This seems aggressive....What are you hearing on 10nm wpm by year end 2020? I want to see if what I'm hearing aligns with what you're hearing.
-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
Discussion Intel current and future Lakes & Rapids thread
Page 245 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
This seems aggressive....What are you hearing on 10nm wpm by year end 2020? I want to see if what I'm hearing aligns with what you're hearing.
uzzi38
Platinum Member
AMD has officially stated their cadence is 12 to 18 months.
I still don't get where this whole "Zen 3 delayed" mindset comes from.
I'm sure you didn't know what 'cooked' actually means in this context, but damn, you've really made a funny one here 🙂
for real competition's sake, let's hope it's not 😉
It originates from desires - which is strange, because the people throwing it around are actually smart. Most of you people here are 🙂
[PATCH] x86/cpu: Add Lakefield, Alder Lake and Rocket Lake to Intel family
www.mail-archive.com
In case you were unconvinced that Alder Lake is using Hybrid, Intel has now confirmed it.
arandomguy
Senior member
I'd think the likelihood is delays, industry wide, due to the impact of Covid. They might not be considered delays based on the technicality that no firm timelines have yet be set but I find it hard to imagine the announcement, launch and release schedule for products would be identical in a non-Covid timeline.
Everything from design, bringing products to market, supply chain, demand changes requiring different prioritization have been impacted.
NostaSeronx
Diamond Member
Info is only on 7nm, however, it largely depends on the tools Intel has ordered to sustain volume. For 10nm to be viable for overlap with 14nm and 7nm it will need around 100,000 wpm.
"10nm ~80K wpm & 10nm+ ~100K wpm"
With 2020 being 25% the max target, thus for 2020 it is around ~125K wafer starts.
Maximum total available is expected to ~180K(+60 if there is another fab I missed) eventually.
7nm Fab = 10nm Fab plus a EUV corridor of NXE:3400C machines // Essentially 7nm uses the high yield portion of 10nm's BEOL and replaces the low yield of 10nm FEOL/MOL with high yield 7nm of the same type. High 10nm production also in turn technically means high 7nm production, if the correct HVM EUV scanners are present.
In that context means the design kit is actually above 1.0 in the 1.x range and physical design kit is above 1.6 and probably in the 2.x range.
Intel at least yoinked a customer from (Samsung Foundry)SAFE 7nm-4nm to their FIP 7nm.
Last edited:
ajc9988
Senior member
It comes from people's ... we'll go with imaginations.
About the time that intel's road map dropped that no longer had rocket lake s on it, there were two digitimes articles saying zen 3 was pushed to 2021 and might even be on 5nm.
People also couldn't understand the purpose of releasing the XT models.
So to make sense of the minimal bump that XT models got, they bought into the notion that zen 3 is delayed. They then also speculated AMD will do a major price hike and say buy Zen 2 XT for cheaper parts, even though AMD has said those are enthusiast SKUs, they expect low volume, and they are in all likelihood are to make AMD look slightly more competitive in some tasks with the 10th series coming out of Intel.
They then ignore AMD has limited fab time, but are releasing GPU products first to compete against Ampere, are releasing console APUs right after that, then releasing CPUs, all within like 2-3 months from each other (September through October or November). Next year, at least console volume will be less in Flux.
So, they want Intel to be their savior, making claims of stagnation and delays, even though AMD, after the digitimes articles, reached out to tech press and YouTubers and asked them specifically to tell the public they are on time and will be out this year.
And so they then changed it from not meeting cadence to on that cadence releasing Zen 4 in Q1 2022 as a "got 'em" moment, ignoring the numerous times that Intel has pushed a quarter or two and that this isn't pushing a quarter but staying on cadence. Even i think it will be Q1 2022. But that isn't late.
All the slide shows is a 2022 at the end. Until now, that has meant before that year on other AMD slides (which also suggests RDNA 3 next year, just saying which is abnormal because GPUs can take 2 years normally).
But what that really is is being pedantic.
Neither do I. And unless mistaken, desktop Zen 4 is slated for early 2022 so I don't quite see what's the problem with an October 2020 Zen 3 release. But then again I won't claim I am 100% sure I am up to speed with either vendor's roadmaps.
uzzi38
Platinum Member
The next person to say Zen 3 is delayed gets a slap.
How many times does this stuff need to be reiterated until people get the idea?
trexfromouterspace
Member
The real Zen 3 was the friends we made along the way
The next person to say Zen 3 is delayed gets a slap.
How many times does this stuff need to be reiterated until people get the idea?
