1 00:00:00,000 --> 00:00:04,160 Intel has been a Fortune 500 company for as long as I've been alive, 2 00:00:04,160 --> 00:00:09,440 consistently controlled well over 80% of the PCCPU market for most of the 2010s, 3 00:00:09,440 --> 00:00:13,760 and hit a 20-year stock price high only about three years ago. 4 00:00:13,760 --> 00:00:18,160 But this month, the company lost its spot on the Dow Jones industrial average, 5 00:00:18,160 --> 00:00:22,640 and headlines say they may potentially be bought by another company. 6 00:00:22,640 --> 00:00:25,680 But how bad could things really be for Team Blue? 7 00:00:25,680 --> 00:00:29,360 Did those shenanigans, with their 13th and 14th gen chips being unstable, 8 00:00:29,360 --> 00:00:38,720 really have that huge of an impact on the company? To be clear, as well publicized as those instability issues were within the gaming PC community, 9 00:00:38,720 --> 00:00:46,240 they didn't hurt Intel very much financially. The affected chips represented a relatively small number of higher-end models 10 00:00:46,240 --> 00:00:51,360 for a company that sells mostly mid-range chips to computer manufacturers like the Dells 11 00:00:51,360 --> 00:00:58,800 and Lenovo's of the world. No, to understand how Intel sank into this position, we have to go all the way back to 2015, 12 00:00:58,800 --> 00:01:04,480 when Intel's struggles with making chips on the 10nm process started hitting the tech news. 13 00:01:04,480 --> 00:01:08,720 Hitting 10nm would have meant more powerful and more energy-efficient chips, 14 00:01:08,720 --> 00:01:12,320 which is something Intel obviously needed eventually. 15 00:01:12,320 --> 00:01:17,040 But at that point, they weren't investing in expensive EUV manufacturing technology, 16 00:01:17,040 --> 00:01:20,560 which makes it easier to create chips with smaller transistors. 17 00:01:20,640 --> 00:01:25,120 The powers that be at Intel already took a generally conservative approach 18 00:01:25,120 --> 00:01:28,320 with any significant changes to their manufacturing processes, 19 00:01:28,320 --> 00:01:32,080 but another big part of the reason the EUV investment didn't happen 20 00:01:32,080 --> 00:01:36,480 is because Team Blue just didn't have pressure from competitors back then. 21 00:01:36,480 --> 00:01:40,960 AMD was still struggling with its anemic bulldozer family of CPUs, 22 00:01:40,960 --> 00:01:44,640 and Apple was still buying processors from Intel to use in its Macs. 23 00:01:44,640 --> 00:01:49,280 But not long after 2015, the landscape really started to shift. 24 00:01:49,280 --> 00:01:54,640 AMD's new Zen architecture hit the scene in 2017 and was an almost immediate hit, 25 00:01:54,640 --> 00:02:00,320 and that same year, TSMC, a massive contract manufacturer that now makes Ryzen chips, 26 00:02:00,320 --> 00:02:04,240 started using EUV to make products with smaller transistors. 27 00:02:04,240 --> 00:02:11,200 Then Apple broke up with Intel in 2020 in order to use its own Apple Silicon processors in its Macs. 28 00:02:11,200 --> 00:02:15,920 All these developments meant that Intel was suddenly significantly behind its competitors 29 00:02:15,920 --> 00:02:19,360 in terms of their ability to crank out leading edge chips. 30 00:02:19,360 --> 00:02:23,040 Team Blue finally got its act together somewhat in 2021 31 00:02:23,040 --> 00:02:27,680 when they brought back former executive Pat Gelsinger to be their new CEO. 32 00:02:27,680 --> 00:02:32,880 Gelsinger wanted to embrace EUV, following infighting among Intel board members 33 00:02:32,880 --> 00:02:40,320 over how to modernize its chip strategy. Intel started releasing chips on what they called their Intel 7 process, 34 00:02:40,320 --> 00:02:44,560 which was really a fixed 10nm process that same year, 35 00:02:44,560 --> 00:02:48,560 and started mass producing chips using EUV in 2023. 36 00:02:48,560 --> 00:02:54,000 But the factors leading to Intel's struggles go beyond merely being behind their competitors 37 00:02:54,000 --> 00:03:02,400 in terms of manufacturing. Unlike competitors like AMD and Apple, Intel owns its own fabs for manufacturing its chips, 38 00:03:02,400 --> 00:03:06,960 and these fabs are very expensive to build, run, and maintain. 