Intel Says Process Tech to Lag Competitors Until Late 2021, Will Regain Lead with 5nm (archive)
It appears that 2020 and 2021 are going to be long years for Intel. CFO George Davis presented at the Morgan Stanley conference yesterday covering a wide range of topics, but noted that despite being "undoubtedly in the 10nm era," the company felt that it would not reach process parity with competitors until it produces the 7nm node at the tail end of 2021. Davis also said that Intel wouldn't regain process leadership until it produces the 5nm node at an unspecified date.
Davis commented that the company was "definitely in the 10nm era" with Ice Lake client chips and networking ASICs already shipping, along with the pending release of discrete GPUs and Ice Lake Xeons. Intel is also moving well along the path of inter-node development, which consists of "+" revisions to existing processes. Davis said the 10nm inter-node step provides a "step-function move" with the Tiger Lake chips based on the 10nm+ process as the company awaits its 7nm process.
However, Davis noted that in spite of the shipping products and pending "+" revisions to the 10nm process, its process node still lags behind competitors, stating:
"So we bring a lot of capability to the table for our customers, in addition to the CPU, and we feel like we're starting to see the acceleration on the process side that we have been talking about to get back to parity in the 7nm generation and regain leadership in the 5nm generation."
Previously:
Intel Launches Coffee Lake Refresh, Roadmap Leaks Showing No "10nm" Desktop Parts Until 2022
Intel's Jim Keller Promises That "Moore's Law" is Not Dead, Outlines 50x Improvement Plan
Intel Roadmap Shows Plans for "5nm", "3nm", "2nm", and "1.4nm" Process Nodes by 2029
Related Stories
Intel 9th Gen Core Processors: All the Desktop and Mobile 45W CPUs Announced
Dubbed 'Coffee Lake Refresh', the 9th generation of Intel's Core CPU product line is a direct refresh of its 8th generation Coffee Lake hardware, with minor enhancements such as a better thermal interface on the high end processors, support for up to 8 cores, and newer chipsets with integrated USB 3.1 Gen2 (10Gbps) and CNVi-enabled Wi-Fi. The hardware is still fundamentally the original 6th Gen Skylake microarchitecture underneath, from 2016, but built on Intel's latest 14nm process variant, in order to extract additional frequency and efficiency, and with more cores at the high-end.
Intel may continue to be largely stuck on a "14nm" process for years to come:
The latest roadmaps come from Tweakers and detail both the Client Commercial CPU products and the Client Mobile CPU products which would be introduced in the future. The authenticity of these roadmaps cannot be confirmed but they are referenced back to the Intel's SIP program and DELL so there might be some legitimacy to them.
[...] It looks like Intel will stick with 14nm++ for a while as the roadmap reveals. Around Q2 2020, Intel will launch their Comet Lake-S processors, featuring up to 10 core SKUs. These would be followed by Intel's Rocket Lake-S parts which would also be based on an optimized 14nm process node. It looks like we can expect a 10nm or sub-10nm part from Intel only around 2022 which is about the same time Intel is expected to launch their Ocean Cove CPU architecture.
Ocean Cove is a future chip architecture under development at Intel which will launch after Golden Cove (2021), the successor to Willow Cove (2020) which itself is the successor to Intel's Sunny Cove (Ice Lake) core's architecture.
The roadmap shows Intel using "10nm" sooner for some mobility (laptop) CPUs.
Intel's Senior Vice President Jim Keller (who previously helped to design AMD's K8 and Zen microarchitectures) gave a talk at the Silicon 100 Summit that promised continued pursuit of transistor scaling gains, including a roughly 50x increase in gate density:
Intel's New Chip Wizard Has a Plan to Bring Back the Magic (archive)
In 2016, a biennial report that had long served as an industry-wide pledge to sustain Moore's law gave up and switched to other ways of defining progress. Analysts and media—even some semiconductor CEOs—have written Moore's law's obituary in countless ways. Keller doesn't agree. "The working title for this talk was 'Moore's law is not dead but if you think so you're stupid,'" he said Sunday. He asserted that Intel can keep it going and supply tech companies ever more computing power. His argument rests in part on redefining Moore's law.
[...] Keller also said that Intel would need to try other tactics, such as building vertically, layering transistors or chips on top of each other. He claimed this approach will keep power consumption down by shortening the distance between different parts of a chip. Keller said that using nanowires and stacking his team had mapped a path to packing transistors 50 times more densely than possible with Intel's 10 nanometer generation of technology. "That's basically already working," he said.
The ~50x gate density claim combines ~3x density from additional pitch scaling (from "10nm"), ~2x from nanowires, another ~2x from stacked nanowires, ~2x from wafer-to-wafer stacking, and ~2x from die-to-wafer stacking.
Related: Intel's "Tick-Tock" Strategy Stalls, 10nm Chips Delayed
Intel's "Tick-Tock" is Now More Like "Process-Architecture-Optimization"
Moore's Law: Not Dead? Intel Says its 10nm Chips Will Beat Samsung's
Another Step Toward the End of Moore's Law
Intel's Manufacturing Roadmap from 2019 to 2029: Back Porting, 7nm, 5nm, 3nm, 2nm, and 1.4 nm
One of the interesting disclosures here at the IEEE International Electron Devices Meeting (IEDM) has been around new and upcoming process node technologies. Almost every session so far this week has covered 7nm, 5nm, and 3nm processes (as the industry calls them). What we didn't expect to see disclosed was an extended roadmap of Intel's upcoming manufacturing processes.
[...] Intel expects to be on 2 year cadence with its manufacturing process node technology, starting with 10nm in 2019 and moving to 7nm EUV in 2021, then 5nm in 2023, 3nm in 2025, 2nm in 2027, and 1.4 nm in 2029. This is the first mention on 1.4nm for Intel on any slide, so this confirms where Intel is going, and just for context, if that 1.4nm is indicative of any actual feature, would be the equivalent of 12 silicon atoms across.
It is perhaps worth noting that some of the talks at this year's IEDM features dimensions on the order of 0.3nm with what are called '2D self-assembly' materials, so something this low isn't unheard of, but it is unheard of in silicon. Obviously there are many issues going that small that Intel (and its partners) will have to overcome.
Inbetween each process node, as Intel has stated before, there will be iterative + and ++ versions of each in order to extract performance from each process node. The only exception to this is 10nm, which is already on 10+, so we will see 10++ and 10+++ in 2020 and 2021 respectively. Intel believes they can do this on a yearly cadence, but also have overlapping teams to ensure that one full process node can overlap with another.
The interesting element to this slide is the mention of back porting. This is the ability for a chip to be designed with one process node in mind, but perhaps due to delays, can be remade on an older '++' version of a process node in the same timeframe. Despite Intel stating that they are disaggregating chip design from process node technology, at some point there has to be a commitment to a process node in order to start the layouts in silicon. At that point the process node procedure is kind of locked, especially when it goes to mask creation.
What's wrong with Intel, and how to fix it: Former principal engineer unloads (archive)
In a blunt video posted late Thursday evening, outspoken former Intel principal engineer Francois Pidnoel offered his advice on how to "fix" Intel CPUs, criticized current leadership for not being engineers, said AVX512 was a misadventure, and declared that it's only luck AMD hasn't grabbed more market share.
"First, Intel is really out of focus," Piednoel said in the nearly hour-long video presentation. "The leaders of Intel today are not engineers, they are not people who understand what to design to the market."
[...] Pidnoel flat-out dismissed including AVX512 in consumer chips as a mistake. "You had Skylake and Skylake X for a reason," Piednoel said. "AVX512 is designed for a race of throughput that is lost to the GPU already. There's two ways to get throughput. One is to get the throughput is by having larger vectors to your core, and the other way is to have more cores."
[...] "Intel is very lucky AMD cannot get the volume, to be able to compete," Piednoel. "If they were getting volume, the price difference would definitely cost Intel market share a lot more than what they are losing right now."
Related: AVX-512: A "Hidden Gem"?
