He did it! 9:08 He pronounced it “D-ram,” instead of “dram.” Thank you so much, even if it was just this once 😅.
Man, this video brings back memories! I worked for a US computer OEM in the early 90s when this exact transition was happening. We used to source almost exclusively Japanese DRAM for our systems, and I remember how our purchasing department would practically grovel to secure allocations from NEC and Hitachi. Japanese memory had this reputation for being absolutely bulletproof reliable compared to alternatives. Around '92-'93, our supply chain director started pushing us to qualify Samsung chips, and there was MASSIVE resistance internally. Our engineers literally laughed at the suggestion! We had this perception that Korean semiconductors would be inferior - boy were we wrong. I'll never forget running the comparison tests myself and seeing that Samsung's 4MB modules were not only matching Japanese quality metrics but their production consistency was actually better. What the video doesn't fully capture is how FAST the power shift happened on the ground. When we first approached Japanese suppliers about pricing pressure from Samsung, they basically shrugged it off as impossible competition. By 1995, those same companies were desperately offering us price concessions to keep our business. The arrogance transformed into panic practically overnight. The craziest part was watching Japanese companies struggle with decision-making exactly as described. We'd request spec modifications or custom configurations, and Samsung would give us an answer in days while Japanese vendors needed weeks of meetings. Their vertical integration became a liability when they couldn't pivot quickly.
Japanese firms in the 1980s had the advantage of being able to build new fabs in less time than US fabs could get permits approved. And Japanese firms were maniacal in their focus on maximizing yields … much more than US firms were.
I'd like to recommend the 4 part documentary 'A Video History of Japan's Electronic Industry'. It's on youtube and though old (and not in such a good quality), it goes into a lot of detail.
Here so early Japan was still "good" at semiconductors
I'm in awe at how easy it was to follow along with you on this topic. Thanks again for making these videos.
Sunday evenings are Asian themed nights. I usually make some noodles and some veggies. I don’t know a damn thing about Asian food but I give it a good shot with my homegrown veggies and cheap ass noodles. Thank you, Mr Deer Head Man. You make me just a tad bit smarter and I’m thankful for it. It’s hard to get content and conversations like this in Mississippi. 😂
A related story that would be interesting to tell is how the Japanese invited an American, W. Edwards Deming, to help them improve their quality control. Having initially been deaf to Deming's message, the Americans ended up getting the Japanese to teach them how to get quality control right.
What a salient explanation! Thank you! I see some parallels between the semiconductor and automotive industries. For example, I’ve seen first-hand the power of Keiretsu (symbiotic relationships between companies) and also the advantages of lifetime employment. The Toyota engineers were so versed in their respective areas of expertise precisely because of their tenure. In contrast, the American revolving door scheme meant engineers had no “history” or “battle scar” experiences to draw upon when it came to product improvement. Serwus!
Also, do not forget the "power semiconductor" industry. Our company started with GE and Westinghouse, but as time went on, we gravitated toward Mitsubishi, Toshiba, Fuji, and others. They all have their niches.
I think another important and often under talked about "culture piece" has been talked about by Craig Barrett in interviews he did from his time visiting Japan when he was rebuilding Intel's manufacturing organization. During his tours in Japan, what he realized was Japan's "special sauce" was data drive engineering built around the idea of continuous improvement and statistical process control. American semiconductor fabs at the time were by contrast more artisanal craft production. Having things like operators with stop watches determining process time and operators/engineers trying to learn the intuition for technical wizardry if you will. Implementation of SPC, data driven engineering, and as Craig Barrett said just "running your chip factories like a factory rather than a laboratory" drove massive product quality improvements and greatly leveled the field with foreign chipmakers. Having recently read *the machine that changed the world*, I can see alot of similarities between the comparisons of Lean production and Craft production. The Japanese semi makers were no Toyotas, but they were definitely at least Ford-like and employed some elements of Lean manufacturing. Meanwhile, IBM, Intel, MOSTEK, and National were more Benz-like filled with proud "artisans". The problem with craft production is that Lean production always beats out mass production on cost and craft production on quality. I do wonder in an alternate timeline if American chipmakers moved away from craft production sooner if there would have been more American leading edge memory makers today. Micron made the transition and came out the otherside doing pretty well for themselves. TI, Intel, and AMD all made great strides in improved manufacturing discipline and methodologies in the late 80s thru 90s, but began/completed pivots out of commodity memories to design rich products. I feel like they could have replicated Micron's success and in the case of TI and Intel would have presumably been more stable on account of the greater financial firepower/scale. But seeing how things worked out in the late 90s and 2000s I don't think TI, Intel, or AMD are particularly broken up about moving to higher margin differentiated products.
That Plaza Accord seems to be the problem with everything. It seems to be mentioned in every video about Japans economy. If that had not happened I wonder if Japan could have continued.
Thank you again Asianometry for another great video. We appreciate your work Sir.
All of my girfriends talks about the HItachi. i guess japan won
Starting in the late 1980s U.S. companies like Motorola adopted more theoretical and mathematical methods to improve and maintain quality, speed up the development process. At Motorola we developed the “ Six Sigma Methodology”. that has now spread worldwide. In the Chip Packaging area to reduce labor input / cost ( plus get higher performance ) we developed Advanced. Packaging ( Flip Chip tech, which uses area array of microscopic solder bumps on the Chip face, aligned and attached to the pads on substrates using adv. M/c vision based fast yet precision Robots pretty much home made w/ own algorithms & software that have now become industry standards and power AI hardware. Since the early 1990s on behalf of both Motorola & Intel have been to Japan over 25 times. They were NOT very well educated ( not enough PhDs or fluent in English ) so could n’t speed up their development work using theory or other peoples Research.. In contrast to Japan, key people in So Korea & Taiwan all had PhDs and that too from the US. These basic technical factors combined w/ business and social factors made Japan uncompetitive, now they are demoralized looking up at TSMC even for technologies ( licensed from the U.S. ) that SONY has been using since 2016 !
Thanks!
I remember Elipda being so sought after by overclockers back in the day. Their chip were the creme de la creme when it comes to overcloking.
Thx a lot for all the work you put on conservation of history I've learned so much
There was a saying I remember hearing in the 1970’s, “America invents something, Russia claims they invented it first, while Japan is shipping it…” I also was told that Sony never made any products with vacuum tubes; they built transistorized high-tech from day one.
@janzettergren7988