Couldn't find the path - Intel kills two more businesses

[BachChain]

Summary

 

Continuing CEO Pat Gelsinger's war on unprofitability, Intel has announced the end of two more divisions, bringing the total to seven since Pat took the reigns. The two now on the chopping block are NEX, their network switch brand, and Pathfinder, their RISC-V IDE program. These cuts come following their Q4 earnings call where Intel reported a loss of over six-hundred and fifty million dollars.

 

Quotes

Quote

NEX continues to do well and is a core part of our strategic transformation, but we will end future investment on our network switching product line, while still fully supporting existing products and customers. Since my return, we have exited seven businesses, providing in excess of $1.5 billion in savings.

Quote

We regret to inform you that Intel is discontinuing the Intel® Pathfinder for RISC-V program effective immediately.

Since Intel will not be providing any additional releases or bug fixes, we encourage you to promptly transition to third-party RISC-V* software tools that best meet your development needs.

 

My thoughts

Intel made switches? How many more businesses does Intel have left to burn through?

 

Sources

https://download.intel.com/newsroom/2023/corporate/q422-investor-call-remarks.pdf

https://pathfinder.intel.com/

 

[Needfuldoer]

9 minutes ago, BachChain said:

Intel made switches?

I think they were mostly in the high end / exotic cluster infrastructure category, competing against Cisco and Mellanox.

 

The regular Ethernet edge switch market is pretty well saturated with established players, from Cisco, HPE/Aruba, Broadcom, and Extreme Networks all the way down to Linksys and Netgear. They didn't have anything competing against the likes of TP-Link.

 

I wonder who's going to buy that tech off them. Ubiquiti? HPE? The ghost of Proteon?

[HenrySalayne]

8 minutes ago, BachChain said:

and Pathfinder, their RISC-V IDE program.

I'm wondering what this means for Intel's investment in RISC-V in general. Is this a goodbye to RISC-V or just the IDE?

[Blqckqut]

2 minutes ago, HenrySalayne said:

I'm wondering what this means for Intel's investment in RISC-V in general. Is this a goodbye to RISC-V or just the IDE?

whats RISC-V

[porina]

1 minute ago, Blqckout said:

whats RISC-V

Open CPU architecture, so alternative to x86 and Arm. New kid on the block relatively speaking but it is probably the next big thing, eventually.

[Blqckqut]

1 minute ago, porina said:

Open CPU architecture, so alternative to x86 and Arm. New kid on the block relatively speaking but it is probably the next big thing, eventually.

isnt intel mostly known for cpus? why would he shut it down, is he secretly an AMD enthousiast and uses AMD cpus but wants to continue working on INTEL GPUS so hell get better performance lol

[williamcll]

That explains why they had a collaboration with an SiFive to make a RISC-V board with DDR5/PCIe5 ports. It's just that they no longer have their own development team anymore.

[igormp]

36 minutes ago, HenrySalayne said:

I'm wondering what this means for Intel's investment in RISC-V in general. Is this a goodbye to RISC-V or just the IDE?

It was mostly an IDE for 3rd party, they seem to still be investing on RISC-V and even released a new CPU in partnership with SiFive:

https://www.tomshardware.com/news/sifive-intel-hifive-risc-v-development-board

30 minutes ago, Blqckout said:

isnt intel mostly known for cpus? why would he shut it down, is he secretly an AMD enthousiast and uses AMD cpus but wants to continue working on INTEL GPUS so hell get better performance lol

Intel does lots of stuff, but x86 is exclusive to only a specific segment and has a weird license where no one manufactures can effectively make use of it. Since ARM is growing even stronger, but also has some fucked up licensing, vendors are looking into moving to RISC-V since it's royalty free, and Intel wants their share of the cake, they're investing in RISC-V CPUs too.

[Coaxialgamer]

2 hours ago, Blqckout said:

isnt intel mostly known for cpus? why would he shut it down, is he secretly an AMD enthousiast and uses AMD cpus but wants to continue working on INTEL GPUS so hell get better performance lol

Intel's RISC V efforts ≠ their entire CPU program. RISC-V is an open-source architecture that came out a few years ago. It's mostly seen as an eventual competitor to ARM, so Intel investing in the RISC V space now means they would have a leg up if it ever overturned ARM.

