Wednesday, October 1st 2025

Intel Reaffirms $28 Billion Ohio Expansion Despite Delay to 2030

Intel reiterated this week that its long-planned Ohio semiconductor campus remains a core part of its manufacturing strategy after Republican Senator Bernie Moreno demanded clearer answers about delays and potential costs to local taxpayers. In a brief response to Moreno, the company stated that it is coordinating with Ohio officials, the state's congressional delegation, and local stakeholders as it revises the schedule, and emphasized that the site "remains an important part of our long-term plans to expand leading-edge manufacturing on U.S. soil." Intel acknowledged that the timeline has shifted—commercial production is now not expected until at least 2030—and said it will maintain the flexibility to adjust construction and ramping based on customer demand. Moreno's letter highlighted roughly $2 billion in public incentives and nearly $700 million in infrastructure commitments, and he pressed the company for specifics on economic impact and any steps to limit costs to Ohio taxpayers. However, the company's public reply did not answer those questions in detail.

Leadership changes, rounds of workforce reductions and the challenge of attracting outside foundry customers have slowed momentum for Intel, while new private investments and a recent federal equity stake have altered financial and strategic calculations. The government involvement has added a national-security angle to the debate and increased scrutiny from state officials and residents who had expected faster job and supply-chain benefits. Intel insists the Ohio campus remains strategic and says it will keep working with stakeholders to align build-out with market realities. For now, the ambitious proposal survives as a long-term bet. Meanwhile, TrendForce reports citing Commercial Times that says Intel started limited 18A node shipments to U.S. customers in Q3, with 18A wafers already in production and initial output of its own CPUs expected in Q4. At least some parts of the manufacturing roadmap remain according to schedule, meaning that the $28 billion Ohio plant project is safe for now.
Source: Bloomberg
Add your own comment

16 Comments on Intel Reaffirms $28 Billion Ohio Expansion Despite Delay to 2030

#1
Philaphlous
YEa whoever the GC is for this build is screwed... I'm sure its a pain to build to say the least but contract terms should have had a definitive end date with any delay costing $X per day beyond the estimated completion date....
Posted on Reply
#2
TristanX
So, look like 14A fabs are now guaranteed, then why these speaks it can be scrapped ?
Posted on Reply
#3
kondamin
TristanXSo, look like 14A fabs are now guaranteed, then why these speaks it can be scrapped ?
could be skipping 14A for foundry services just doing a "limited" internal run like they are doing intel 3
or they could sell their NA-euv and stick to EUV
Posted on Reply
#4
remixedcat
And the longer it takes the more expensive it'll be and it'll be dated by the time they get it up n running.
Posted on Reply
#5
kondamin
remixedcatAnd the longer it takes the more expensive it'll be and it'll be dated by the time they get it up n running.
doesn't have to be, those buildings are as expensive as they are because of what they are built for and all the infrastructure needed near the site to keep them running.

while the EUV machines might no longer be cutting edge at the time, swapping them out for what is current at that point fixes that so the investment is still good.
Posted on Reply
#7
Hecate91
kondaminwhile the EUV machines might no longer be cutting edge at the time, swapping them out for what is current at that point fixes that so the investment is still good.
It takes time to replace the machines and get the machines running up to acceptable yields, by the time they replace machines and get the fab running they'll likely be behind again.


Intel must be giving their competitors some very attractive deals to use an Intel node, though it wouldn't be wise for AMD to hand their designs over to their main competitor.
Posted on Reply
#8
kondamin
Hecate91It takes time to replace the machines and get the machines running up to acceptable yields, by the time they replace machines and get the fab running they'll likely be behind again.


Intel must be giving their competitors some very attractive deals to use an Intel node, though it wouldn't be wise for AMD to hand their designs over to their main competitor.
Sure, but the buildings need to be ready before those machines get installed, so they will come last
Posted on Reply
#9
Dr. Dro
Hecate91It takes time to replace the machines and get the machines running up to acceptable yields, by the time they replace machines and get the fab running they'll likely be behind again.


