Worldren

111 posts

Worldren

Worldren

@WorldrenChang

Private investor focusing on AI, Space, Digital currency

Katılım Kasım 2023
265 Takip Edilen15 Takipçiler
Worldren
Worldren@WorldrenChang·
@jukan05 Is it a huge risk if US further tighten the purchasing from China
English
1
0
1
1.6K
Jukan
Jukan@jukan05·
To those shorting ASML, you might want to be careful. TSMC earnings are tomorrow, right? We already know what Uncle C.C. Wei is going to say there. $ASML
English
25
25
576
74K
Worldren
Worldren@WorldrenChang·
@pbdes @RocketLab 200 is not a surprising number and not enough to serve constellation. But it is a good start and need to watch its potential
English
0
0
0
27
Worldren
Worldren@WorldrenChang·
@jukan05 Wow, what is Samsung thinking now?
English
0
0
1
696
Jukan
Jukan@jukan05·
What the hell is even going on here?
Intel@intel

Intel is proud to join the Terafab project with @SpaceX, @xAI, and @Tesla to help refactor silicon fab technology. Our ability to design, fabricate, and package ultra-high-performance chips at scale will help accelerate Terafab’s aim to produce 1 TW/year of compute to power future advances in AI and robotics. It was fun hosting @elonmusk at Intel this past weekend!

English
115
42
1.1K
235.5K
Worldren
Worldren@WorldrenChang·
@firstadopter Based on the committed new production launch schedule , the delay is surely not a surprise in semi industry. No delay is a surprise indeed. But it does not matter in such tight supply and hungry demand situation. Selling more rubin or more blackwell no big differences.
English
0
0
0
396
tae kim
tae kim@firstadopter·
KeyBanc on Nvidia. Doesn't seem to be a big deal compared to the worried SemiAnalysis note two weeks ago $NVDA : "NVDA supply of CoWoS in 2026 remains unchanged at 650K, while secured supply in 2027 is at 840K, while Rubin ramp is slightly delayed due to HBM4 qualification issues.” “We estimate that this CoWoS supply in 2026 should support 5.5M-6M Blackwell GPUs, 1.5M Rubin, and 1M Hopper GPUs.” “The ramp of NVDA’s Rubin GPU has been delayed due to issues related to the qualification of HBM4. We believe NVDA is targeting production of 1.5M Rubin GPUs, down from 2M originally as a result of this delay” My general view on slight product delays and hiccups is that they are not a big deal as long as demand remains strong and competitive positioning doesn't change. It is expected frankly, when you are running at the Speed of Light with an aggressive annual cadence. Who cares if the revenue for a next-gen product comes few months later? What matters is long-term earnings power and moat durability.
English
4
5
69
13.6K
Worldren
Worldren@WorldrenChang·
@jukan05 If US ban all China purchasing from ASML, which is under drafting and China sales takes a lot in ASML revenue, the upside can fill the gap?
English
0
0
0
871
Jukan
Jukan@jukan05·
I've been calling ASML since October last year, predicting that DRAM would face a severe supply shortage, and that as the three major DRAM makers ramp up capacity expansions, ASML's EUV tools would become the subject of an intense scramble. My thesis is gradually playing out. ASML's EUV slots are already fully sold out through 2027, and negotiations for 2028 allocations are now underway. Most recently, Samsung alone ordered 20 EUV systems for a single fab. ASML is still cheap.
Jukan tweet media
Jukan@jukan05

