On 11/12/18 12:31 PM, Liam Proven wrote:

On 12/11/2018 12:24, Juan Erbes wrote:

...

Tumbleweed is only for PowerPC and the buggy and insecure Intel microprocessors?

The page says

Intel & AMD (64-bit and 32-bit)

... right there... (?)

since a few minutes - check history ;) Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org

¿They are no difference betwen AMD and Intel?

openSUSE uses a subset of instructions that are supported by both vendors. Some parts may use CPU-specific instructions, but only after checking for support and only with alternative code path(s) for other CPUs. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org

El lun., 12 nov. 2018 a las 9:33, Michal Suchánek () escribió:

On Mon, 12 Nov 2018 08:41:42 -0300 Juan Erbes wrote:

...

El lun., 12 nov. 2018 a las 8:31, Liam Proven () escribió:

...

On 12/11/2018 12:24, Juan Erbes wrote:

...

Tumbleweed is only for PowerPC and the buggy and insecure Intel microprocessors?

The page says

Intel & AMD (64-bit and 32-bit)

... right there... (?)

Yes, I see today says this.

Revision as of 11:25, 12 November 2018

But not yesterday!

¿They are no difference betwen AMD and Intel?

There are no more differences between Intel and AMD than between different generations of supported Intel CPUs. Of course, the brand is different but from a technical point of view different CPU generations are no less different than CPUs from different vendors implementing compatible instruction set.

HTH

Cool, thanks! -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org

On 12/11/2018 13:19, Aleksa Sarai wrote:

(I think you'd need to do gcc -march=native or use very new Intel-specific intrinsics).

Beware -- much of that sort of thing and you'll start wanting to switch to Gentoo. ;-) https://funroll-loops.teurasporsaat.org/

And please remember that AMD's CPU business wasn't doing too well until little over a year ago -- and recently did they become very competitive with (and in some cases superior to) Intel. Mentioning AMD CPU support in 2016 wouldn't have made much sense or been impressive. (We aren't well-known for updating the wiki in a timely manner. :P)

That's not really fair. It's not just Intel and AMD and almost never was. Since the IBM PC came out, x86 chips have come from: * Intel * NEC * AMD * Harris * Cyrix * IDT * Via * Transmeta ... and that's from memory. Around the Pentium 1 period, the best bang-for-the-buck was Cyrix. Even IBM manufactured and used them. The Cyrix 6x86 was about 15-20% faster, clock for clock, than Intel. Quake killed it. https://liam-on-linux.livejournal.com/49259.html Then around the Pentium 4 period, the best performance for both money and for electricity consumption was AMD: https://www.pcper.com/news/Editorial/Yes-Netburst-really-was-bad-CPU-archite... Intel had to do a humiliating climb-down, cancel the entire P4 line and switch to a derivative of the Pentium-M instead: https://www.theregister.co.uk/2005/05/13/intel_confirms_netburst_end/ The Pentium-M being a derivative of the old Pentium 3, i.e. of the Pentium Pro from 1995. https://en.wikipedia.org/wiki/P6_(microarchitecture) So basically for at 3 periods during the time while x86 has been dominant -- NEC V20/V30 versus 8088/8086, 1982-1984; Cyrix 6x86 versus P5, 1996-1997; and AMD Sledgehammer versus Netburst P4, 2003 to 2006 -- Intel has *not* been the vendor of the fastest x86 processors. Yes, it's Intel's architecture. Yes, Intel has always been the dominant vendor. But it's not had it all its own way. -- Liam Proven - Technical Writer, SUSE Linux s.r.o. Corso II, Křižíkova 148/34, 186-00 Praha 8 - Karlín, Czechia Email: lproven@suse.com - Office telephone: +420 284 241 084 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org

On 12/11/2018 22:59, Juan Erbes wrote:

You forgot The Pentium was based on patents stolen to Digital from the Alpha risc processor http://archive.fortune.com/magazines/fortune/fortune_archive/1997/06/09/2275...

News to me. It looks like DEC tried to argue this and failed.

Yes, dominant with monopolic and mafia tactics, but no "because is the best"!

That is paranoid and I submit absurd. Intel *invented* x86. Did anyone force HP to share PA-RISC? Did anyone force Mostek to share the 6502? If anyone copies ARM, then ARM Ltd stops them. x86 is Intel's invention. Some customers insisted that the chip be second-sourced -- in other words, that other vendors were allowed to make it to remove dependence on a single manufacturer. So Intel licensed it to other vendors. Some enhanced it. But accusing Intel of being unfair, monopolistic etc. in controlling _its own invention_ is ridiculous.

Other time when AMD launched the X86-64 Opterons, and dominated the server market: http://www.novell.com/news/press/archive/2003/suse_archive/cray.html

That was one of my points, yes.

In November 12, 2009 Intel and AMD settle, agree cross-licensing deal, on which Intel has payed AMD $1.25bn and the two companies will share patent rights for the next five years, while AMD has cancelled all antitrust litigation against Intel https://www.zdnet.com/article/intel-and-amd-settle-agree-cross-licensing-dea...