Intel must be giving their competitors some very attractive deals to use an Intel node, though it wouldn't be wise for AMD to hand their designs over to their main competitor.
This Intel vs. AMD thing only exists in consumers' heads. Same for Apple vs. Samsung. Samsung manufactures a ton of hardware for Apple (and to this day, manufactures iPhone panels etc.), and even made Apple's chips for a long time (while their lithography made sense, anyway - for example, A9 came in both TSMC and Samsung variants). If the Intel node is competitive and/or makes logistical sense for AMD, they should use it to bolster volume and/or increase profit margins.
Posted on Reply
#10
Tomorrow
Dr. DroThis Intel vs. AMD thing only exists in consumers' heads.
Really? Beyond standards such as PCIe etc that both are part of there's almost no collaboration between the two.
The last time there was, was when Intel Made Kaby Lake-G NUC with AMD licensed Vega graphics.
Intel's marketing has continuously called out AMD with dubious and ridiculous claims such as the "snake oil" and "glued together"

I think it's safe to say there's no love lost between the two. Even more animosity between Nvidia and Apple for example.
Yes i agree that Samsung vs Apple is largely performative since Samsung is a major supplier of various components to Apple.
And despite their mocking of Apple they end up copying the same things.
Posted on Reply
#11
igormp
TomorrowReally? Beyond standards such as PCIe etc that both are part of there's almost no collaboration between the two.
Well, they don't collaborate that much in end products, but there still is collaboration between companies. Latest example would be the x86 Advisory Group:
newsroom.intel.com/corporate/october-2024-intel-news

But there's also collaboration/negotiations in terms of CPU extensions and whatnot. Apart from that, they have founded and worked a lot in many different consortiums/standards, such as OCP, TCG, AI Alliance, UEFI Forum, etc etc.
TomorrowIntel's marketing has continuously called out AMD with dubious and ridiculous claims such as the "snake oil" and "glued together"
So did AMD, it's no different from the Samsung vs Apple example. They will bad mouth each other to try to improve their own sales, but will gladly shake hands if there are profits to be made. It's just how business goes.
Posted on Reply
#12
Dr. Dro
TomorrowReally? Beyond standards such as PCIe etc that both are part of there's almost no collaboration between the two.
Yes, really. Money takes no sides, it's a lesson to life. For money, even the most embittered rivals will unite if it means that they're both making a killing here :toast:
TomorrowIntel's marketing has continuously called out AMD with dubious and ridiculous claims such as the "snake oil" and "glued together"
Haha, it's just business. AMD got back at them for their digs, too, hanging these at airports:









Posted on Reply
#13
Tomorrow
igormpWell, they don't collaborate that much in end products, but there still is collaboration between companies. Latest example would be the x86 Advisory Group:
newsroom.intel.com/corporate/october-2024-intel-news

But there's also collaboration/negotiations in terms of CPU extensions and whatnot. Apart from that, they have founded and worked a lot in many different consortiums/standards, such as OCP, TCG, AI Alliance, UEFI Forum, etc etc.
That is what i meant by giving the PCI-SIG example.
Dr. DroYes, really. Money takes no sides, it's a lesson to life. For money, even the most embittered rivals will unite if it means that they're both making a killing here :toast:
I can only deduce that there has been zero money to be made then because of zero collaboration on end user products.
Or that there has been case of money to be made, but animosity between the two does exist after all?

I mean look at Intel's leadership. It wasn't just marketing. Their (now former) leader said at one point that "AMD is in the rear view mirror"
For a long time Intel acted as if AMD did not even exist and while that might have been logical prior to 2016, then after that they should have stood up and payed attention.

There's always potential for collaboration. Even between AMD and Nvidia, but we rarely see it.
If as you say money talks then they should be constantly collaborating on products by combining the strong sides of both, but there's nothing.
Sure, Nvidia did/does use EPYC in their racks, but that's because it was so much better than Xeon.
Posted on Reply
#14
TechBuyingHavoc
I can see Intel slow-walking this expansion since most of the early costs are less expensive one-time charges (planning, buying land, permitting, constructing the long-term assets like the physical building) while the costs yet to be incurred are the majority of the money (clean room construction, fab equipment purchases and installation, validation efforts, etc). If there is a concern about cash flow and viability of future products, Intel may want to slow down investment here.
Posted on Reply
#15
igormp
TomorrowThat is what i meant by giving the PCI-SIG example.
That's a lot of consortiums that both are often leads on. I'd say that's a ton of collaboration.
You seem to just be focused on specific end-products, such as an Intel CPU with an AMD iGPU or something of the sorts. Collaboration is not restricted to that.
Posted on Reply
#16
Dr. Dro
TomorrowI can only deduce that there has been zero money to be made then because of zero collaboration on end user products.
Yeah, because past Intel nodes are unsuitable, and it takes some 5 years from planning to release, so it's going to take another few years to see if IFS is even viable IMO
Posted on Reply
Add your own comment
Nov 20th, 2026 18:08 PST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts