A new Intel roadmap has leaked, predicting a start for a 10nm Intel “Alder Lake” as before long as late 2020. Normally, when I publish up a rumor, it’s because I’ve by now believed it about and made the decision it has a good prospect of getting real. In this scenario, I’m executing some thing a small distinct. I want to chat about this rumor in conjunction with various other individuals that have surfaced about the final 6 months.
The new rumor, in accordance to THG, surfaced on Sina Weibo and is made up of two slides:
Here’s how Tom’s Components describes the contents of the to start with slide: “the slide statements improved yields, significant boost in generation ability and a collection of 10nm products and solutions launching in 2020. As for the 7nm lineup, the leak states the main merchandise will land in 2021, though the comprehensive merchandise portfolio can be envisioned in 2022. The data also suggests overall performance enhancements just about every year.” For comparison, here’s the first variation of that slide that Intel launched to the public:
The THG description of this slide could indicate that it has been up to date considering the fact that we noticed it back in 2019, which is absolutely some thing organizations occasionally do — or that anyone started from a known Intel slide to give their creation a veneer of authenticity and then made wrong information to overwrite what had formerly been offered.
Other Modern Rumors
In the past 6 months, key Intel roadmap rumors have bundled the pursuing:
- Intel has canceled all 10nm desktop CPUs. Desktop continues to be on 14nm via 2021.
- Intel will introduce a new 14nm desktop CPU architecture named Rocket Lake in late 2020, with backported 10nm attributes.
- Intel will start 10nm desktop Alder Lake in an 8+8 significant.Minimal configuration as an forthcoming comply with-up to LGA 1200 CPUs. Implied 2021 start day.
Now, let’s insert in a couple of things that we in fact can say for specific. We know that Intel has a comply with-up architecture to Ice Lake, dubbed Tiger Lake, coming on 10nm. We know that immediately after Ice Lake’s Sunny Cove arrives Willow Cove (assumed for Tiger Lake) and then Golden Cove. Alder Lake could be based mostly on possibly Willow or Golden in theory and use reduced-electrical power Atom cores (Tremont or Gracemont) for the “little” chips.
If the to start with circumstance is real, the other two naturally aren’t. If the second circumstance (Intel backporting 10nm attributes) is real, the 3rd most likely isn’t. Why would Intel start Rocket Lake immediately after using the monumental issues to backport its capabilities, only to transform all over and start a “true” 10nm desktop CPU at specifically the same time? It wouldn’t.
Intel explained to the press at its 2018 Engineering Day that it had by now backported specific AVX-512 attributes to 14nm, but it under no circumstances stated this was the only time it would ever do so. It also has stated nothing at all about 10nm desktop CPUs — not final drop, not this spring. When you look at Intel’s presentations, the conversations of 10nm that mention platforms chat about cellular and server. Does this imply Intel isn’t launching 10nm desktop chips? No. But it usually means Intel isn’t talking about launching desktop 10nm chips, at the very least not still, and it ordinarily would be by now. And still, we also know Intel not too long ago canceled most of Cooper Lake, which indicates the organization is seeing improved 10nm CPU yields than envisioned — specifically as this slide reportedly states.
Following Broadwell, Intel stated it wouldn’t skip a system with a node yet again, but that was various CEOs back and a distinct set of situations. When you look at these a few rumors, I’d argue we have the most help for #1 and #2. These rumors also do not contradict each and every other. This 3rd rumor refers to Intel’s Snow Ridge 5G base station on the to start with slide, a system which is by now launched, and it makes some significant guarantees all over foreseeable future Intel products and solutions. It’s feasible that the slide is genuine, but was made at an before place in time when Intel had a distinct concept of what its roadmap would look like. At times, the purpose rumors show up to be wrong is because they were being real when the leaks happened, but were being later on canceled.
Some of you may possibly try to remember rumors that AMD would build a tremendous-APU with a substantial-clocked Piledriver-derived CPU and a wide desktop GPU for graphics, all linked to a unified pool of HBM, in a socket capable of withstanding the electrical power draw. Sources within AMD confirmed that this was a serious merchandise the organization was looking at building, but ultimately abandoned. The rumors about Tarnhelm weren’t wrong they just did not arrive real. This rumor could unquestionably however be real, but offered that it contradicts before rumors, we know that some of the data in circulation about Intel’s forthcoming ideas is incorrect. Salt accordingly.