Yes, so?

https://www.sec.gov/Archives/edgar/data/2488/000119312509236705/dex102.htm

A year ago according to a report by Fundzilla, Intel has reached a deal to license AMD’s graphics technology, though neither of the two companies have acknowledged or refuted the news. During the last year, there has been a lot of speculation that Intel was contemplating switching over to AMD graphics when its cross-licensing deal with Nvidia ends in March 2017 https://www.forbes.com/sites/greatspeculations/2017/05/17/licensing-deal-wit...

*Speculation*. No proof. Intel is going its own way with GPUs and they're good enough, so far. Both AMD and nVidia mainly focus on the lucrative gamer market. Intel has picked up that business users -- a huge sector -- and people who want better battery life from portables don't _need_ high-power GPUs. My little Mac mini drove a pair of 24" screens off its Core i5's built-in GPU and it was fine, even for the little gaming I do.

When I say "the intel microprocessors are buggy and insecure", I refer to this: https://arstechnica.com/information-technology/2018/01/intel-ceos-sale-of-st...

Not only Intel is vulnerable to this new class of speculative-execution exploits. -- Liam Proven - Technical Writer, SUSE Linux s.r.o. Corso II, Křižíkova 148/34, 186-00 Praha 8 - Karlín, Czechia Email: lproven@suse.com - Office telephone: +420 284 241 084 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org

On 11/13/18 12:51 PM, John Paul Adrian Glaubitz wrote:

On 11/13/18 12:29 PM, Liam Proven wrote:

...

Did anyone force HP to share PA-RISC? Did anyone force Mostek to share the 6502? If anyone copies ARM, then ARM Ltd stops them.

If I remember correctly, the 6502 was actually licensed by a lot of companies like NEC, Sharp and so on which is why it was so widely adopted.

Please stop discussing history here. This is completely irrelevant. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org

On 11/13/18 3:34 PM, Liam Proven wrote:

...

To which vendors? There are only two vendors on this planets which have an x86 license, those are AMD and VIA, the latter having sublicensed to some Chinese companies.

Today. Via owns both Cyrix and Centaur/IDT.

Yes, that's what I was talking about.

Harris made x86 chips up to the 80286: https://en.wikipedia.org/wiki/Intel_80286#History_and_performance

NEC's clones were unofficial: https://en.wikipedia.org/wiki/NEC_V20

This article: https://en.wikipedia.org/wiki/List_of_x86_manufacturers

Those are all historic. I'm talking about the status quo. Intel is not giving out any new licenses for x86, ARM does on their IP.

Told me about Zhaoxin, of whom I'd not previously heard. https://en.wikipedia.org/wiki/Zhaoxin

AFAIK those are also licensed through VIA.

...

Intel has actually a very tight grip on their IP unlike other vendors of CPU designs like ARM, Sun and IBM.

AFAIK nobody makes ARM chips without a licence...?

That's not what I said. The difference between ARM and Intel is that ARM is actually giving out licenses while Intel is not.

...

Actually, there was an antitrust lawsuit from the EU against Intel for exactly that reason. I don't know what the current status of that lawsuit is though.

Intel won it.

https://www.ft.com/content/f460ef98-930f-11e7-bdfa-eda243196c2c

https://www.reuters.com/article/us-eu-intel-antitrust/landmark-intel-judgmen...

That's not what the articles state. They say, that the case has been sent back to another court, quote: "The ECJ sent the case back to the general court to reconsider the commission’s decision and its analysis of whether the rebates harmed competition." That's not the same as winning a case. It just means the case needs another review. This is usually what higher courts do. Review a case and decide whether the decision is ok or needs to be revised. They don't actually speak a verdict. Intel might still get convicted in the end.

And it wasn't about control of x86, it was illegal sales practices, such as giving customers rebates if they dealt exclusively with Intel.

Well, it was an abuse they could only lever because of their tight grip on x86. Adrian -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org

On 2018-11-12, Juan Erbes wrote:

When I say "the intel microprocessors are buggy and insecure", I refer to this:

I think everyone on this list is aware of the series of speculative execution attacks seen in the past year-or-so. It should be noted that AMD also does speculative execution, and thus was also somewhat vulnerable to SPECTRE and similar attacks (though the speculation wouldn't cross privilege boundaries so it was much less severe). I would argue that "buggy and insecure" is a label that you could apply to most chips made in the past 15 years (if "vulnerable to speculative execution side-channels" is the bar for "buggy and insecure"). While I appreciate the reason why you have a problem with Intel (I get it, for a while it felt they were approaching monopoly-like control over mainstream CPUs), I don't think that presenting it on an openSUSE mailing list is the best place for such discussion. The original topic of the thread (that openSUSE supports AMD chips just as well as Intel ones) has already been resolved. Thank you. -- Aleksa Sarai Senior Software Engineer (Containers) SUSE Linux GmbH https://www.cyphar.com/