 PM, Melzzzzz wrote:
>> On Thu, 21 Jan 2016 18:28:04 -0500
>> DFS <nospam@dfs.com> wrote:
>>
>>> On 1/21/2016 3:23 PM, Melzzzzz wrote:
>>>
>>>
>>>> AMD has open sourced their GPU hardware which means much better
>>>> drivers for their cards. I might shift to AMD when such drivers
>>>> appear, I hope very soon. AMD could be preferred choice on
>>>> Linux....
>>>
>>>
>>> Where's a link to AMD talking about open sourcing their GPUs?
>>>
>>> I don't see a mention of it here:
>>>
>>> AMD Open Source Zone
>>> http://developer.amd.com/tools-and-sdks/open-source/
>>>
>>>
>>>
>>
>> I posted link to article that described all in detail here in group
>> some time ago.
>
>
> So again you post wild-ass claims with zero evidence.
>
> You can do better, SMelzzzzz.  We already have idiots like 7 and Dumb 
> Willie for that.
>

I don't follow (or care much) about the GPU market. From what I've seen AMD 
is doing something called OpenGPU which has nothing to do with the GPU 
itself. They're open sourcing various parts of their video driver software 
stack.

-- 
"It's not so much the threat of divorce, as making your every moment at home 
a living hell. My wife used to dole out my weekly allowance, $20, and then 
quiz me about where I spent it, as she didn't expect me to buy snacks and 
such."

Chris Ahlstrom - emasculated wimp without a spine.
http://groups.google.com/group/comp.os.linux.advocacy/msg/ecaf21d667b4afca?hl=en


[toc] | [prev] | [next] | [standalone]


#341243

From"Ezekiel" <zeke@nosuchemail.com>
Date2016-01-21 10:40 -0500
Message-ID<n7qtva$tu3$1@dont-email.me>
In reply to#341199
"Jeremy Bentham" <nobody@anemone.mooo.com> wrote in message 
news:7cc3ac0cc41ea4325ef6dcc848bacd0a@anemone.mooo.com...
> In article <n7op4t$r15$1@dont-email.me>
> "Ezekiel" <zeke@nosuchemail.com> wrote:
>>
>>
>> "Slimer" <.m@nsn.s> wrote in message news:n7o5k3$9rb$1@dont-email.me...
>> > On 2016-01-19 9:30 PM, Nomen Nescio wrote:
>> >> In article <n7mc4p$ktl$10@dont-email.me>
>> >> GreyCloud <mist@cumulus.com> wrote:
>> >>>
>> >>> On 01/19/16 14:55, Anonymous Remailer (austria) wrote:
>> >>>> In article<n7kufj$v6o$1@dont-email.me>
>> >>>> "Mr. Man-wai Chang"<toylet.toylet@gmail.com>  wrote:
>> >>>>>
>> >>>>> On 1/19/2016 5:04 PM, JJ wrote:
>> >>>>>>
>> >>>>>>    From what I can see is that they've haven't tested Skylake 
>> >>>>>> well,
>> >>>>>> and the
>> >>>>>> product is kind of rushed to the market. They *assumed* that such
>> >>>>>> heavy load
>> >>>>>> will never happen. Well, strike two, Intel.
>> >>>>>> Regardsless, it's still a fact that there's a bug.
>> >>>>>
>> >>>>> This is NOT the first time Intel released a CPU with bug if my 
>> >>>>> memory
>> >>>>> serves....
>> >>>>
>> >>>> Every Intel CPU ever released has bugs.  Most are not critical
>> >>>> and get fixed during the manufacturing life if deemed important
>> >>>> enough.
>> >>>>
>> >>> Even tho cpu vendors have errata sheets, IBMs have a much shorter 
>> >>> errata
>> >>> sheets than Intel.
>> >>
>> >> Intel is like Cisco, bugs are "features"
>> >>
>> >> http://download.intel.com/design/processor/specupdt/320836.pdf
>> >>
>> >> 136 "No fix" in this one.
>> >>
>> >> AMD had a shitwad too.  Really disappointed in AMD.  When they
>> >> brought those losers from ATI into the fold, the end was written
>> >> for AMD.
>> >>
>> >> http://support.amd.com/TechDocs/25759.pdf
>> >
>> > Well, to be honest, the graphics department they inherited from ATI is
>> > the only thing keeping the company alive at the moment. Their 
>> > processors
>> > are being ignored (except those who are looking for a decent mobile 
>> > chip
>> > with decent graphics performance) but their GPUs provide excellent
>> > performance for a fraction of the price of an NVIDIA card. If I had a
>> > desktop, I'd be looking long and hard at procuring an AMD GPU for $500
>> > rather than an NVIDIA equivalent at about $800.
>> >
>> > Either way, AMD's A10 and A8 chips are great for laptops but it seems
>> > that nobody wants them.
>> >
>>
>> AMD is in trouble. They lost about $100 million dollars last quarter and 
>> the
>> market-cap for the entire company is only about $1.4 billion. That's 
>> roughly
>> the same market-cap as the company that makes Ugg shoes.
>>
>>     "Advanced Micro generated revenues of $0.96 billion, down 9.7%
>> sequentially and down 22.7% year over year"
>>
>> I have nothing against AMD and wish them well. Looking at what they have,
>> it's hard to see some silver-lining that's going to turn their fortunes
>> around.
>
> They need to make the smaller processors more powerful.  That is
> why they are not selling.  It is purely a performance issue.
>

That pretty much sums up the objective for any processor. Small and powerful 
along with being power efficient.

The problem as I see it for AMD is that they're in a no-mans land with their 
processors. Intel dominates the x86 market for PCs and servers and ARM 
processors own the mobile market.