Intel CEO Blames "10nm" Delays on Aggressive Density Target, Promises "7nm" for 2021
Intel's Process Nodes Will Trail Behind Competitors Until at Least Late 2021
Linus Torvalds: Don't Hide Rust in Linux Kernel; Death to AVX-512
Intel Engineering Chief Out After 7nm Product Delays
Intel Faces Class-Action Lawsuit Over "7nm" Delays
(Score: 2) by Booga1 on Wednesday March 04 2020, @10:27PM (6 children)
Well, this was exactly the kind of news I was going to start digging for in the next few weeks.
A friend of mine is considering building a new computer and he's seriously considering AMD, even though he's been an Intel user for decades. Performance, power, and features like PCI 4.0 are all in favor of AMD now. There's no way he'll wait until the end of 2021 just for Intel to catch up.
(Score: 2) by takyon on Wednesday March 04 2020, @10:50PM (2 children)
If one company leapfrogs the other in performance, the other company can try to compete on price. Except with Intel, the CPUs currently use 2x the energy or worse to hit those high clock speeds. Even if you think Intel's upcoming 10-core 10900k [wccftech.com] is a good idea, the motherboard will be more expensive, you'll have to pick up a cooler, and it will use a lot more power under load.
I think the consideration right now should be whether to stay on the AM4 socket or wait for AM5. AMD's Zen 3 should be out by the end of this year and will be the last on AM4. You could build with a cheap Zen+/Zen2 chip now and switch to Zen 3 later. The AM5 socket should switch to DDR5 and maybe PCIe 5.0, but it will all be more expensive initially (especially the memory). That could offer an upgrade path to... Zen 6?
For your friend, I would recommend AM4, see if something cheap like a Ryzen 2700X will beat whatever he is using now, and he could upgrade to up to 16-core Zen 3 years later when that is cheaper, assuming the motherboard can handle it.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 3, Interesting) by Booga1 on Wednesday March 04 2020, @11:03PM (1 child)
He's willing to spend money to build what he wants now, but he's not very patient. I expect he'll settle for a Zen2 chip or MAYBE I can talk him into waiting for the holiday season if Zen3 looks like it'll arrive on time. This coronavirus thing could be a big disruption if things get bad.
(Score: 4, Interesting) by takyon on Thursday March 05 2020, @03:42AM
Yeah, coronavirus could delay product launches by months across the tech industry, or result in shortages/paper launches.
Zen 3 may have a more staggered launch than Zen 2 did, with some chips coming out this year and maybe the 12 and/or 16-core coming out next year.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by Grishnakh on Thursday March 05 2020, @03:30AM (2 children)
Your friend should definitely wait for Intel to catch up, or just get one of their current systems even if it costs more and uses more power and performs worse.
After all, if he's been a die-hard Intel fan for "decades", that means he must have been an Intel user during the P4/RAMBUS days, when an Intel system cost far more than a comparable AMD system and used way more power too because the P4 was such a power hog. If he was that blinded by corporate loyalty to not jump ship back then, why should he change now?
(Score: 2) by toddestan on Thursday March 05 2020, @04:23AM
The AMD's back then were faster and cheaper, but you were often saddled with crappy VIA chipsets and SDRAM. The Intel systems cost more, but you got something that was at least stable. I remember those days, the old Socket A systems never quite ran right (including AMD's own efforts at a chipset) until the end when nVidea jumped in with the nForce. And that was only really temporary, as the next nVidia chipsets for AMD's 64-bit offerings were terrible. I'd also call the power difference a wash - the 64-bit AMD chips ran cool, but the original Athlons that competed against the P4/Rambus systems ran crazy hot.
I remember back about 10-12 years ago when those later 32-bit PC's no longer had any real economic value, but were still good enough for someone to use, I'd fix up the P4 systems to give away, and get parts by scrapping the AMD systems because they just weren't worth dealing with. At least Windows 7 would run on a Pentium 4, if a bit sluggish, whereas getting anything past Windows XP to run on an Athlon XP (I see what you did there AMD...) was a bit of a crapshoot.
I ran AMD systems for years, but in 2012 I finally had enough and built my first Intel system. At this point I'm not sure whether it would be best to try rolling the dice on AMD or to just get another Intel system. Luckily for me the system I built in 2012, despite its age, should be good for a few more years so I can just wait it out.
