News Intel Sunsets Network Switch Biz, Kills RISC-V Pathfinder Program

[Admin]

Intel Sunsets Network Switch Business, Kills RISC-V Pathfinder Program : Read more

 

[InvalidError]

If you put all of your eggs in the same "high return" basket, you are setting yourself up for a hard crash-and-burn should anything happen to that basket such as the current downturn in demand for all things x86 and GPUs.

 

[setx]

Cutting RISC-V looks silly: it might not be high-margin area for Intel CPUs/chipsets, but it could become high-volume external demand for their foundry.

 

[InvalidError]

Cutting RISC-V looks silly: it might not be high-margin area for Intel CPUs/chipsets, but it could become high-volume external demand for their foundry.

Intel can fab others' RISC-V designs without developing its own RISC-V stuff. TSMC has little to no design input in its customers' design besides providing them with the primitives libraries to plug into their EDA tools so they can spit out physical designs from HDL.

 

[DuelMaster]

Intel is sacrificing the long term benefits to please the shareholders in the short term. Eliminating all of these businesses will eliminate what makes Intel "special". There was nothing like Optane memory in the industry. The elimination of their storage business with the Optane business eliminates any chance for them to create a unique product in the space. They could have made serious gains in the networking switch space because of their unique position as a processor business. In regards to the processor business, their decision with the RISC-V pathfinder business is strange to me. RISC-V is increasingly looking to be the future of computing. I am also seriously concerned for Arc graphics with those losses that they posted with the graphics team as mentioned in another article. Intel, in its efforts to reduce losses, may end Arc graphics and any chance to eventually make it in that space with it. I don't know. I'm not an analyst or one of the people on leadership or finance at Intel or anything like that, but I think they need to start telling investors that their businesses are in it for the long term, not the short term. The continued shortsightedness could be the slow death of Intel.

 

[kjfatl]

Is is possible that potential customers look at Intel's RISC-V as a direct competitor to their business model? If so, dropping the Intel IISC-V business might get 3 or 4 RISC-V IP suppliers to target Intel's fabs instead of just TSMC and Samsung.

 

[bit_user]

Cutting RISC-V looks silly:

They're not. They're just cutting an IP-neutral SDK that they put together for it.

I'd never heard of Pathfinder. Here's how Intel described it:

​

"Designed for SOC architects and system software developers, Intel Pathfinder for RISC-V is a pre-silicon development environment that supports IP selection via testing for compatibility and performance, as well as early-stage software development using Intel FPGA and simulator platforms."​

​

That ties in nicely with what the article said about it.

their decision with the RISC-V pathfinder business is strange to me. RISC-V is increasingly looking to be the future of computing.

Regarding RISC-V, what Intel cancelled was a small potatoes development tool. This Barcelona Supercomputing Center project sounds very interesting, and potentially much more relevant to their long-term strategic interests in this area.

BTW, the RISC-V Pathfinder program was aimed mostly at helping people get started on developing for non-Intel RISC-V cores. Were Intel to launch its own RISC-V IP, Pathfinder would run somewhat contrary to its interests. As such, it could signify Intel leaning into the space, rather than backing away from it. I don't put much stock in that take, but it might be something.

 

[bit_user]

Intel is sacrificing the long term benefits to please the shareholders in the short term.

I think Gelsinger is looking at their strategic interests and cutting anything that's not. For now, their discrete graphics business appears to be safe.