[WSJ] Intel to Stop Making Modem Chips for 5G Smartphones

[ehume]

Intel is exiting the business of making modem chips for 5G smartphones, the company said, hours after its chief rival in that market, Qualcomm, resolved a yearslong legal dispute with Apple.

 

[CHADBOGA]

Intel has so many failures outside their core competency, it is not funny.

 

[positivedoppler]

Soooo, Charlie at sa was spot on calling out Intel a looong time ago

 

[NostaSeronx]

I wonder if they will focus the rest of the development towards Lantiq stuff. It would be nice to get 802.11be production drafts sooner than later.
https://www.intel.com/content/www/us/en/smart-home/connected-home/anywan-grx750-home-gateway.html
^-- be nice to get a 10nm quad-core as well.

I don't see why they should drop 5G Modems entirely, there is still the always connected laptops/tablets.
Random search quote; "The chip maker is teaming up with Microsoft, Dell, HP and Lenovo to create new PCs with 5G connectivity using Intel’s XMM 8060 5G multimode modem. The company expects the first 5G-connected PCs to be launched in the second half of 2019, which should give some time to carriers around the world to build their 5G networks."

 

[ehume]

. . . I don't see why they should drop 5G Modems entirely, there is still the always connected laptops/tablets.

They're not. They're just not going to produce stuff for smartphones. IOT is different, somehow.

Edit to add cite: "Hours after the Apple-Qualcomm settlement was announced, surprise, surprise, Intel declared it will quit building 5G modems for smartphones. It will also "complete an assessment of the opportunities for 4G and 5G modems in PCs, internet of things devices and other data-centric devices."

 

[scannall]

I have to wonder if this interview applies. It seems to.

Xerox Failure

 

[Markfw]

I have to wonder if this interview applies. It seems to.

Xerox Failure

I am deaf, and there is no cc on that video. Could you sum it up for me please ? (not joking)

Since I was let go from Xerox as they cleaned house at Tektronix, I am really curious, and I hope Xerox the worst.

 

[scannall]

I am deaf, and there is no cc on that video. Could you sum it up for me please ? (not joking)

Since I was let go from Xerox as they cleaned house at Tektronix, I am really curious, and I hope Xerox the worst.

https://lybio.net/steve-jobs-on-why-xerox-failed/people/amp/

 

[gorobei]

I am deaf, and there is no cc on that video. Could you sum it up for me please ? (not joking)

Since I was let go from Xerox as they cleaned house at Tektronix, I am really curious, and I hope Xerox the worst.

its just the Steve Jobs interview where he outlines how companies with monopolies(like pepsi and xerox) end up promoting marketing personnel to the decision making tiers while shutting out the technical product makers. if there is no benefit for making a better copier then the only one who can increase profits is the advertising guys. end result is management has no clue whether a product is good or bad. the parc xerox team called the executives 'toner heads'. Jobs says xerox could have owned the pc market if not for the bad.

 

[Markfw]

https://lybio.net/steve-jobs-on-why-xerox-failed/people/amp/

Well, thanks, but that really didn;t tell me much.

Here is what I know about Xerox. Tektronix had a very successful printer line based on solid ink. Xerox could not compete, so they bought out the division of Tektronix that I worked in, and laid off most of the people, then killed the solid ink printer line. Stupid move IMO, they were great printers. (circa 2002)

If you can't beat the competition, buy them, and kill their product.

So thats why I hate Xerox. And technically I retired from there with 20 years service. But no severance package.

Also, I agree with the statement that upper management does not know anything about the product. The company that I worked for After Tektronix/Xerox was the same. Upper management are idiots, and I could prove it, including a VP that I knew very well, but at the top, stupidity reins supreme ! They got millions for loosing money and bad customer service.

 

[ehume]

Think about what happened to Syracuse China, or to Lechmere. Same deal: buy the competition, kill off the product line, lay off the people. Look at what happened to Maxtor as opposed to Hitachi in HD. Teddy Roosevelt started trust-busting. Antitrust enforcement was supposed to save capitalism, but today we have generic med makers making monopoly profits by not making meds that compete, and their alleged competitors don't make meds that would compete, so the consumers pay higher prices. And now we see Intel . . . not competing. Clearly, their marketers run the company.

My late father used to say, long ago, "Engineers like work and salesmen like money. So they give engineers work and salesmen money." Things haven't changed.