 

Intel produces a lot of processors, including the x86 chips that form the backbone of client and server processing, but they also make microcontrollers (their Quark line) and apparently they design RISC-V stuff too. Intel retreating from this space just means they won't be pursuing developments in that particular aspect of RISC V, but certainly not that they're abandoning CPUs as a whole.

 

CEO must think that the RISC V gamble is one that won't or can't pay off for Intel, at least when it comes to IDEs. He might also believe that it's one of the less critical elements of Intel Corp and thus more deserving of being shut down in order to cut losses.

[Supreme Calamitas]

Gelsinger ain’t a bad ceo, I expected him to be since I never heard of him prior

[leadeater]

On 1/28/2023 at 4:50 AM, Needfuldoer said:

Ubiquiti? HPE? The ghost of Proteon?

Ubiquity not a chance, I highly doubt they want to enter that market segment. It'll be someone like Broadcom or Cisco, one already in this sector but want to take over existing Intel based customers.

[AlTech]

On 1/27/2023 at 3:40 PM, BachChain said:

How many more businesses does Intel have left to burn through?

A lot but not enough. Though one of them is their CPU division if it doesn't stop losing money.

 

Raptor Lake has basically zero profit below the i9 products. Intel has way too many leftover Alderlake dies and so needs to sell them for unprofitable amounts + rebadging the low end 13th gen just to get rid of stock.

 

Intel can't afford to keep fighting AMD the way they're doing it for too much longer given how much money Intel is haemorraging. 

 

Imho Intel's gotta move to chiplets for consumers and just take the latency penalty that comes with it. Their monolithic dies are clearly too expensive for them to keep making if Raptor Lake below i9 is selling at cost.

 

On 1/27/2023 at 3:40 PM, BachChain said:

Sources

https://download.intel.com/newsroom/2023/corporate/q422-investor-call-remarks.pdf

https://pathfinder.intel.com/

 

 

[porina]

35 minutes ago, AluminiumTech said:

Though one of them is their CPU division if it doesn't stop losing money.

CCG still has positive operating income, just a lot lower than it has been in past. Only AXG and IFS were at operating loss, but those are relatively new and longer term plays. 

 

It looks like the overall operating loss reported for the most recent quarter is mostly a result from investing in the future. They are still on their recovery path from 14nm.

 

35 minutes ago, AluminiumTech said:

Imho Intel's gotta move to chiplets for consumers and just take the latency penalty that comes with it.

This has been their plan on record for years. Next gen Meteor Lake is moving to tiles on higher end, with monolithic believed to remain on lower end. Not much difference to how AMD has chiplet offerings and monolithic APUs depending on placement.

 

35 minutes ago, AluminiumTech said:

Their monolithic dies are clearly too expensive for them to keep making if Raptor Lake below i9 is selling at cost.

Where did you see that? Couldn't find it in a quick search.

[Lurick]

On 1/27/2023 at 10:50 AM, Needfuldoer said:

I think they were mostly in the high end / exotic cluster infrastructure category, competing against Cisco and Mellanox.

 

The regular Ethernet edge switch market is pretty well saturated with established players, from Cisco, HPE/Aruba, Broadcom, and Extreme Networks all the way down to Linksys and Netgear. They didn't have anything competing against the likes of TP-Link.

 

I wonder who's going to buy that tech off them. Ubiquiti? HPE? The ghost of Proteon?

Intel made the ASICs (after they bought Tofino) and switches, they were a true ODM for that stuff. There is no way Ubiquiti would or could buy that up because they would need fab capacity and the ability to utilize the ASICs. We're not talking lower end 10 gigabit stuff, we're talking 100gig, 400 gigabit and up. As @leadeater said it would be someone akin to Cisco, Juniper, or Broadcom most likely. I know back in the day Cisco was supposedly gunning for Tofino Networks but Intel beat them to the punch but we'll see who scoops this up now.

 

Edit:

What's crazy is Intel bought Tofino Networks just 3.5 years ago and they're already killing the division.

[AlTech]

3 hours ago, porina said:

Where did you see that? Couldn't find it in a quick search.

MLID.