Samsung Electronics Orders ~20 EUV Tools for P5… "First Cleanroom Completion Next Year" Samsung Electronics has reportedly placed orders for approximately 20 extreme ultraviolet (EUV) lithography tools—critical equipment for sub-10nm advanced processes—with Dutch semiconductor equipment maker ASML. Including deep ultraviolet (DUV) tools, the total lithography equipment order reaches roughly 70 units. Samsung plans to leverage this overwhelming number of lithography tools to maintain a decisive lead over competitors such as SK Hynix and Micron in advanced process technology. According to multiple semiconductor industry sources on April 6, Samsung Electronics has issued purchase orders (POs) to ASML and Japan's Canon for approximately 70 lithography tools to be installed in Phase 1 of its Pyeongtaek Campus Fab 5 (P5). Notably, the ~20 EUV lithography tools alone are valued at over KRW 10 trillion. These tools will be deployed to ramp production capacity on Samsung's 1c node—its 6th-generation 10nm-class DRAM process. As 1c process productivity improves, output of 6th-generation High Bandwidth Memory (HBM4) built on this node will also increase. Samsung mass-produced HBM4 in February this year—a world first—and has been supplying it to NVIDIA, the world's largest AI semiconductor company. HBM4 is mounted on NVIDIA's latest high-performance GPU, Rubin. Rubin is expected to begin shipping in earnest in the second half of this year, with supply going to U.S. big tech companies including Google and Amazon. The industry projects Rubin will generate over $1 trillion in revenue. Samsung is scaling up HBM4 production at its Hwaseong H3 Line 17 and Pyeongtaek P3/P4 fabs in line with the Rubin launch. Once the newly ordered EUV tools are delivered sequentially from ASML, Samsung is expected to simultaneously expand both DRAM and HBM4 output, solidifying its dominant position in the memory semiconductor market. The lithography tools from this order are scheduled for delivery in time for the P5 Phase 1 cleanroom build-out, expected in Q1 next year. Given that semiconductor equipment shipping typically takes about one year, Samsung is projected to begin installing the EUV and other lithography tools in the P5 cleanroom by Q2. Accordingly, Samsung's 1c DRAM and HBM4 production capacity is highly likely to see a significant increase in 1H next year. Industry observers note that this large-scale lithography equipment order marks the beginning of Samsung widening the technology gap over competitors in advanced process technology. SK Hynix, which is in fierce competition in the HBM market, signed an EUV supply contract with ASML worth approximately KRW 12 trillion for around 20 units late last month. SK Hynix plans to bring its total EUV fleet to roughly 40 units to strengthen its competitiveness in advanced processes. However, with Samsung ordering ~20 additional EUV tools, the equipment gap between the two companies in advanced processes is likely to persist. Samsung currently operates approximately 40 EUV tools—roughly double SK Hynix's fleet. With the addition of 70 lithography tools including EUV, Samsung can continue to lead in the race for advanced process supremacy against SK Hynix. Furthermore, analysts believe this positions Samsung favorably in the development race for the 1d node—the 7th-generation 10nm-class DRAM expected to be adopted starting with HBM5E. Samsung plans to deploy approximately 20 EUV tools at Pyeongtaek P5. If all four phases of P5 are configured as DRAM production lines, Samsung could produce more than double the volume of SK Hynix. An industry source familiar with the matter explained: "In the past, it was standard practice to determine NAND flash lines first when building a new fab, but this time the decision was made to expand DRAM lines first due to the expected increase in HBM4 shipments. EUV tools are expected to be installed sequentially starting in Q2 next year." $ASML