 

[dark zero]

They should partnered with Mediatek and Spreadtrum in order to make their 5G modems competent... seems that the US-China Trade War was even more critical than expecting.

And Intel is clearly suffering the IBM destiny... but this time by their own hand and because they did not innovated in time.

 

[NTMBK]

Wow, the diversification from the PC market is going great.

 

[jpiniero]

Have to assume it's pretty much another product killed off because they can't really fix 10 nm. There was no way they would have been able to produce enough chips for Apple.

I would have thought they would have seen this coming and tried to move it to a functional node, but guess not or was too late.

 

[Dayman1225]

Have to assume it's pretty much another product killed off because they can't really fix 10 nm. There was no way they would have been able to produce enough chips for Apple.

I would have thought they would have seen this coming and tried to move it to a functional node, but guess not or was too late.

I don’t think you can blame every failure of Intel on 10nm... especially given rumours that XMM 8160 was to be fabbed at TSMC. I think this decision is primarily due to new management, it had one customer that was ultimately going to abandon them in the future... why continue investing into this money pit?

 

[NTMBK]

Semiaccurate has a fairly entertaining story up on this: https://semiaccurate.com/2019/04/16/qualcomm-just-beat-apple-into-sumbission/

 

[DrMrLordX]

I don’t think you can blame every failure of Intel on 10nm... especially given rumours that XMM 8160 was to be fabbed at TSMC. I think this decision is primarily due to new management, it had one customer that was ultimately going to abandon them in the future... why continue investing into this money pit?

If Charley is to be believed, it's because neither the 8060 nor 8160 ever worked. Would be interesting to know if the 8060 samples were fabbed on Intel 10nm or TSMC 7nm.

Semiaccurate has a fairly entertaining story up on this: https://semiaccurate.com/2019/04/16/qualcomm-just-beat-apple-into-sumbission/

Heehee. Pity Qualcomm killed their server CPU business. That would have gotten interesting with all that extra money flowing into the coffers from Apple.

Also, what about that Apple modem team? 6 years of Qualcomm modems? Hmmmm

 

[Dayman1225]

Also, what about that Apple modem team? 6 years of Qualcomm modems? Hmmmm

6 year licensing agreement and a "multi year chipset agreement", not 6 years of chipset agreement

 

[jpiniero]

I don’t think you can blame every failure of Intel on 10nm... especially given rumours that XMM 8160 was to be fabbed at TSMC. I think this decision is primarily due to new management, it had one customer that was ultimately going to abandon them in the future... why continue investing into this money pit?

There were rumors of fabbing at TSMC; but either way it was either on 10 nm, or it was being ported from 10 nm to TSMC and they couldn't complete it in time for Apple.

 

[piokos]

There were rumors of fabbing at TSMC; but either way it was either on 10 nm, or it was being ported from 10 nm to TSMC and they couldn't complete it in time for Apple.

Why would Intel want to use an expensive 10nm process on modems is beyond me. It's a low margin, low power part. It could be done on 14nm just like the LTE modems they make for Apple now.

Truth is this was never a good deal for Intel, but it could be excellent for Qualcomm.
Both companies gained on stock exchanges today. Qualcomm by enormous 35% and Intel by excellent 5% (given that they "lost" a contract, not got one ).

 

[coercitiv]

Why would Intel want to use an expensive 10nm process on modems is beyond me. It's a low margin, low power part. It could be done on 14nm just like the LTE modems they make for Apple now.

Power. Power. Power.

In smartphones everything is low power.

 

[jpiniero]

Why would Intel want to use an expensive 10nm process on modems is beyond me. It's a low margin, low power part. It could be done on 14nm just like the LTE modems they make for Apple now.

The first gen 5g modem was on Intel's 14 nm and sucked up way too much power.

 

[deathBOB]

6 year licensing agreement and a "multi year chipset agreement", not 6 years of chipset agreement

Good catch. I doubt anyone can build a modem without licensing Qualcomm IP.

 

[Mopetar]

I wonder how many years that "multi-year" chipset agreement is, because that should tell you when Apple plans to have their own baseband. They bought up plenty of patents of their own and you know that they have no problems with paying for good engineering talent.

 

[DrMrLordX]

6 year licensing agreement and a "multi year chipset agreement", not 6 years of chipset agreement

So the plot thickens. I agree with @Mopetar in wanting to know the length of the chipset agreement.