You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Here is what Intel published before the meeting on 7nm and 10nm:
10nm Process Technology: Intel's first volume 10nm processor, a mobile PC platform code-named “Ice Lake,” will begin shipping in June. The Ice Lake platform will take full advantage of 10nm along with architecture innovations. It is expected to deliver approximately 3 times faster wireless speeds, 2 times faster video transcode speeds, 2 times faster graphics performance, and 2.5 to 3 times faster artificial intelligence (AI) performance over previous generation products. As announced, Ice Lake-based devices from Intel OEM partners will be on shelves for the 2019 holiday season. Intel also plans to launch multiple 10nm products across the portfolio through 2019 and 2020, including additional CPUs for client and server, the Intel® Agilex™ family of FPGAs, the Intel® Nervana™ NNP-I (AI inference processor), a general-purpose GPU and the “Snow Ridge” 5G-ready network system-on-chip (SOC).
Building on a model proven with 14nm that included optimizations in 14+ nm and 14++ nm, the company will drive sustained process advancement between nodes and within a node, continuing to lead the scaling of process technology according to Moore’s Law. The company plans to effectively deliver performance and scaling at the beginning of a node, plus another performance improvement within the node through multiple intra-node optimizations within the technology generation.
7nm Status: Intel's 7nm process technology that will deliver 2 times scaling and is expected to provide approximately 20 percent increase in performance per watt with a 4 times reduction in design rule complexity. It will mark the company’s first commercial use of extreme ultraviolet (EUV) lithography, a technology that will help drive scaling for multiple node generations.
The lead 7nm product is expected to be an Intel Xe architecture-based, general-purpose GPU for data center AI and high-performance computing. It will embody a heterogeneous approach to product construction using advanced packaging technology. On the heels of Intel’s first discrete GPU coming in 2020, the 7nm general purpose GPU is expected to launch in 2021.
It's impressive in volume Intel is arguably the best in manufacturing high performance semiconductors.
I'm not sold on 7nm. It's Maricopa county and is really authoritarian, not a place for engineers. Let's be honest the guy that knows how the equipment work and maintains it may not be the owner but it can be his company more or less.
Impressive? Intel has been trailing TSMC and Samsung for a long time now in volume and process technology. TSMC does ~1 million 7/10nm wafers a year, or around 500 million chips. In 2018 1.4 billion mobile phones were sold, but only 260 million PCs (desktops, laptops etc). Today the most advanced high performance semiconductors are in mobile phones (which have CPUs which will beat your laptop or even desktop).
Impressive? Intel has been trailing TSMC and Samsung for a long time now in volume and process technology. TSMC does ~1 million 7/10nm wafers a year, or around 500 million chips. In 2018 1.4 billion mobile phones were sold, but only 260 million PCs (desktops, laptops etc). Today the most advanced high performance semiconductors are in mobile phones (which have CPUs which will beat your laptop or even desktop).
Maybe EUV made Intel 7nm much easier to develop/ramp? TSMC seems to think so. I'm wondering why they did not mention 7nm FPGAs? Hopefully they will be sampling in 2021. They won't need HVM FPGA until a year or two after that due to the design cycle.
I don't think they can find the engineers and mechanics willing to work at the Maricopa campus. The place is too authoritarian, same is true in Washington county. I don't expect much from 7nm with intel.
Impressive? Intel has been trailing TSMC and Samsung for a long time now in volume and process technology. TSMC does ~1 million 7/10nm wafers a year, or around 500 million chips. In 2018 1.4 billion mobile phones were sold, but only 260 million PCs (desktops, laptops etc). Today the most advanced high performance semiconductors are in mobile phones (which have CPUs which will beat your laptop or even desktop).
1. Mobile SoC are not considered as high performance. And it only beats laptop / desktop in certain cases.
2. Intel has and still leads in high performance node AND volume, that is the 14nm I don’t know how many plus sign behind it. Although that is likely not the case once AMD launch 7nm Zen2.
My guess would be that is the only product line they desperately need to compete. It is a new segment ( GPGPU ) and Intel will need invest ( a lot ) into it. After all Nvidia aren’t sitting still, we all know how competitive Jensen is.
1. Mobile SoC are not considered as high performance. And it only beats laptop / desktop in certain cases.
2. Intel has and still leads in high performance node AND volume, that is the 14nm I don’t know how many plus sign behind it. Although that is likely not the case once AMD launch 7nm Zen2.
They were not considered high performance. A modern mobile phone core is more advanced than Skylake. Mobile performance still increases by 30% per year (every year people say it is not possible, and they are proven wrong). All of this is possible by process improvements and rapid micro architecture development, instead of just chasing frequency blindly.
It's impressive in volume Intel is arguably the best in manufacturing high performance semiconductors.
I'm not sold on 7nm. It's Maricopa county and is really authoritarian, not a place for engineers. Let's be honest the guy that knows how the equipment work and maintains it may not be the owner but it can be his company more or less.
AZFSM do an outstanding job at making wafter and have a great track record. There is no process node development happening in Maricopa country, it's all high volume manufacturing there. That's all done in Oregon.
I think Intel has more to gain in GPU than CPU short term. The cloud is being populated with Intel CPUs and Nvidia/AMD GPUs. Intel wants this business. I also do not think AMD is putting as much pressure on Intel CPUs as advertised.
If you look at one of the comments: "[FONT="]AMD Milan will have an amazing 15 chiplets."[/FONT]
My speculation is these AMD Milan servers in 2020 will be 1 IO die, 10 CPU die, AND 4 GPU die on infinity fabric. THIS IS WHY INTEL NEEDS GPUS STAT.... Intel is in a heap of trouble.
I think Intel has more to gain in GPU than CPU short term. The cloud is being populated with Intel CPUs and Nvidia/AMD GPUs. Intel wants this business. I also do not think AMD is putting as much pressure on Intel CPUs as advertised.
AZFSM do an outstanding job at making wafter and have a great track record. There is no process node development happening in Maricopa country, it's all high volume manufacturing there. That's all done in Oregon.
Cool story david duke. I work in manufacturing so i know the culture. Both washington county and maricopa county are too authoritarian to succeed. The new thing is to harass "ethics" by calling the cops on them. if you want to live there get a lawyer the specializes in neighbor disputes.
10 nm with intel is manufactured with ai and both places have no answers. Let's be honest there's a lack of people willing to do the work let alone trustworthy.
you're doing the same thing goebbels did, repeat the same lie (that's all done in oregon) over and over again. Oregon has never had a significant successful venture without israeli help in engineering in my lifetime at least. The engineers just don't have the firepower and it probably will change the.demographics of the state.
This is laughable PR, trying to say Intel 14nm is competing against TSMC 10nm. Except TSMC has been on 7nm for almost a year now and very few products are still on 10nm, while Intel is still not shipping 10nm parts in any volume. Intel's financial engineering is also falling apart, since they changed their depreciation schedule based on the Tick-Tock-Tock cadence, so they will need to take a massive writedown on 10nm.
Austrians discovered another flaw (Zombieload): CPU.fail
What that means: To software-patch this hardware flaw, is lower performance for cloud-providers.
So they need to buy even more Xeons to make up for this loss. More demand means, Intel could try to price those Xeons higher, and just quit making Core i3's altogether (leave that market for AMD, as there currently is no fab capacity to make more Xeons _and_ provide for i3's _and_ discrete GPU's.
But AMD doesn't seem to have these new flaws! So if customers can't buy Xeon's because of shortages, or they are more expensive as customers are competing for INTC fab capacity, guess which alternatives they'd be looking at...
So it looks like INTC's problem just got way worse!
