4 ms·

It's not, they've fallen all the way back to 22nm while AMD is at 7nm, while maintaining healthy production capabilities of current process nodes. And that's no

It's not, they've fallen all the way back to 22nm while AMD is at 7nm, while maintaining healthy production capabilities of current process nodes. And that's not even taking into account the fact that AMD architected an objectively better processor

alexhutcheson 8y ago
The article is about chipsets[1], not processors. Intel is only falling back to 22nm for the chipset, which is a less performance-sensitive component than the processor. They are still making 14nm processors. [1] https://en.wikipedia.org/wiki/Chipset https://en.wikipedia.org/wiki/Chipset
- ZiiS 8y ago
  AMD (or rather TSMC) are not (yet) producing 7nm X86 processors for sale. Though you can buy 7nm ARM chips made by TSMC for Apple and Huawei.
- rbanffy 8y ago
  It'd seem they are doing that because they never moved CPUs to 10nm facilities, thus not freeing 14nm capacity for the chipsets and now they want to increase production of 14nm CPUs and they can only move chipsets to 22nm. If the chipsets are simpler, less critical and easier to design than CPUs, shouldn't they start new processes with the least critical stuff?
  - bloomer 8y ago
    There is much less value in producing chipsets on the more advanced process line, which would basically waste the investment in the new process. Also depending on the issue it is likely to affect chipsets and cpus similarly so there may not even be a manufacturing advantage.
  - monocasa 8y ago
    Typically the new process nodes are seen as a rapidly depreciating asset. You put the highest return on investment as quickly as you can.
    - rbanffy 8y ago
      But won't you be able to deliver the high-value product sooner if you get early experience with the low-value one first? You can design around unreliable processes by increasing redundancy and these simpler chips seem excellent as technology validation.
      - monocasa 8y ago
        Not for the time frame generally seen until nowish which is a new process every 18 months. If you wait to smooth out your process you'll be left in the dust. Also, easier on the RTL isn't really easier on the process. The digital guys and the process guys have very different constraints, so there's probably not even a benefit to having "simpler" chips on a newer process first.
        rbanffy 8y ago
        Simpler chips also means smaller dies, which, for a new process also means better yield even at a higher defect rate. Not only chipsets could benefit from it, but also smaller processors such as the Atom lineup. I assume a 10nm Quark X would be hard to see with the naked eye ;-)
        monocasa 8y ago
        But the margins are already much lower on the smaller dies. Already something like a quark is next to free on the process level; the recurring cost is nearly all packaging and testing. Additionally there's some practical limits to how small you can make a chip since you still need the analog side of the io pins and their bond pads, and you can't shrink those very much.
tinus_hn 8y ago
It doesn’t matter what AMD is building. The only thing that matters for Intel is how many products they ship because that’s where the money is made.
- ben_w 8y ago
  If Intel’s competitors are building better products than Intel, then Intel ships less products or has to charge less for their products. It matters what AMD produces.
  - rbanffy 8y ago
    What's better for one person may not be better for all others.
    - ben_w 8y ago
      In principle I agree with you. Unfortunately, your response doesn’t really deal with the point I made: not only does it apply equally to both companies, but also I cannot think of any economically significant use case where Intel will beat AMD while costing more ($ or watts) per gflop.
      - loeg 8y ago
        Absolute single thread perf and/or memory latency still matter for many applications. AMD definitely wins on a $/gflop basis, especially at high core counts. But some applications are happy to trade absolute gflops for low latency and pay Intel out the nose for it.
        geezerjay 8y ago
        > But some applications are happy to trade absolute gflops for low latency and pay Intel out the nose for it. Can you provide one example of where Intel products beat AMD's in latency in a way that prospective customers are willing to pay a premium for that performance delta?
yborg 8y ago
Intel is currently projecting full-year revenue for 2018 of $69.5 billion.
- adventured 8y ago
  Intel second quarter profit: $5 billion AMD full-year 2017 revenue: $5.3 billion You see, Intel is gasping for breath, drowning in the $18 billion of profit they'll generate this year. How they'll survive another year at this rate is anyone's guess.
  - losteric 8y ago
    Gasping for breath != dead; Intel is stumbling here, AMD will start eating into those numbers unless Intel can regain their pace.

