4 ms·
One slide noted that their 7nm processor had 13.28 billion transistors per 331 mm^2. Although 7nm/10nm have mostly become marketing terms, is this something di
by at1as 8y ago
One slide noted that their 7nm processor had 13.28 billion transistors per 331 mm^2.
Although 7nm/10nm have mostly become marketing terms, is this something directly comparable between fabs/companies?
Does anyone have a comparison for Intel? Best I can find is this list: https://en.wikipedia.org/wiki/Transistor_count https://en.wikipedia.org/wiki/Transistor_count (and if that's accurate and up to date, it looks as though AMD will be well ahead).
- at1as 8y ago^ err. Actually, those figures are for the GPU. Nevermind.
- fermienrico 8y agoThe best way to compare CPUs would be performance/watt benchmarks which is what matters ultimately to the customer; in this case, data center customers. Let me clarify - we are not talking about Geekbench type benchmarks. Each customer has their own validation & qualification process for new datacenter chip procurement. Public gets hung up with all kinds of marketing terms (7nm, 7nm+, 10nm, 10nm+, 10nm++, etc). What does 7nm+ even mean!? It is purely a marketing term and large datacenter customers know this. They run their workloads on test samples and make a decision to go with Intel or AMD. Furthermore, there is also the aspect of maintainability, servicing and infrastructure inertia that is priced into Intel's server chips. Apple-to-Apple chip comparison (sorry for the pun, not intended) from Intel & AMD would not be priced the same since Intel knows that there is a giant amount of switching inertia for a customer to switch to AMD Epyc. Furthermore, datacenter customers want predictability and proven performance. In this case, Intel again wins with its history and you betchya its modeled in the pricing. So, this is all business as usual. HN loves beating on Intel but their numbers in quarterly reports depict a different story. Let me repeat: No sane customer gives a shit about 7nm or 10nm. My comments are only applicable to datacenter customers. Desktop/Client chips are a whole another enchilada where marketing plays a bigger role (have you seen the ridiculous packaging from AMD & Intel? This is to please the RGB Gamer crowd).
- Dylan16807 8y agoThe feature size is relevant for understanding how the chips progress over time. It's not relevant for a point-in-time purchase, but it's not a marketing thing to ignore either. The plus means a generation of optimization on how to use roughly the same lithography tech, which can give you a big difference when they're so hard to use.
- akvadrako 8y agoIt really doesn't matter. What if going from 10nm to 1nm means you need to use simpler features, reducing performance? CPUs won't necessarily evolve toward smaller feature sizes - it's just what we've seen so far.
- Dylan16807 8y agoYou can always use your 1nm tech to craft more precise circuits of the same size, packing them better. When it comes to the working chips that are made, assuming some tighter lithography that's equally functional to the previous one, it's only going to benefit.
- makomk 8y agoAs I understand it, Intel's 14nm+ and ++ process revisions actually increased the effective feature size compared to their previous 14nm processes. As you say, transistor density isn't everything.
- ac29 8y agoThese links might be useful: https://en.wikichip.org/wiki/10_nm_lithography_process https://en.wikichip.org/wiki/10_nm_lithography_process https://en.wikichip.org/wiki/7_nm_lithography_process https://en.wikichip.org/wiki/7_nm_lithography_process
- xioxox 8y agoThis page [1] puts the 10nm Intel density as 100.8 million transistors per mm^2. Your number is 40.1 million/mm^2. [1] https://hexus.net/tech/news/cpu/119699-intel-10nm-density-27x-improved-14nm-node/ https://hexus.net/tech/news/cpu/119699-intel-10nm-density-27...