(Score: 2) by Booga1 on Thursday March 05 2020, @06:28AM
You're correct, he did have a Pentium 4 system with Rambus RAM. I also had one, but we got them at the bargain basement price of free when the company he worked for went under during the dot-com collapse. (The 72 port switch with 3 gigabit back plane was fantastic for LAN parties.)
He's considering the change now because he keeps hearing about all the good reviews regarding AMD's latest updates to the Ryzen processor line. I built a Ryzen 1700X system two years ago for half the cost of his eight year old Intel rig. My machine outperforms his on every single metric except the video card and I know that bugs the crap outta him.
I even suggested he stick with Intel because it's what he's most familiar with, but the lure of AMD seems to be calling him. AMD has been playing their cards right if they're starting to sway hardcore Intel fans.
(Score: 0) by Anonymous Coward on Wednesday March 04 2020, @10:39PM
We are projected to be all caught up with today's standars in 2 years. By then hopefully the competition hasn't advanced any and we would be all caught up.
(Score: 0) by Anonymous Coward on Wednesday March 04 2020, @11:18PM (3 children)
Process tech relies on volume, and Intel focusing on processor chips can't compete with contract foundries (TSMC) and dram foundries (Samsung).
(Score: 2) by takyon on Wednesday March 04 2020, @11:29PM
TSMC is laying down the cash [tomshardware.com] it needs to stay at the head of the pack:
Intel is still trying to increase its "14nm" capacity [guru3d.com].
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by Grishnakh on Thursday March 05 2020, @03:38AM (1 child)
It used to be that Intel could do both. They could focus on processor chips on one hand, and process tech on the other; they had enough volume and profitability to do this because they were a near-monopoly, and also because everyone was buying a new computer every 2 years because the technology was moving so fast. These days, it's slowed down; an 8-year-old laptop is really fine for basic tasks (web, office, etc.), and isn't really noticeably slower than a brand-new one, though it will have worse battery life. The growth is in mobile devices, and a lot of people now are even giving up PCs and relying solely on phones and/or tablets. Intel totally missed the boat on mobile CPUs. They tried; they did StrongARM/XScale for a little while, but then sold that off, they had some other ARM-based mobile CPUs for a while, then sold that to Marvell, then they tried Atom, but that didn't go too far. Now ARM is unbeatable for mobile phones. So now they're the dominant player in a shrinking market, and their long-time competitor, while smaller, keeps getting better.
(Score: 3, Interesting) by takyon on Thursday March 05 2020, @03:47AM
Intel's problem is not the slowdown of the PC market (which has stabilized). They have had capacity shortage problems for years now.
https://articles2.marketrealist.com/2019/12/intels-cpu-shortage-impact-chip-giants-pc-makers/ [marketrealist.com]
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 0) by Anonymous Coward on Thursday March 05 2020, @06:18AM (3 children)
Why is Intel so far behind?
Is there some critical trick they don't have to make the next density jump?
(Score: 2) by Freeman on Thursday March 05 2020, @05:01PM (2 children)
It's called, having to focus on security issues, while developing the next latest and greatest thing. Which likely were using said tricks, but left gaping security holes. So, they can't use those tricks anymore.
Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
(Score: 0) by Anonymous Coward on Thursday March 05 2020, @07:45PM (1 child)
I think the trend seems to be that you get your product out first and then try to focus on making it secure later. From a marketing perspective that makes a lot more sense. It's easier to start with and sell a new less secure product now that does more and is easy to use and to try to make it more secure later than to start with a more secure product that does less and is harder to use and then try to add features later on. That's because it takes longer for users to compare how secure one product is to another than it does for them to compare the functionality and ease of use of one product to another.
(Score: 0) by Anonymous Coward on Thursday March 05 2020, @11:36PM
That is answering an architectural question which drives the schematic. That's another interesting story.
I was asking a process question as to why they are behind at least two different other companies?
It takes both tick and tock to be great in that area. I pointed out tick, you pointed out tock. Both together makes a sad story.