By: Paul A. Clayton (paaronclayton.delete@this.gmail.com), February 27, 2019 2:32 pm
Room: Moderated Discussions
Doug S (foo.delete@this.bar.bar) on February 27, 2019 11:45 am wrote:
> Paul A. Clayton (paaronclayton.delete@this.gmail.com) on February 27, 2019 10:25 am wrote:
> > If (when?) anyone can trivially assemble ARM design components, ARM will be hurting the profit
> > margins of those licensing its designs. When a MediaTek chip that just assembles ARM design components
> > is nearly as good as a Qualcomm product, it would become difficult to justify extra design effort
> > from Qualcomm and higher prices (to fund such effort and increase profit).
>
>
> What extra design effort? Qualcomm already gave up doing custom cores several years ago, now
> they use standard ARM cores. Maybe they are tweaking them a bit, but if they are it isn't enough
> to show up in benchmarks. So Qualcomm has no room to complain about ARM's licensing making them
> feel the pain of competition, because they are simply licensing cores themselves.
>
> The only Android OEM designing custom cores is Samsung,
I was using "MediaTek" as a name for "company that mainly just assembles prepared components and sends the design to a fab" and "Qualcomm" as a name for "company that designs custom cores and other components". Samsung seemed a less good example because its association with its own fabs.
> Qualcomm's value add is in the integration of everything you need, including cellular and other
> stuff like fingerprint readers. That's where they will have to do better than the rest to keep
> their competitive edge, but when modems from Samsung, Huawei, Mediatek and Intel become "good
> enough", and the other stuff is "good enough" then there isn't much room for them to get a bigger
> profit margin off their SoCs. That won't be ARM's fault though, anymore than ARM is to blame
> for the smartphone market being saturated and replacement cycles lengthening.
If ARM was providing the designs for the modems and "other stuff", then the shrinking profit margin would rightly be attributed to ARM even when the products are produced by other companies. The better and more complete ARM design components and integration become (for a given target) and the easier and cheaper it is to use them, the harder it becomes for others to justify higher prices.
An arms dealer selling mercenary services as well as effective weapons and training would not be that much more an economic threat to a mercenary company than one that selling just weapons and training when anyone can get good recruits. For ARM's licensees, fab technology is not a distinguishing factor ("anyone can get good recruits"). If ARM's design offerings become "good enough" (and complete enough), ARM effectively becomes an indirect competitor.
(Arms dealers and dog food — my mind is strange.)
My post was meant to point out that ARM selling hardware would not be substantially increasing competition compared to providing "turnkey" solutions. Licensees might rightly see their money going to subsidize the competition, but that is true whether ARM sells hardware or not (and it already true to a less extent just in cheaply licensing RTL for cores). It is quite possible that selling hardware would have a greater emotional effect, so selling hardware might push existing licensees to look elsewhere.
> Paul A. Clayton (paaronclayton.delete@this.gmail.com) on February 27, 2019 10:25 am wrote:
> > If (when?) anyone can trivially assemble ARM design components, ARM will be hurting the profit
> > margins of those licensing its designs. When a MediaTek chip that just assembles ARM design components
> > is nearly as good as a Qualcomm product, it would become difficult to justify extra design effort
> > from Qualcomm and higher prices (to fund such effort and increase profit).
>
>
> What extra design effort? Qualcomm already gave up doing custom cores several years ago, now
> they use standard ARM cores. Maybe they are tweaking them a bit, but if they are it isn't enough
> to show up in benchmarks. So Qualcomm has no room to complain about ARM's licensing making them
> feel the pain of competition, because they are simply licensing cores themselves.
>
> The only Android OEM designing custom cores is Samsung,
I was using "MediaTek" as a name for "company that mainly just assembles prepared components and sends the design to a fab" and "Qualcomm" as a name for "company that designs custom cores and other components". Samsung seemed a less good example because its association with its own fabs.