alexhutcheson 8y ago

The article is about chipsets[1], not processors. Intel is only falling back to 22nm for the chipset, which is a less performance-sensitive component than the processor. They are still making 14nm processors. [1] https://en.wikipedia.org/wiki/Chipset https://en.wikipedia.org/wiki/Chipset

ZiiS 8y ago
AMD (or rather TSMC) are not (yet) producing 7nm X86 processors for sale. Though you can buy 7nm ARM chips made by TSMC for Apple and Huawei.
rbanffy 8y ago
It'd seem they are doing that because they never moved CPUs to 10nm facilities, thus not freeing 14nm capacity for the chipsets and now they want to increase production of 14nm CPUs and they can only move chipsets to 22nm. If the chipsets are simpler, less critical and easier to design than CPUs, shouldn't they start new processes with the least critical stuff?
- bloomer 8y ago
  There is much less value in producing chipsets on the more advanced process line, which would basically waste the investment in the new process. Also depending on the issue it is likely to affect chipsets and cpus similarly so there may not even be a manufacturing advantage.
- monocasa 8y ago
  Typically the new process nodes are seen as a rapidly depreciating asset. You put the highest return on investment as quickly as you can.
  - rbanffy 8y ago
    But won't you be able to deliver the high-value product sooner if you get early experience with the low-value one first? You can design around unreliable processes by increasing redundancy and these simpler chips seem excellent as technology validation.
    - monocasa 8y ago
      Not for the time frame generally seen until nowish which is a new process every 18 months. If you wait to smooth out your process you'll be left in the dust. Also, easier on the RTL isn't really easier on the process. The digital guys and the process guys have very different constraints, so there's probably not even a benefit to having "simpler" chips on a newer process first.
      - rbanffy 8y ago
        Simpler chips also means smaller dies, which, for a new process also means better yield even at a higher defect rate. Not only chipsets could benefit from it, but also smaller processors such as the Atom lineup. I assume a 10nm Quark X would be hard to see with the naked eye ;-)
        monocasa 8y ago
        But the margins are already much lower on the smaller dies. Already something like a quark is next to free on the process level; the recurring cost is nearly all packaging and testing. Additionally there's some practical limits to how small you can make a chip since you still need the analog side of the io pins and their bond pads, and you can't shrink those very much.

tinus_hn 8y ago

It doesn’t matter what AMD is building. The only thing that matters for Intel is how many products they ship because that’s where the money is made.

ben_w 8y ago
If Intel’s competitors are building better products than Intel, then Intel ships less products or has to charge less for their products. It matters what AMD produces.
- rbanffy 8y ago
  What's better for one person may not be better for all others.
  - ben_w 8y ago
    In principle I agree with you. Unfortunately, your response doesn’t really deal with the point I made: not only does it apply equally to both companies, but also I cannot think of any economically significant use case where Intel will beat AMD while costing more ($ or watts) per gflop.
    - loeg 8y ago
      Absolute single thread perf and/or memory latency still matter for many applications. AMD definitely wins on a $/gflop basis, especially at high core counts. But some applications are happy to trade absolute gflops for low latency and pay Intel out the nose for it.
      - geezerjay 8y ago
        > But some applications are happy to trade absolute gflops for low latency and pay Intel out the nose for it. Can you provide one example of where Intel products beat AMD's in latency in a way that prospective customers are willing to pay a premium for that performance delta?

yborg 8y ago

Intel is currently projecting full-year revenue for 2018 of $69.5 billion.

adventured 8y ago
Intel second quarter profit: $5 billion AMD full-year 2017 revenue: $5.3 billion You see, Intel is gasping for breath, drowning in the $18 billion of profit they'll generate this year. How they'll survive another year at this rate is anyone's guess.
- losteric 8y ago
  Gasping for breath != dead; Intel is stumbling here, AMD will start eating into those numbers unless Intel can regain their pace.