AFAIK, some guys that restricted by NDA has tested Zen3 ES which is made in 2019, so a release before 2021 is plausible.😛
(wait, what? this is intel icelake thread?@_@)
DrMrLordX
Lifer
AMD didn't say "our launch cadence is 12-18 months" when they released a roadmap of Zen2 in 2019, Zen3 in 2020, and Zen4 in 2021 (a roadmap they have never retracted). If Zen3 launches so late that Zen4's market window is now 2022, it's late. I don't know how many other ways I can say this, or why we're discussing this in an Intel thread, when the whole reason why this got started was that I was reacting to one of @Dave2150 's posts where he seemed to think Intel's upcoming releases (which are horribly late) are somehow exciting, when they really aren't.
uzzi38
Platinum Member
AnandTech Forums: Technology, Hardware, Software, and Deals
Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
www.anandtech.com
Mark Papermaster: We’re on a 12-18 month cadence, and we believe that is sustainable. It’s what the industry demands from us.
DrMrLordX
Lifer
@uzzi38
www.techspot.com
So I'm corrected, Papermaster actually said that BEFORE AMD pushed their roadmap showing Zen4 in 2021. How irresponsible of them. Specifically:
Zen4 is slated for 2021. Period. If they miss then it's late, and part of the reason why it may be late is that they couldn't/didn't want to get Zen3 out the door quickly enough. They will have 14 months at most to make good on their roadmap claims before Zen4 becomes a 2022 CPU. It's not like they're Intel and they're stuck on an old node or limping along on a newer, flawed node.
And let's be realistic, anyone who thinks that Alder Lake-S will be out in Q4 2021 will probably be due for disappointment.
AMD's CPU roadmap reveals Zen 3, Zen 4, and 5nm
The consumer and enterprise CPU roadmaps were unveiled at AMD's Financial Analyst Day. In the former category, the graph only goes as far as Zen 3. The...
www.techspot.com
So I'm corrected, Papermaster actually said that BEFORE AMD pushed their roadmap showing Zen4 in 2021. How irresponsible of them. Specifically:
Zen4 is slated for 2021. Period. If they miss then it's late, and part of the reason why it may be late is that they couldn't/didn't want to get Zen3 out the door quickly enough. They will have 14 months at most to make good on their roadmap claims before Zen4 becomes a 2022 CPU. It's not like they're Intel and they're stuck on an old node or limping along on a newer, flawed node.
And let's be realistic, anyone who thinks that Alder Lake-S will be out in Q4 2021 will probably be due for disappointment.
DrMrLordX
Lifer
Hey, at least I mentioned Alder Lake.
Speaking of delays: Tiger Lake-U is allegedly launching on Sept. 2nd. Any bets on if Rocket Lake will make it onto the market before April/May 2021?
tamz_msc
Diamond Member
Rocket Lake brings new platform features, it depends on how quickly board partners validate for those features; so it's likely that you're right - no Rocket Lake before Q1'21.
DrMrLordX
Lifer
I'm thinking Q2 might be possible though. Comet was massively delayed if you believed the initial reports of it hitting in Q4 2019 (it didn't). Intel may do well to get LGA1700 ready by March 2021.
Yet @Dave2150 says intel will be the one laughing when they finally release a 10nm desktop product (Alder Lake) in the same general time frame AMD will release an 5nm one. He must know something we're not aware of.
tamz_msc
Diamond Member
Was PCIe 4.0 on Z490 always planned? If it wasn't then that might explain the Comet Lake delay. I'm not too keen to believe the rumors regarding power delivery causing delays.
Anyway Rocket Lake brings some much needed platform features - dedicated x4 PCIe for NVMe and x8 DMI link to chipset, but I would hope that those aren't too difficult to get right once PCIe 4.0 support has been laid out.
coercitiv
Diamond Member
I still believe Intel will attempt one early SKU launch to mitigate the media fallout from the limited 2020 AMD launch. However, until I see Intel becoming a lot more talkative about their future plans I will always expect delays for high volumes, no matter the node.
Zen 3 should be enough motivation for them to keep it to around CES. But anything is possible.
Delay was I believe because they delayed production to open up room for more Cascade Lake XCC.
Icelake SP Geekbench 5 submission - Linux
Stepping 4 -
Single - 1089
Multi -27708
browser.geekbench.com
Stepping 4 -
Single - 1089
Multi -27708
Supermicro SYS-620C-TN12R - Geekbench
Benchmark results for a Supermicro SYS-620C-TN12R with an Intel $0000% processor.
browser.geekbench.com
tamz_msc
Diamond Member
Interesting cache configuration. L2 is 1.25MB/core like TGL up from 512KB/ core in ICL-U but L3 is down from 2MB/core on consumer ICL-U to 1.5MB/core.Icelake SP Geekbench 5 submission - Linux
Stepping 4 -
Single - 1089
Multi -27708
Supermicro SYS-620C-TN12R - Geekbench
Benchmark results for a Supermicro SYS-620C-TN12R with an Intel $0000% processor.
Looks like Intel may be focusing on throughput over latency with the server parts. Wonder if L3 is a victim cache like SKL-SP.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