39 00:03:06,960 --> 00:03:10,960 In order to be profitable, they need to be running at close to full capacity. 40 00:03:10,960 --> 00:03:14,400 However, during this Intel struggle period we mentioned, 41 00:03:14,400 --> 00:03:17,040 some of their fabs were only running at around 70%. 42 00:03:17,920 --> 00:03:21,520 Therefore, Intel was still pouring about the same amount of money into these fabs 43 00:03:21,520 --> 00:03:24,560 that they would have been had they been running at full capacity, 44 00:03:24,560 --> 00:03:27,760 but far fewer chips were coming out of them to sell. 45 00:03:27,760 --> 00:03:32,320 So does this mean that Intel is losing so much money from keeping its fabs open 46 00:03:32,320 --> 00:03:36,800 that the company will be sold off to the highest bidder? Well, not quite. 47 00:03:36,880 --> 00:03:40,880 The situation isn't so dire that Intel is on the verge of bankruptcy, 48 00:03:40,880 --> 00:03:44,880 and Team Blue does have a plan to get their fabs running at full capacity. 49 00:03:44,880 --> 00:03:49,440 One big piece of this is that Intel is going to perform contract manufacturing 50 00:03:49,440 --> 00:03:57,040 for other chip designers at its fabs. Historically, this wasn't possible because Intel's factories used proprietary chip design 51 00:03:57,040 --> 00:04:00,880 software that no one outside of Intel really knew how to use. 52 00:04:00,880 --> 00:04:05,680 However, Intel is moving to industry standard tools for both its existing fabs 53 00:04:05,680 --> 00:04:11,440 and the new fabs it's currently building, which are on track to start producing chips in 2027. 54 00:04:11,440 --> 00:04:16,000 Intel may then try to get outside investors to own a part of its fab business 55 00:04:16,000 --> 00:04:22,640 and operate it like a subsidiary, but the company would have to be in very dire straits for a total spin-off to happen, 56 00:04:22,640 --> 00:04:25,920 like what AMD did back in 2009 with their fabs. 57 00:04:25,920 --> 00:04:30,400 AMD's former fab business is now a separate company called Global Foundries. 58 00:04:30,400 --> 00:04:34,320 Indeed, it's more likely that Intel would sell off parts of the business 59 00:04:34,320 --> 00:04:41,600 rather than the whole enchilada. This is already happening in part as ASUS has taken over Intel's NUC business, 60 00:04:41,600 --> 00:04:48,240 Apple got Intel's smartphone modem technology, and Intel is currently trying to sell a minority stake in Altera, 61 00:04:48,240 --> 00:04:53,040 which is a line of FPGAs. Another Intel department that might be ripe for the taking 62 00:04:53,040 --> 00:05:00,320 is their data center AI business, which is lagging far behind NVIDIA and AMD since Intel was late to the party. 63 00:05:00,400 --> 00:05:03,440 And outside of Intel's plans to write itself, 64 00:05:03,440 --> 00:05:07,120 there just aren't a lot of companies that would have the desire and ability 65 00:05:07,120 --> 00:05:12,960 to buy all of Intel outright. Apple, who has been mentioned as a potential buyer in the press, 66 00:05:12,960 --> 00:05:17,600 has already committed to the ARM architecture rather than Intel's x86, 67 00:05:17,600 --> 00:05:22,560 and they already have a good relationship with TSMC, who manufactures Apple Silicon. 68 00:05:22,560 --> 00:05:28,480 Meanwhile, Samsung's name has also been thrown around, but they're having issues of their own with manufacturing 69 00:05:28,480 --> 00:05:34,000 and would probably face regulatory hurdles, as a foreign company trying to buy a US-based business 70 00:05:34,000 --> 00:05:37,360 that the American government views as critical to its economy, 71 00:05:37,360 --> 00:05:40,560 hence its recent funding of Intel through the CHIPS Act. 72 00:05:40,560 --> 00:05:44,640 So I wouldn't expect Intel to become a subsidiary of some other tech company, 73 00:05:45,280 --> 00:05:48,320 but who knows, maybe one day the Walton family will get bored 74 00:05:48,320 --> 00:05:52,400 and suddenly will have great value brand CPUs on the shelves at Walmart. 75 00:05:52,400 --> 00:05:57,200 Hey, thanks for watching. If you liked this video, check out our other video on why Intel and AMD 76 00:05:57,200 --> 00:06:00,240 are the only two companies making x86 CPUs.