KEil bought by ARM

Keil compilers have been bought by ARm.

see

formatting link

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
 Click to see the full signature
Reply to
Chris Hills
Loading thread data ...

One is left wondering why they have done this.

And yes, I did read the press release.

Why would ARM want compliers for a load of processors that compete with their own? Especially as ARM seem to be doing their best to alienate their own compiler customers by forcing non-compatable upgrades on them at regular intervals.

As Kiel have, IIUC, recently releasae an ARM version of their complier methinks they are just taking out some competition.

tim

Reply to
tim (moved to sweden)

"Why would ARM want compliers for a load of processors that compete with their own? '

Probably they will sell off the non ARM compliers (unless Keil owns the lion share of that market, then I suppose they will just let them die)

Reply to
joep

"Why would ARM want compliers for a load of processors that compete with their own? '

Probably they will sell off the non ARM compliers (unless Keil owns the lion share of that market, then I suppose they will just let them die)

Reply to
StephensDigital

A good move for Arm.. in my opinion.. Keil knows how to build very good compilers.. and they have done so for many years.. if they do the same for ARM as they did with

8051 then it is indeed a good aquisition.. only time will tell..
Reply to
TheDoc

In article , joep writes

Why comment if you don't know?

AFAIK Keil have about 80% of the 8051 market and 50% of the 166 market.

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
 Click to see the full signature
Reply to
Chris Hills

In article , tim (moved to sweden) writes

Keil have 8051/251 166 and ARM 7 al in C They have they majority market share in 51 and 166

KEil's own compiler is C for ARM7. Others have C and C++ for 7,9,11 XScale etc. Mind you their ARM7 C compiler was very good.

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
 Click to see the full signature
Reply to
Chris Hills

In article , snipped-for-privacy@gmail.com writes

I can't see ARM killing off the cash cow that is the Keil 8051 market.

The Keil compiler is very good and covers the vast majority of 8051 parts, more than anyone else. That will be the case even if they freeze it now (which I doubt).

The 8051 market is not likely to expand with radically new types any more. New types yes but not extensions to the architecture I think.

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
 Click to see the full signature
Reply to
Chris Hills

In article , TheDoc writes

I think it is a good move for ARM (and for Keil for many reasons). The technologies of the two companies fit well. IT fits their strategies and is good for the marketplace.

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
 Click to see the full signature
Reply to
Chris Hills

The big 'win' in the silicon industry is going to be the processor that offers the best migration path from 8/16 bit to 32 bit.

In terms of developed systems, the 8051 is probably (I have no idea other than experience) the most popular embedded processor on the planet.

If ARM have the customer base for a large majority of those customers, then they are now in the best position to promote conversion to ARM.

I wonder who is going to buy IAR. :)

Reply to
Sonic

Actually Open Source is no match for good commercial compilers in so many ways.

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
 Click to see the full signature
Reply to
Chris Hills

Ah, the wonders of closed source coupled with the vagaries of commercialism. No match at all for Open Source.

Ian

Reply to
Ian Bell

Good for ARM - certainly. Imagine ARM trying to pitch the Cortex, when someone elses ARM7 compiler/library creates smaller/faster code !? Customer: "So what _was_ the benefit of this new core, again ?

- we already have a (cheaper?) ARM7 license "

With control of what was looking like the better embedded ARM compiler, NOW ARM can spin things to make Cortex look a LOT better.

Good for Keil's shareholders ? - yes.

Good for Keil's 166 user base ? sorry, I think not.

Good the Keil's 80C51 user base - hmm; most likely, the good programmers will be co-opted to ARM's Cortex.

As happens in these take-over cases, the resources are cherry-picked.

Also, you can be sure the user database can expect many "Special offer, Upgrade to ARM !!" mailers....

-jg

Reply to
Jim Granville

Mmm, you know I was rebuilding an old version of arm-elf-gcc just a couple of days ago, and I thought "If only this was a proprietary compiler, I'd be able to spend two weeks and thousands of dollars finding the right dongle version on eBay - and an old computer to work with it - so that I could run this old compiler again for a legacy project that only needs a two-line code change".

BTW, I was speaking to a Philips recruiter yesterday, and it looks as if an old project you and I talked about could be revisited ;)

Reply to
larwe

LOL

Ian

Reply to
Ian Bell

In article , larwe writes

Though it can take weeks to finding the old versions of Open source compilers and especially getting all the right parts.

With commercial compilers you can usually get the original vendor to provide the right version overnight. I have myself provided specific old versions of compilers (amended to work with the current dongle) for customers in 48 hours and at no cost as they were on support.

SO if you need to make a 2 line change and you have archived the program it is no problem otherwise it could take a LOT longer to get EXACTLY the right parts to recreate the open source compiler, especially if it was some years ago, than pick up from the compiler vendor an exact version of commercial SW.

Your illustration falls over when you compare like with like. If you have the source (or executables) for the old SW then the problems are the same. Unless you are suggesting that the old compiler SW is that old that it will not run on modern OS?

In which case how exact will the translation be using the old Open source compiler re-built for the later OS? I assume you compiled it with the same compiler used originally? if you rebuilt it with a newer compiler then you have a different compiler and again the illustration falls over. You may as well use the new version of the commercial compiler. Assuming you are still using it (or have a newer version that runs on your current hardware.

--
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
\/\/\/\/\ Chris Hills  Staffs  England     /\/\/\/\/
 Click to see the full signature
Reply to
Chris Hills

Also, even if you accept that the commercial compilers are "better" at code generation (I have not seen this):

- gcc is getting better and better. Presumably there is a limit to how good a commercial compiler can be, so any gap should be closing.

- code space and MHz are getting cheaper and cheaper, so it makes less and less difference anyway as time goes on.

- It is getting easier and easier to justify using 32 bit parts (like ARM) in new designs. I would expect the architecture of these parts to more closely match gcc's abilities than, say, the 8051 where Keil have been the leader.

--

John Devereux
Reply to
John Devereux

You don't *need* to be able to *build* the compiler, you can still just archive the binaries, in the same way that you would for a commercial compiler. Having the source is an *additional* benefit, in that there is then the possibility of running the compiler on different platforms where the binary might not have run.

IAR did not do this for me. We have a product that is now not maintained for this reason. We use gcc for everything, now.

My experience is that gcc has had less bugs than the commercial compilers I have used. My IAR8051 could not even take the address of a function!

At any rate 99.99% of the bugs in my software are due to me, not the compiler :(

--

John Devereux
Reply to
John Devereux

I tried to do this recently. I was buying a compiler from Altium/Tasking to modify/update code on a project a client has licensed. I wanted to at least start with being able to produce a binary image that matched the current version so I asked for a quote to include a previous version ( I thought at least covering media costs was reasonable).

The response I got back was that they couldn't provide any versions earlier than x (x being a minor differnce from their shipping version). So this assertion strikes me as overly optimistic.

Don't get me started on dongles and Flex

Robert

Reply to
R Adsett

And which compiler would that be?!? :-)

Good for Keil you mean - now they have access to the best ARM compiler in the world. Bundle it with a tiny library suitable for the low-end embedded space and you have an unbeatable solution. I'm sure some competitors in the MCU space are not very happy with the acquisition...

If you think the Keil compiler is better than the ARM compiler you're dreaming. Amongst many things the Keil compiler doesn't support Cortex. ARM's compiler got mentioned in the press statement, but not Keil's, so which compiler do you think Keil will be using in the future?

In the press statement it is made clear that support for non-ARM architectures will be continued. Most of Keil's profit is from 8- and

16-bit micro controllers - not ARM, so it would be stupid to kill that revenue. Of course I'm sure the goal would be to migrate users to ARM, but that is happening already.

Wilco

Reply to
Wilco Dijkstra

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.