Intel reportedly dismantles Jim Keller's revolutionary Royal Core project and cancels Beast Lake

[Redaktion]

Intel has reportedly canceled the upcoming Beast Lake architecture according to a new leak. According to past reports, Beast Lake had a considerable focus on single-core performance. In addition to canceling Beast Lake, the report from Moore's Law Is Dead also claims that we may not see the entirety of the Royal Core project materialize in one architecture.

https://www.notebookcheck.net/Intel-reportedly-dismantles-Jim-Keller-s-revolutionary-Royal-Core-project-and-cancels-Beast-Lake.871208.0.html

[Real NikoB (old nm block)]

Intel finally gave in, started making chips at TSMC, which means it got the same or even better conditions in terms of process technology as AMD. Nothing now prevents it from competing "fairly" with AMD (and probably with Apple, since Intel's market share is 5 times greater than AMD's market share and is comparable in pieces to Apple's chip production, which means it can get much more favorable conditions at TSMC than AMD and even Qualcomm, which is actually on par with Apple in terms of contract profitability), even if the cost of batches has sharply increased at TSMC. At the very least, to show something more energy-efficient in terms of performance per 1W, in order to at least return some abstract "reputation" for its development team, as the best team with equal process technology with competitors.

We will see this soon..

Even in this regard, improving its reputation is now important for Intel.

[usacomputer]

In short, Intel has been on the defensive because it had a MONOPOLY for a few generations, neglecting power specs to get the coveted top performance spots in revisions. This showed that AMD's microarchitecture was much more polished than Intel's which doctored its processors with so much power that over 20% of them burned out. AMD did this to a lesser extent with Ryzen 7000; nearly doubling the stock power consumption of the 7950X gave them enough extra percentage points in performance to score wins. And now it's being proven and we've seen that if you cap AMD and Intel's TDP to a more reasonable power limit like 125W, AMD barely loses any performance while Intel takes a considerable hit of over 50% performance loss, further proof of AMD's microarchitectural superiority. The most serious problem is that there are 2 generations of Intel Core13 and Core14 and INTEL KEPT QUIET and now that Intel is forced to lower voltages we see that INTEL'S PREVIOUS VICTORIES were cheating since they took the silicon to the maximum ignoring that they burned out after a year.
Now the danger is that this happens with LunarLake and Intel Core15 since Intel still doesn't know if it's a voltage, oxidation or microcode issue, what is certain is that here in the US they are suing Intel for bad practice and that in the US the complaints for factory faults are paid by the manufacturer and the worst of all is that people tired of so much scam are opting for AMD.

[Ping K Meep]

Honestly Intel might even miss out on some or all CHIPS Act funding due to layoffs. Still loosing market share in the most profitable segments. In less profitable segments they just extended warranty periods by another two years due to ongoing defects. Dramatic drops in PR and advertising spends when they arguably need to spend more. Stock price went from $51 to $19 in less than a year. I would like to say maybe it might be worth the risk soon but Intel also banked on fab capacity for other companies. And now that they have TSMC they have to meet schedules or pay for capacity that they don't use. What is the worst that could happen? WW3 AND stagflation into brutal recession. Can Intel survive?

[ChipDip]

Intel finally gave in, started making chips at TSMC, which means it got the same or even better conditions in terms of process technology as AMD. Nothing now prevents it from competing "fairly" with AMD (and probably with Apple, since Intel's market share is 5 times greater than AMD's market share and is comparable in pieces to Apple's chip production, which means it can get much more favorable conditions at TSMC than AMD and even Qualcomm, which is actually on par with Apple in terms of contract profitability), even if the cost of batches has sharply increased at TSMC. At the very least, to show something more energy-efficient in terms of performance per 1W, in order to at least return some abstract "reputation" for its development team, as the best team with equal process technology with competitors.

We will see this soon..

Even in this regard, improving its reputation is now important for Intel.

Exactly. This is actually the only comment this far that has actual facts instead of doom and gloom antics. Everything at Intel is in a state of change. Our government (US) has invested to much into Intel for it to actually fail. They are building a massive factory have Switched to TSMC so I have absolute faith they will pull through.

Thier cores I believe are superior and with the technology referred to in this article plays out we should see some interesting results. Yes Intel fell victim to success just like every other corpo tart. They get fat and lazy and ride off reputation till it explodes in their face.

Now they just wipe the pie from their eyes and call it a day b

[Spencer]

Looking forward to Intel's Jumpintha Lake series.

[cfb]

Does everyone remember MLID telling all of us that intel was closing up its gpu division entirely soon after Alchemist launched? And then went on f-bomb laden tirades about how he was sure that he was correct?

I sure do, and haven't listened to a "leak" from him since.

[ArsLoginName]

Does everyone remember MLID telling all of us that intel was closing up its gpu division entirely soon after Alchemist launched?

Intel can't 'close up its GPU division' after Alchemist due to needing the core of those GPUs to make better iGPUs. But it does seem like Intel will not launch 6 variants of Battlemage or Celestial like they did with Alchemist (A310, A380, A580, A750, A770 8 GB, A770 16 GB). But what is the real difference between closing up completely and reducing to only 1 or 2 dedicated models? Not much since it probably involved reducing staff counts by a large %. Predicted to be only 2 Battlemage aimed at the entry to mid level not 6 variants covering the entire range.

[BetterDecisionsTel]

Did Intel's dip drop a rare tech that can be outlicensed? Surely anything else would be anticompetitive?

Looking forward to having many semiconductor competitors, and many labor unions of applicable workers and their orgs, certainly there can be many sectors started now that only robots can fill and improve labor adequately.

[Prodro]

They all make it seem like it is some kind of Intel's architecture for not being "efficient" enough. As a result of this I am convinced I live in the twilight zone. If Intel gets in its hands the nodes Apple has been working with, it will produce about the same efficiency and performance chips. This goes without saying. If incompetent Apple managed to do so, it will be a walk in the park for Intel.

A note on performance per watt. Some people just divide gigaflops per watt and make a conclusion about efficiency.  But lack of common sense and basic education makes them reach at completely wrong conclusions. Letting aside that they disregard lithography altogether, they have also little understanding how performance per watt scales from low power to high power systems. So , they take, on same lithography[sometimes different ones which is hilarious] , 2 chips one at 5 watts and one at 150 watts. When they calculate the low power system they see for example it is 10 gigaflops per watt. So , they expect ,like the ignorant they are, the high power system to output 150 x10 =1500 gigaflops. When they see it produces "only" 150 gigaflops, they freak out:"yes it is more powerful but inefficient" ROFL. So if an 100 hp car does zero to 100 km/sec in  10 seconds, according  to them, then a car with 1500hp it should do it in  0.6 secs . Really?

[Henry Molter]

I think intel has lost alot of trust for people who closely follow the tech world. The people who dont follow the tech will unfortunately most likely not know about the microcode patch and fail to update in time to save their current  chips. I think it will make alot of non "tech" people more hesistant to go with intel. I hope they are able to fix the issues and boost performance in the upcoming chips but i think alot of people effected will be going amd till they rebuild the trust they lost from the way they handled the issue. I will be reluctant to go back to intel untill i have seen a couple generations come out without any problems. They have there work cut out.

[Yarrow Yul]

Intel gets massive property tax breaks in Oregon, they probably can withstand this storm without much trouble.