I've been betting this would happen for years, and was always wrong. Until now. It's a low-margin business that Intel originally created as a "standard high volume server" enabling initiative. With companies like Supermicro, Inspur, and Lenovo as players (not to mention Dell and HPE), they stuck it out longer than Intel usually does in low-margin businesses.
Intel sold off its server business
- Thread starter blueone
- Start date
Rack servers are low margin because the barrier to entry in that market is not high, as compared to say semiconductor design and manufacturing. Intel has competitive modern efficiency. But Intel needs to use their R&D funding on businesses that provide better gross margins, so they're focusing on what does. Intel and their customers don't need Intel to be in the server chassis and board business the way they did in 2003, so it got pruned. IMO, that business should have been sold off a while ago.
I also wonder about their PC NUC business...
This is a parallel business unit. They are not removing staff from their more efficient business, correct?
They make CPUs. They have their own foundry sophisticated foundry. They have interposers to package them. They should make DRAM (or make a deal with Micron). It seems to me that they should have the best servers.
Not having your own foundry and sophisticated packaging seems to me a potential barrier to entry. If Intel comes close to breaking even on the server business, it is still worth keeping. Geopolitics is changing rapidly right now. i would stand... Pat.
They make CPUs. They have their own foundry sophisticated foundry. They have interposers to package them. They should make DRAM (or make a deal with Micron). It seems to me that they should have the best servers.
Not having your own foundry and sophisticated packaging seems to me a potential barrier to entry. If Intel comes close to breaking even on the server business, it is still worth keeping. Geopolitics is changing rapidly right now. i would stand... Pat.
It is about focus and Intel definitely needs it. I have been amazed by some of the moon shot acquisitions Intel has made over the years only to shut them down.
Acquisitions are a completely different story.
I assume their server business is close to breaking even. It uses their chips. It proves in their flow. They will continue to improve the SIPs and reduce the board and components within. They should change the culture and improve.
I assume their server business is close to breaking even. It uses their chips. It proves in their flow. They will continue to improve the SIPs and reduce the board and components within. They should change the culture and improve.
I think they're profitable, as is the NUC business, but IMO to Intel this sort of assembly, contract design, and logistics business is just a distraction they don't need.
Yup, like two Ethernet switch acquisitions, both of which have been shutdown. (Fulcrum and Barefoot Networks) In the early 2000s it was billions in communications companies. Remember the nonsensical WiMax moonshot to attempt to displace cellular? If it isn't an x86 CPU, manufacturing chips, or creating enabling software, Intel senior management just can't seem to get a winning attitude about it. I think the jury is still out on FPGAs and GPUs, but I wish I was more hopeful. I think we agree that IFS needs to urgently succeed to save the company as we know it.