So what's left for AMD? There's video (ATI) but I'm not sure how big this 
is. For most people who don't use their PCs for gaming the built-in graphics 
chipset is more than adequate.

It does leave AMD in a difficult position. One possibility is if VR (virtual 
reality) takes off at all then perhaps they can use their GPUs to drive some 
of that.

-- 
>Problem with Chromebooks in the enterprise - when IBM, HP, eBay,
>Citi Bank, etc writes an email, internal document or spreadsheet
>they don't want Google to see each and every word they write?

Didn't think you had the brains that God gave a cockroach, fsckwit.

How about encrypting things that they don't want others to see?

turdv "thinks" that encryption will magically provide privacy on the docs 
and spreadsheets you edit, calculate and spell-check online. Google can 
spell-check your document but if it's encrypted they won't be able to read 
what they're spell-checking.
Message-ID: <gh8rf81ca9gjjjrt1pghah4tc41t41068l@4ax.com>





[toc] | [prev] | [next] | [standalone]


#341251

FromPAS <abc@doremi.net>
Date2016-01-21 12:16 -0500
Message-ID<n7r3jm$lqv$2@dont-email.me>
In reply to#341243
On 1/21/2016 10:40 AM, Ezekiel wrote:
> "Jeremy Bentham" <nobody@anemone.mooo.com> wrote in message
> news:7cc3ac0cc41ea4325ef6dcc848bacd0a@anemone.mooo.com...
>> In article <n7op4t$r15$1@dont-email.me>
>> "Ezekiel" <zeke@nosuchemail.com> wrote:
>>>
>>> "Slimer" <.m@nsn.s> wrote in message news:n7o5k3$9rb$1@dont-email.me...
>>>> On 2016-01-19 9:30 PM, Nomen Nescio wrote:
>>>>> In article <n7mc4p$ktl$10@dont-email.me>
>>>>> GreyCloud <mist@cumulus.com> wrote:
>>>>>> On 01/19/16 14:55, Anonymous Remailer (austria) wrote:
>>>>>>> In article<n7kufj$v6o$1@dont-email.me>
>>>>>>> "Mr. Man-wai Chang"<toylet.toylet@gmail.com>  wrote:
>>>>>>>> On 1/19/2016 5:04 PM, JJ wrote:
>>>>>>>>>     From what I can see is that they've haven't tested Skylake
>>>>>>>>> well,
>>>>>>>>> and the
>>>>>>>>> product is kind of rushed to the market. They *assumed* that such
>>>>>>>>> heavy load
>>>>>>>>> will never happen. Well, strike two, Intel.
>>>>>>>>> Regardsless, it's still a fact that there's a bug.
>>>>>>>> This is NOT the first time Intel released a CPU with bug if my
>>>>>>>> memory
>>>>>>>> serves....
>>>>>>> Every Intel CPU ever released has bugs.  Most are not critical
>>>>>>> and get fixed during the manufacturing life if deemed important
>>>>>>> enough.
>>>>>>>
>>>>>> Even tho cpu vendors have errata sheets, IBMs have a much shorter
>>>>>> errata
>>>>>> sheets than Intel.
>>>>> Intel is like Cisco, bugs are "features"
>>>>>
>>>>> http://download.intel.com/design/processor/specupdt/320836.pdf
>>>>>
>>>>> 136 "No fix" in this one.
>>>>>
>>>>> AMD had a shitwad too.  Really disappointed in AMD.  When they
>>>>> brought those losers from ATI into the fold, the end was written
>>>>> for AMD.
>>>>>
>>>>> http://support.amd.com/TechDocs/25759.pdf
>>>> Well, to be honest, the graphics department they inherited from ATI is
>>>> the only thing keeping the company alive at the moment. Their
>>>> processors
>>>> are being ignored (except those who are looking for a decent mobile
>>>> chip
>>>> with decent graphics performance) but their GPUs provide excellent
>>>> performance for a fraction of the price of an NVIDIA card. If I had a
>>>> desktop, I'd be looking long and hard at procuring an AMD GPU for $500
>>>> rather than an NVIDIA equivalent at about $800.
>>>>
>>>> Either way, AMD's A10 and A8 chips are great for laptops but it seems
>>>> that nobody wants them.
>>>>
>>> AMD is in trouble. They lost about $100 million dollars last quarter and
>>> the
>>> market-cap for the entire company is only about $1.4 billion. That's
>>> roughly
>>> the same market-cap as the company that makes Ugg shoes.
>>>
>>>      "Advanced Micro generated revenues of $0.96 billion, down 9.7%
>>> sequentially and down 22.7% year over year"
>>>
>>> I have nothing against AMD and wish them well. Looking at what they have,
>>> it's hard to see some silver-lining that's going to turn their fortunes
>>> around.
>> They need to make the smaller processors more powerful.  That is
>> why they are not selling.  It is purely a performance issue.
>>
> That pretty much sums up the objective for any processor. Small and powerful
> along with being power efficient.
>
> The problem as I see it for AMD is that they're in a no-mans land with their
> processors. Intel dominates the x86 market for PCs and servers and ARM
> processors own the mobile market.
>
> So what's left for AMD? There's video (ATI) but I'm not sure how big this
> is. For most people who don't use their PCs for gaming the built-in graphics
> chipset is more than adequate.
>
> It does leave AMD in a difficult position. One possibility is if VR (virtual
> reality) takes off at all then perhaps they can use their GPUs to drive some
> of that.
>

Being the CPU supplier to Microsoft And Sony for the Xbox and 
Playstation helps AMD but I wonder to what extent.

[toc] | [prev] | [next] | [standalone]


Page 4 of 6 — ← Prev page 1 2 3 [4] 5 6  Next page →

Back to top | Article view | comp.os.linux.advocacy


csiph-web