English
23
57
968
178.4K
墓碑科技
墓碑科技@mubeitech·
华盛顿街头黑压压全是人。 主流媒体却集体装瞎了。 这是一场支持川普的大游行。 人群中站着大量伊朗人。 他们高喊着一句话。 “谢谢您,川普总统。” 这么大的阵仗,新闻台硬是一秒都没播。 他们的转播车全开去了另一边。 去拍那场叫“No Kings”的抗议秀。 拿钱办事举牌子的群演,享受着全天候的滚动直播。 自发上街的民意,直接被物理抹除。 媒体手里的静音键,按得比谁都准。
中文
72
740
2.5K
83.3K
Worldren
Worldren@WorldrenChang·
@damnang2 Don’t have to persuade everyone. If you know you know
English
0
0
0
198
Damnang2
Damnang2@damnang2·
I’m honestly exhausted by the endless stream of people asking whether TurboQuant will reduce memory demand going forward and whether HBM is no longer needed. At this point, just believe whatever you want.
English
8
1
81
9.1K
Worldren
Worldren@WorldrenChang·
@AlecStapp thans to AI. if you do fact check with ChatGPT you will find the example over-simplify the loss. Tons of reasons behind it.
English
0
0
0
196
Alec Stapp
Alec Stapp@AlecStapp·
Reminder that the United States could have been the world leader in 5G technology instead of China if we had just given *one guy* a green card when he needed one.
Alec Stapp tweet mediaAlec Stapp tweet media
English
204
1.2K
7.7K
412.9K
Worldren
Worldren@WorldrenChang·
@firstadopter it shows more memories can be used to improve model performance and reduce usage of memory of per token.
English
0
0
0
93
Worldren
Worldren@WorldrenChang·
@damnang2 not reasonable in item 4. should be double-checked
English
1
0
0
285
Damnang2
Damnang2@damnang2·
서브스택에 memory 산업 전반에 대한 인사이트를 얻을 수 있는 아티클을 작성했습니다. 오늘 참석했던 SK hynix의 세미나 내용에 제 인사이트를 곁들인 내용이고, 장담하건데 어디서도 들을 수 없는 내용이라고 확신합니다. open.substack.com/pub/damnang2/p… 혹시 아직도 메모리가 cycle 산업이다, 지금이 메모리 슈퍼사이클이다 라고 떠드는 사람이 있다면 반드시 이 아티클을 읽게 해주세요. 감사합니다. Table of Contents 1. This Is Not a Memory Super Cycle 2. The Three Requirements to Buy HBM 3. cHBM: Memory Becomes Foundry Like 4. What It Takes to Build a 1GW Data Center 5. Is Hybrid Bonding Mandatory for HBM? 6. 3D DRAM and the China vs. Korea Dynamic 7. The Limits of PIM and Google 8. SK Hynix ADR, and Micron
Damnang2 tweet media
한국어
5
17
88
11.4K
Worldren
Worldren@WorldrenChang·
@rklb_invest About all the statements from Iran officially it is very hard to know these statements are shown to their own people for saving their faces and uniting their people or they really mean that.
English
0
0
1
256
Jordy Beuving
Jordy Beuving@BeuvingJordy·
So guys, what's the next big thing after opto/photonics ??? Because investors are impatient... And after witnessing the rotation GPU > Memory > Semi Equip > Opto I'd rather front-run the next trend a bit early 😂
GIF
English
10
0
15
2.1K
Worldren
Worldren@WorldrenChang·
@best181p What’s th proof of success of Tesla 4680 dry electrode in terms of yield and cost
English
0
0
0
21
JH의메모
JH의메모@best181p·
I've recently been having some concerns. Now that Tesla's 4680 dry electrode process has succeeded, they will gradually start equipping their vehicles with these batteries. And the AI5 chip will also be released soon. In that sense, it feels a bit regrettable/pitty to buy a Tesla Model Y right now… but even so, would it still be better to go ahead and buy one?
English
2
0
2
459
Worldren
Worldren@WorldrenChang·
@rklb_invest It is interesting to know the moat of making solar panel with space requirements. Once the TAM becomes that big, everyone will get in?
English
1
0
2
447
Rocketesla
Rocketesla@rklb_invest·
1 TW(=1,000GW) 의 전력을 우주에서 태양광으로 만들려면 필요한 Starship 발사 횟수 약 1.6만 ~ 5만 회 이상 결국 발사체는 Starship이 독점 될 예정. 하지만, 패널 수요는? $RDW $RKLB
Rocketesla tweet media
한국어
12
12
116
19.6K
심월
심월@mindmoon_108·
@latel_coin 꼭 일론 때문은 아니더라도 반도체 하나 만드는게 만만치 않다는 이야기를 하고 싶었습니다
한국어
1
0
1
37
Worldren
Worldren@WorldrenChang·
@RealChickenBoy9 I don’t think Tesla solved 4680 issues, yield and cost. Where is your information from?
English
0
0
0
40
리얼치킨보이
리얼치킨보이@RealChickenBoy9·
2020년 Battery Day 때 전문가들은 똑같이 말했었습니다. “Dry electrode 공정? 용매 없이 어떻게 균일하게 코팅하나? 100년 넘게 습식 공정이 표준인데 이건 공상과학이다.” “Tabless + 실리콘 애노드? 레이저 용접 불량에 실리콘 팽창(300%) 때문에 사이클 수명 500회도 못 간다.” “5배 에너지? 56% 비용 절감? 숫자만 보고 말하는 거다. 실제 양산 수율 잡는 데 5~10년 걸린다.” CATL 회장까지 공개적으로 “그 기술은 실패할 것”이라고 했죠. 당시 업계 분위기는 “4680은 또 다른 vaporware(증기웨어)다. Cybertruck에 들어간다더니 영원히 프로토타입일 거다”였습니다. 그때는 정말 “Dry electrode? Silicon? Tabless? 절대 안 된다”는 반응이 지배적이었음. 그런데 2026년 지금, 상황이 완전히 달라졌습니다. Cybertruck에 4680이 제대로 들어가고, 일부 Model Y에도 다시 탑재되기 시작했어요. Dry electrode도 안정화됐고, 비용 경쟁력도 상당 수준까지 올라왔습니다. 전문가들은 이제 그때처럼 크게 말하지 않아요. 불가능을 현실로 만들어서 조용해졌습니다.. 그리고 지금 똑같은 일이 테라팹에서 반복되고 있습니다. “전체 클린룸을 Class 1으로 유지해야 한다.” “2nm GAA nanosheet 균일성 제어는 불가능하다.” “수율은 돈으로도 안 된다.” 전문가들이 또 “안 된다”고 말하고 있어요. 일론은 그 말을 들을 때마다 항상 같은 질문을 던집니다. “이게 물리적으로 진짜 불가능한가, 아니면 그냥 ‘항상 그래왔으니까’ 불가능한가?” 그 질문의 대표적인 예가 바로 “팹 안에서 시가 피우겠다” 발언입니다. 업계는 폭소했죠. ISO Class 1 클린룸에서 시가 한 대 피우면 입자 수십억 개, EUV 미러가 죽는다는 건 사실입니다. 하지만 일론이 진짜 물은 건 다른 거예요. “왜 건물 전체를 ISO Class 1으로 유지해야 하는가?” 웨이퍼를 질소 퍼지 + 완전 밀폐된 micro-environment pod에 넣고, 로딩 순간만 초미세 localized vacuum chamber로 처리하면 어떨까? 그러면 전체 팹 공기량을 70~80% 줄이고, HVAC 비용과 에너지 소비를 30~50% 줄일 수 있다. TSMC·삼성도 이미 2010년대부터 비슷한 방향으로 조금씩 바꿔왔어요. 일론은 그걸 한 단계 더 밀어붙이는 거죠. 장비 병목도 마찬가지예요. ASML High-NA EUV는 리드타임이 2년이라 기다려야 하지만, 그래서 테라팹은 패키징부터 시작합니다. SpaceX Texas FOPLP 라인에 이미 2025년 9월 장비가 들어갔고, 2026년 3분기부터 소량 생산 들어갑니다. CoWoS 병목을 먼저 푸는 거예요. 2nm GAA nanosheet도 마찬가지입니다. 전문가들은 “양자 구속 효과 때문에 두께 변동 0.5nm만 생겨도 Vth가 크게 흔들린다”고 하지만, 삼성 Taylor에서 테슬라 엔지니어들이 ALD 공정을 실시간으로 제어하고 있고, 변동을 상당히 줄이고 있는 것으로 알려져 있습니다. 수율도요. 내부 수요(FSD, Optimus, xAI)가 워낙 크니까 low-yield wafer도 바로 테스트하면서 데이터가 폭발적으로 쌓입니다. 4680 때처럼 학습 루프를 훨씬 빠르게 돌릴 수 있습니다. PDK는 삼성 걸 실전으로 배우고 있고, Intel과도 협상 중입니다. 로직과 메모리도 처음부터 한 지붕에 다 넣을 생각은 없어요. 초기에는 로직 fab + advanced packaging 통합 형태로 가면서 점진적으로 진행할 계획. 결국 테라팹의 핵심은 “100% 자체 생산”이 아니라 “TSMC·삼성에 인질 안 잡히겠다”는 옵션을 만드는 거예요. 일론은 늘 그랬습니다. “항상 그래왔으니까 불가능하다”는 가정을 깨는 사람. 그리고 과거와는 다른게 현재는 AI가 있다는 것. 불가능을 가능으로 만드는 속도가 더욱 빨라 질 수 있다는 것.
한국어
22
69
276
37.1K
Worldren
Worldren@WorldrenChang·
@zhongwen2005 No way. Most guys like TV, refrigerator and sofa.
English
0
0
0
282
大趙
大趙@zhongwen2005·
Beijing to Release 80,000 Additional Purchase Quotas for New Energy Passenger Vehicles in 2026; Tesla Emerges as the Top Choice for Buyers Report by a journalist
English
7
10
83
10.5K
Aaron
Aaron@Arronwei3n·
From my recent checks, it appears $NVDA has secured ALL of $TSM's A16 capacity (with Super Power Rail) for the Feynman GPU. As a result, we are likely to see two distinct versions of the Feynman architecture: one leveraging the A16 node, and another built on the N2P node to balance performance with volume availability.
Aaron tweet media
English
3
14
146
64K
Worldren retweetledi
Worldren
Worldren@WorldrenChang·
@Arronwei3n key and biggest beneficiary while external components built within packaging
English
0
1
1
73