> Qualcomm's value add is in the integration of everything you need, including cellular and other
> stuff like fingerprint readers. That's where they will have to do better than the rest to keep
> their competitive edge, but when modems from Samsung, Huawei, Mediatek and Intel become "good
> enough", and the other stuff is "good enough" then there isn't much room for them to get a bigger
> profit margin off their SoCs. That won't be ARM's fault though, anymore than ARM is to blame
> for the smartphone market being saturated and replacement cycles lengthening.
If ARM was providing the designs for the modems and "other stuff", then the shrinking profit margin would rightly be attributed to ARM even when the products are produced by other companies. The better and more complete ARM design components and integration become (for a given target) and the easier and cheaper it is to use them, the harder it becomes for others to justify higher prices.
An arms dealer selling mercenary services as well as effective weapons and training would not be that much more an economic threat to a mercenary company than one that selling just weapons and training when anyone can get good recruits. For ARM's licensees, fab technology is not a distinguishing factor ("anyone can get good recruits"). If ARM's design offerings become "good enough" (and complete enough), ARM effectively becomes an indirect competitor.
(Arms dealers and dog food — my mind is strange.)
My post was meant to point out that ARM selling hardware would not be substantially increasing competition compared to providing "turnkey" solutions. Licensees might rightly see their money going to subsidize the competition, but that is true whether ARM sells hardware or not (and it already true to a less extent just in cheaply licensing RTL for cores). It is quite possible that selling hardware would have a greater emotional effect, so selling hardware might push existing licensees to look elsewhere.
Topic | Posted By | Date |
---|---|---|
ARM announces Ares | nobody in particular | 2019/02/20 08:35 AM |
ARM announces Ares | Adrian | 2019/02/20 08:39 AM |
ARM announces Ares | dmcq | 2019/02/20 10:03 AM |
ARM announces Ares | Michael S | 2019/02/20 10:41 AM |
ARM announces Ares | dmcq | 2019/02/20 12:49 PM |
ARM announces Ares | Michael S | 2019/02/20 01:21 PM |
ARM announces Ares | dmcq | 2019/02/20 02:01 PM |
ARM announces Ares | Wilco | 2019/02/20 02:31 PM |
ARM announces Ares | dmcq | 2019/02/20 03:16 PM |
ARM announces Ares | Wilco | 2019/02/20 03:49 PM |
ARM announces Ares | dmcq | 2019/02/20 04:09 PM |
ARM announces Ares | Wilco | 2019/02/20 04:45 PM |
ARM announces Ares | nobody in particular | 2019/02/20 04:55 PM |
ARM announces Ares | Dan Fay | 2019/02/20 05:44 PM |
ARM announces Ares | Wilco | 2019/02/20 07:06 PM |
ARM announces Ares | Dan Fay | 2019/02/21 08:27 AM |
ARM announces Ares | dmcq | 2019/02/20 05:49 PM |
