Intel's Response: Micron’s Control of 3D XPoint Fab Will Not Disrupt Optane Roadmap
by Anton Shilov on October 22, 2018 12:00 PM EST
Intel late on Friday said that Micron’s intention to gain full control of the IMFT joint venture will not interrupt its plans concerning products based on 3D XPoint memory. The company is confident that it will have enough time to start production of 3D Xpoint memory after Micron exercises its option to buy Intel’s stake in IMFT.
Micron currently controls 51% of IMFT, its joint venture with Intel, and has a right to buyout Intel’s share under certain conditions. After Micron proposes to buy Intel’s stake, as Micron has signified it intends to do next year, the deal will take up to twelve months to close and during that time the fab that exclusively makes 3D XPoint will be working as usual, and will make the memory for Intel Optane-branded products at near manufacturing cost. Furthermore, even after the deal closes, the fab may continue to produce memory for Intel on a foundry basis, which means that that Intel and Micron will have to come to an agreement (and price may increase). What is important is that Intel has the technical capability to produce 3D XPoint memory in different locations as well.
Intel has already finalized its Optane roadmap for the coming quarters and Micron’s intentions to gain 100% control of the fab near Lehi, Utah, will not disrupt it those intentions, Intel has said.
The full statement is presented below:
Micron’s statement is a pre-announcement. They can’t officially make the call until January 1, 2019. The operation of the IMFT factory would not change until after the close of the call, which is at Intel’s discretion for up to one year. There is no near-term change to Intel’s plans in the coming quarters—this has been part of our planning for some time now. Intel has a number of manufacturing options available to us within the time window. We’ve been shipping a broad portfolio of Intel Optane technology products for over a year with a continually expanding product line. We will continue to lead the industry with this exciting new technology.
Related Reading
- What Next for 3D XPoint? Micron to Buy Intel's Share in 3D XPoint Fab
- Intel and Micron To Dissolve 3D XPoint Partnership After 2019
- Intel And Micron To Discontinue Flash Memory Partnership
- Micron Starts Construction of Its Third NAND Fab in Singapore
- Micron Discontinues Lexar Business, Plans to Focus on Higher-Margin Products
Facebook
Twitter
RSS
26 Comments
View All Comments
ilt24 - Monday, October 22, 2018 - link
@BurntMyBacon..."Given the Intel's current fabrication capacity and demand"Intel has a large underutilized 300mm fab in New Mexico. The last major update to the fab was to 32nm logic. It has more recently added R&D and initial production for silicon photonics. Last month Intel announced that they were bringing future memory development to this plant. I imagine this along with the fab in China are the logical places for Intel to produce it's memory once their breakup with Micron is complete.
PaoDeTech - Monday, October 22, 2018 - link
It means:- Micron will be the only manufacturer of Optane - 3D-Xpoint since Intel needs its fabs for their core CPU business / NAND. Intel is already spread too thin (CPU shortages and all).
- Intel will keep developing Optane, but more importantly it will figure out how to use "Optane memory sticks" in Intel servers to provide a 100X(?) addressable memory expansion for specialized workloads.
- AMD will eventually do the same by supporting Micron 3D-Xpoint 2rd generation hyper scale RAM on their server, but at a fraction of Intel price.
- Everybody will make money: Intel, Micron, AMD and the customers that need such massive amount of "level 2 RAM"
- Some PC gamers will buy 3D-Xpoint HDD just because their father will give them the money. But they will have their behind handed to them in Fortnite regardless.
Trust me, I'm a futurologist.
edzieba - Monday, October 22, 2018 - link
Intel can't just up and switch an existing fab to XPoint, it's a rather different chemical process to Silicon IC fabbing so will require extensive retrofitting.The current fab is split 50/50 in capacity between Intel and Micron. Intel are selling every die that can fab, and Micron is... doing something with theirs (which everyone assumes is 'selling them to Intel who actually have buyers'). Whether price needs to move up depends on how long it takes for a fab conversion (or building a whole new fab, depending on just how extensive the modifications are for Chalcogenide processing): Intel now have a 14 month clock to work to before Micron take over the fab, and possibly longer depending on if Intel can muscle them into a limited-time at-cost sale agreement. If that time is not sufficient for Intel to get their own fab up and running, prices may increase, otherwise I'd expect them to decrease with Micron cut out as a middleman for half the production run.
Gondalf - Monday, October 22, 2018 - link
You forget Intel has doubled the clean room of Dalian Fab, so it can meet the production of XPoint easily without retofit Rio Rancho to another chemical process. No problem man.HStewart - Monday, October 22, 2018 - link
I think most people on the net are blind about Intel, and don't realize they have other fabs to do the job. Intel would not have sold the lab to Micron if they actually needed it. My guess they have something better to handle the job.PaoDeTech - Monday, October 22, 2018 - link
I believe that "people on the net" know that Intel has many high value fabs. They also know that fabs are very expensive and apparently Intel is currently maxed out. Yes, they can convert a fab to produce XPoint (and they need to say so to keep Micron behaving...). But should they? Is that the best use of capex? Is 10nm and stopping EPYC from eating their data center lunch more important? Intel has a lot to deal with already. I don't see an obvious answer forming in my little brain.HStewart - Monday, October 22, 2018 - link
I think people like to assume Intel is more expensive and make out and that EPYC is a serious threat. But even though it good to have competition, Intel is a lot bigger than most think. In some ways I think this delay in 10nm and rise of AMD Cpu's and even ARM will be good thing in long run. It gives Intel ambition to focus on the new products and have a product that will significantly better. I am not gullible person that believe just because node size is smaller that it means that it better product.But lets say next year, Intel comes back with product that is so much better than competition, on both the high end and also lower power (ARM) and then people that hate Intel, will be crying Monopoly, Antitrust and such - but they are the one that cause this.
It would be very foolish to believe Intel is sitting back and doing nothing about this stuff.
PaoDeTech - Monday, October 22, 2018 - link
All I'm saying is that Intel has a lot on its plate: whatever is after 14nm, CPU shortages, competition from AMD - NVIDIA - ARM - APPLE, a totally new GPU architecture, Optane optimization and customization, new CPU architecture that's cost efficient at >14nm, chipsets, NAND, spectre, meltdown, the ending of Moore law... Maybe, just maybe that ramping up a 3D XPoint Fab is not high priority if Micron is willing and able to keep selling them wafers.sa666666 - Monday, October 22, 2018 - link
No matter what, you find a way to defend Intel. Even if they were being flushed down the toilet, you'd probably say that it's a good idea, since the toilet needed to be cleaned anyway. I've never encountered someone that was so enamoured with a company that they will justify _anything_ that they do. I really hope you're getting paid for this; you're certainly earning it.ilt24 - Monday, October 22, 2018 - link
@edzieba...." Intel now have a 14 month clock "Probably quite a bit more. While Micron might initiate it's takeover in January it's going to take some 6 to 12 months for them to close the separation deal. That separation deal would include pricing future wafers from Micron after the initial 12 months.