ARM announces Ares | Wilco | 2019/02/20 06:40 PM |
ARM announces Ares | Charles | 2019/02/21 02:16 AM |
ARM announces Ares | Michael S | 2019/02/21 02:26 AM |
ARM announces Ares | anon | 2019/02/20 08:55 PM |
ARM announces Ares | JS | 2019/02/21 12:59 AM |
*has not hasn't (NT) | JS | 2019/02/21 01:01 AM |
ARM announces Ares | Salvatore De Dominicis | 2019/02/21 07:05 AM |
Definitions of RISC | Brendan | 2019/02/21 10:07 AM |
Definitions of RISC | Michael S | 2019/02/21 10:16 AM |
PDP-8 Not Usually Considered RISC | Mark Roulo | 2019/02/21 02:10 PM |
PDP-8 Not Usually Considered RISC | rwessel | 2019/02/21 07:13 PM |
Definitions of RISC | Adrian | 2019/02/21 02:42 PM |
Definitions of RISC (nod to John Mashey and comp.arch) | wumpus | 2019/02/21 06:29 PM |
Definitions of RISC (nod to John Mashey and comp.arch) | none | 2019/02/22 12:32 AM |
Definitions of RISC (nod to John Mashey and comp.arch) | Michael S | 2019/02/22 04:28 AM |
Definitions of RISC (nod to John Mashey and comp.arch) | none | 2019/02/22 08:01 AM |
ARM announces Ares | lockederboss | 2019/02/20 09:56 AM |
stability? (NT) | anonymous2 | 2019/02/20 10:01 AM |
ARM announces Ares | dmcq | 2019/02/20 10:05 AM |
ARM announces Ares | Groo | 2019/02/20 10:11 AM |
ARM announces Ares | Andrei Frumusanu | 2019/02/20 11:49 AM |
ARM announces Ares | Linus Torvalds | 2019/02/20 10:36 AM |
ARM announces Ares | Michael S | 2019/02/20 10:54 AM |
ARM announces Ares | Geoff Langdale | 2019/02/20 03:07 PM |
ARM announces Ares | dmcq | 2019/02/20 03:32 PM |
ARM announces Ares | none | 2019/02/21 12:03 AM |
That last line should have been removed :-) (NT) | none | 2019/02/21 12:04 AM |
ARM announces Ares | Michael S | 2019/02/21 01:47 AM |
ARM announces Ares | none | 2019/02/21 03:59 AM |
ARM announces Ares | Michael S | 2019/02/21 04:45 AM |
ARM announces Ares | dmcq | 2019/02/21 05:18 AM |
ARM announces Ares | Simon Farnsworth | 2019/02/22 09:43 AM |
ARM announces Ares | anon | 2019/02/20 09:27 PM |
ARM announces Ares | Michael S | 2019/02/21 01:53 AM |
ARM announces Ares | Linus Torvalds | 2019/02/21 09:03 AM |
ARM announces Ares | Michael S | 2019/02/21 09:35 AM |
ARM announces Ares | Michael S | 2019/02/21 09:51 AM |
ARM announces Ares | Foo_ | 2019/02/21 02:40 PM |
ARM announces Ares | aaron spink | 2019/02/21 03:56 PM |
ARM announces Ares | Linus Torvalds | 2019/02/21 04:27 PM |
ARM announces Ares | Stoffels | 2019/02/22 12:21 AM |
ARM announces Ares | Michael S | 2019/02/22 04:15 AM |
ARM announces Ares | Simon Farnsworth | 2019/02/22 09:41 AM |
ARM announces Ares | none | 2019/02/22 10:30 AM |
In other words: nobody will ever get fired for choosing x86 (NT) | Gabriele Svelto | 2019/02/24 01:32 PM |
In other words: nobody will ever get fired for choosing x86 | Simon Farnsworth | 2019/02/25 04:53 AM |
ARM announces Ares | Foo_ | 2019/02/22 02:52 AM |
ARM announces Ares | Gabriele Svelto | 2019/02/24 01:31 PM |
ARM announces Ares | Maynard Handley | 2019/02/25 03:57 AM |
ARM announces Ares | Gabriele Svelto | 2019/02/25 04:21 AM |
ARM announces Ares | Michael S | 2019/02/25 04:58 AM |
ARM announces Ares | nobody in particular | 2019/02/25 05:21 AM |
ARM announces Ares | Adrian | 2019/02/26 08:02 AM |
ARM announces Ares | Maynard Handley | 2019/02/26 12:32 AM |
ARM announces Ares | Gabriele Svelto | 2019/02/26 12:14 PM |
ARM announces Ares | David Hess | 2019/03/19 05:34 PM |
ARM announces Ares | none | 2019/02/26 01:34 AM |
ARM announces Ares | Gabriele Svelto | 2019/02/26 12:16 PM |
ARM announces Ares | none | 2019/02/27 12:19 AM |
ARM announces Ares | end of an era | 2019/02/24 03:18 PM |
Word salad bot strikes again (NT) | nanon | 2019/02/25 12:26 AM |
ARM announces Ares | hobel | 2019/02/25 02:10 AM |
ARM announces Ares | Michael S | 2019/02/25 02:52 AM |
ARM announces Ares | hobel | 2019/02/25 10:48 AM |
ARM announces Ares | Andrew Clough | 2019/02/25 02:07 PM |
ARM announces Ares | Maynard Handley | 2019/02/26 12:38 AM |
ARM announces Ares | John Yates | 2019/02/26 03:43 PM |
ARM announces Ares | Howard Chu | 2019/02/27 06:55 AM |
ARM announces Ares | David Hess | 2019/03/19 05:22 PM |
ARM announces Ares | David Hess | 2019/03/19 04:54 PM |
ARMY announces Ares | dmcq | 2019/03/20 03:12 AM |
ARMY announces Ares | David Hess | 2019/03/20 05:47 AM |
ARMY announces Ares | dmcq | 2019/03/20 06:05 AM |
ARM announces Ares | Groo | 2019/02/21 10:23 AM |
ARM announces Ares | Michael S | 2019/02/21 10:29 AM |
ARM announces Ares | David Hess | 2019/03/19 05:44 PM |
ARM announces Ares | juanrga | 2019/02/21 11:52 AM |
ARM announces Ares | anon | 2019/02/21 08:19 PM |
ARM announces Ares | hobel | 2019/02/22 03:40 AM |
software ecosystems | RichardC | 2019/02/21 04:31 PM |
software ecosystems | Foo_ | 2019/02/22 03:15 AM |
ARM announces Ares | dmcq | 2019/02/21 05:31 PM |
+ on hyperscaling | dmcq | 2019/02/22 08:23 AM |
+ on hyperscaling | Maynard Handley | 2019/02/22 08:38 AM |
+ on hyperscaling | juanrga | 2019/02/22 11:57 AM |
+ on hyperscaling | dmcq | 2019/02/22 08:23 AM |
sorry duplicate | dmcq | 2019/02/22 08:26 AM |
ARM announces Ares | anon | 2019/02/21 08:34 PM |
ARM announces Ares | Brett | 2019/02/21 10:20 PM |
ARM announces Ares | A. Wilcox | 2019/02/22 03:52 PM |
ARM announces Ares | Doug S | 2019/02/20 03:30 PM |
ARM announces Ares | dmcq | 2019/02/20 04:14 PM |
ARM announces Ares | Doug S | 2019/02/21 08:40 AM |
ARM announces Ares | aaron spink | 2019/02/21 04:05 PM |
ARM announces Ares | Maynard Handley | 2019/02/22 08:48 AM |
ARM announces Ares | juanrga | 2019/02/21 02:28 AM |
ARM announces Ares | Michael S | 2019/02/21 03:25 AM |
ARM announces Ares | juanrga | 2019/02/22 04:09 AM |
ARM announces Ares | Maynard Handley | 2019/02/22 08:51 AM |
ARM announces Ares | David Soul | 2019/02/22 11:47 AM |
ARM announces Ares | juanrga | 2019/02/22 12:23 PM |
Chicken or egg, both. | Brett | 2019/02/22 02:21 PM |
Chicken or egg, both. | David Soul | 2019/02/22 06:05 PM |
Chicken or egg, both. | Brett | 2019/02/22 09:55 PM |
ARM sales | juanrga | 2019/02/23 03:55 AM |
ARM sales | aaron spink | 2019/02/23 08:47 AM |
Chicken or egg, both. | Ronald Maas | 2019/02/23 06:33 PM |
Chicken or egg, both. | Magagop | 2019/02/24 11:18 PM |
why not? (NT) | Michael S | 2019/02/25 02:03 AM |
why not? | Not the parent | 2019/02/25 10:36 AM |
why not? | Michael S | 2019/02/25 10:43 AM |
why not? | dmcq | 2019/02/25 11:35 AM |
why not? | Michael S | 2019/02/25 12:03 PM |
why not? | dmcq | 2019/02/25 12:29 PM |
why not? | Not the parent | 2019/02/26 10:08 AM |
why not? | Doug S | 2019/02/26 12:05 PM |
why not? | Not the parent | 2019/02/27 04:51 AM |
why not? | Doug S | 2019/02/27 12:15 PM |
why not? | Not the parent | 2019/02/28 01:43 AM |
why not? | anon | 2019/02/26 11:22 PM |
why not? | Not the parent | 2019/02/27 04:59 AM |
why not? | anon | 2019/02/27 06:42 AM |
why not? | Magagop | 2019/02/27 03:31 PM |
why not? | anon | 2019/02/27 05:02 PM |
why not? | dmcq | 2019/02/27 05:25 PM |
why not? | anon | 2019/02/27 05:56 PM |
why not? | dmcq | 2019/02/27 06:19 PM |
why not? | anon | 2019/02/27 06:46 PM |
why not? | Wilco | 2019/02/28 05:49 AM |
why not? | Jukka Larja | 2019/02/28 07:02 AM |
why not? | Michael S | 2019/02/28 08:32 AM |
why not? | Wilco | 2019/02/28 09:40 AM |
why not? | Doug S | 2019/02/28 10:13 AM |
why not? | Foo_ | 2019/03/01 07:35 AM |
why not? | Doug S | 2019/03/01 11:05 AM |
why not? | anon | 2019/03/01 02:32 PM |
why not? | dmcq | 2019/03/01 03:13 PM |
why not? | anon | 2019/03/01 04:13 PM |
why not? | dmcq | 2019/03/02 03:56 AM |
why not? | anon | 2019/03/02 10:09 AM |
why not? | dmcq | 2019/03/02 10:41 AM |
why not? | anon | 2019/03/02 12:19 PM |
why not? | dmcq | 2019/03/03 03:48 AM |
why not? | Doug S | 2019/03/02 04:53 PM |
why not? | Michael S | 2019/03/02 05:06 PM |
why not? | Doug S | 2019/03/03 12:54 AM |
why not? | Michael S | 2019/03/03 02:58 AM |
why not? | anon | 2019/03/03 05:35 AM |
why not? | dmcq | 2019/03/03 06:33 AM |
I considired dry humor a Brittish specialty. It seems, I was wrong about it. (NT) | Michael S | 2019/03/03 06:39 AM |
why not? | anon | 2019/03/03 10:13 AM |
why not? | Michael S | 2019/03/03 12:41 PM |
why not? | dmcq | 2019/03/03 01:21 PM |
why not? | anon | 2019/03/03 04:15 PM |
why not? | anon | 2019/02/28 08:38 PM |
why not? | Not the parent | 2019/02/28 01:48 AM |
strawman (NT) | anon | 2019/02/28 08:31 PM |
why not? | Michael S | 2019/02/27 03:53 AM |
why not? | Not the parent | 2019/02/27 05:26 AM |
why not? | Michael S | 2019/02/27 05:42 AM |
why not? | dmcq | 2019/02/27 05:49 AM |
why not? | Michael S | 2019/02/27 06:21 AM |
why not? | dmcq | 2019/02/27 07:15 AM |
Selling at multiple layers | Paul A. Clayton | 2019/02/27 11:25 AM |
Selling at multiple layers | Doug S | 2019/02/27 12:45 PM |
Selling at multiple layers | Paul A. Clayton | 2019/02/27 02:32 PM |
Selling at multiple layers | Doug S | 2019/02/27 04:04 PM |
Paul A. Clayton | 2019/02/27 07:06 PM | |
Arm reference board? | Simon Farnsworth | 2019/02/28 12:34 PM |
why not? | Doug S | 2019/02/27 12:46 PM |
why not? (NT) | la mama de murgay | 2019/02/25 10:22 PM |
Chicken or egg, both. | Ronald Maas | 2019/02/25 01:23 PM |
ARM announces Ares | Gabriele Svelto | 2019/02/24 03:07 PM |
ARM announces Ares | juanrga | 2019/03/02 05:40 AM |
ARM announces Ares | Gabriele Svelto | 2019/03/02 08:08 AM |
ARM announces Ares | dmcq | 2019/03/02 10:10 AM |
ARM announces Ares | hobel | 2019/02/22 04:33 PM |
ARM announces Ares | David Soul | 2019/02/22 06:20 PM |
Troll | Doug S | 2019/02/23 01:37 AM |
Troll | David Soul | 2019/02/23 08:59 AM |
ARM announces Ares | Ennis | 2019/02/23 02:18 AM |
ARM announces Ares | Foo_ | 2019/02/23 02:38 AM |
ARM announces Ares | anon | 2019/02/23 03:46 AM |
ARM announces Ares | dmcq | 2019/02/23 05:41 AM |
ARM announces Ares | nobody in particular | 2019/02/26 03:58 AM |
expensive | anonymous2 | 2019/02/27 12:46 AM |
ARM announces Ares | Howard Chu | 2019/02/27 07:19 AM |
ARM announces Ares | Howard Chu | 2019/02/27 07:21 AM |
ARM announces Ares | dmcq | 2019/02/27 07:57 AM |
ARM announces Ares | Jukka Larja | 2019/02/27 08:11 AM |
ARM announces Ares | Michael S | 2019/02/27 08:16 AM |
ARM announces Ares | Doug S | 2019/02/27 12:49 PM |
ARM announces Ares | Michael S | 2019/02/27 01:43 PM |
ARM announces Ares | Doug S | 2019/02/27 04:06 PM |
ARM announces Ares | Gian-Carlo Pascutto | 2019/02/28 05:00 AM |
ARM announces Ares | Gian-Carlo Pascutto | 2019/02/28 05:02 AM |
ARM announces Ares | Michael S | 2019/03/01 03:47 AM |
death by a thousand cuts | Michael S | 2019/03/01 07:27 AM |
death by a thousand cuts | Foo_ | 2019/03/01 07:37 AM |
death by a thousand cuts | dmcq | 2019/03/01 12:10 PM |
ARM announces Ares | Jukka Larja | 2019/02/28 06:31 AM |
ARM announces Ares | Howard Chu | 2019/02/28 09:17 AM |
ARM announces Ares | Gabriele Svelto | 2019/02/28 03:02 PM |
ARM announces Ares | aaron spink | 2019/03/01 01:19 AM |
ARM announces Ares | Andrew Clough | 2019/02/28 11:32 AM |
ARM announces Ares | Andrew Clough | 2019/02/28 11:33 AM |
ARM announces Ares | Foo_ | 2019/03/01 01:55 AM |
ARM announces Ares | Jukka Larja | 2019/03/01 05:40 AM |
ARM announces Ares | Howard Chu | 2019/02/27 08:45 AM |
ARM announces Ares | dmcq | 2019/02/27 09:17 AM |
ARM announces Ares | dmcq | 2019/02/27 09:21 AM |
ARM announces Ares | Howard Chu | 2019/02/27 04:03 PM |
ARM announces Ares | Jukka Larja | 2019/02/28 06:03 AM |
ARM announces Ares | none | 2019/02/27 11:49 PM |
ARM announces Ares | Howard Chu | 2019/02/28 04:26 AM |
ARM announces Ares | Peter E. Fry | 2019/02/28 07:31 PM |
ARM announces Ares | Adrian | 2019/02/28 10:09 PM |
ARM announces Ares | Gabriele Svelto | 2019/03/02 12:56 AM |
ARM announces Ares | Howard Chu | 2019/03/02 10:05 AM |
ARM announces Ares | Gabriele Svelto | 2019/03/02 02:31 PM |