Ry Chang (@ryry_chang) profile picture

📍Taiwan

Taipei City, Taiwan
Joined September 2024
Photo by Ry Chang (@ryry_chang)Photo by Ry Chang (@ryry_chang)
Ry Chang retweeted
Why $AAOI & $SIVE represent the biggest asymmetric opportunity in the AI trade.
I went through over 50 earnings calls so you don't have to, and one thing has become crystal clear: Data centers are barreling straight into a brutal laser shortage. Why $AAOI & $SIVE represent the biggest asymmetric opportunity in the AI trade:
8
13
115
10,943
Ry Chang retweeted
fu*cking every Jane Street resource for traders. free. all in one. hft.md/jane-street save rn.
31
401
7
5,332
588,368
Ry Chang retweeted
800 V直流配電的競爭,正從架構選擇走向設備互通 AI機櫃功率增加,讓供電成為資料中心設計的重要限制。Counterpoint Research在2026年11月5日發布的分析中,將800 V直流視為降低高電流配電負擔的發展方向,但不同業者未必會以相同方式完成轉換。 以1 MW負載為例,54 V供電約需18,500安培,800 V則約需1,250安培。較高的配電電壓有助減少銅材與配電空間,也能降低高電流帶來的損耗。不過,800 V進入機房或機櫃周邊,不代表內部運算設備已全面改用800 V。 對既有設施而言,較務實的方法是增加電源櫃,或集中設置列級電源設備,再向運算機櫃配送直流電。交流輸入與機櫃內54 V配電仍可保留。採固態變壓器直接將中壓交流轉為800 V直流,則是較長期的新建機房方案,原文列示的部署目標為2028年第一季。 架構差異也在演變 NVIDIA( $NVDA)採單極800 V,Mt Diablo原以雙極±400 V為主;後者兩條電源軌之間同樣具有800 V電位差。隨著Mt Diablo擴展支援範圍,以及開放運算計畫(OCP)協作推進,兩者未必會形成長期互不相容的市場。 差別也可能來自營運需求 Counterpoint認為,新型GPU雲端業者較重視快速上線,較可能採取現成參考設計與局部改裝;大型雲端業者則希望供電設備能跨越不同運算世代,同時支援GPU及自研晶片,降低對單一供應商的依賴。 因此,800 V的採用不能只看電壓或單項設備效率。供電路徑效率、轉換器峰值效率與整座資料中心的PUE,是不同指標。最終仍須確認設備互通、安全保護、實際可靠度,以及整體部署成本。規格逐漸靠攏,是形成共同市場的一步,卻還不是全面普及的證明。 counterpointresearch.com/en/…
4
58
6,208
Serenity with a ~$16M port at peak. Did the math: +4,502% in May and currently +594% now, Then there is also the -$13.6M Oct drawdown that was shared. If that's peak to now, then we can determine that Serenity started this journey around ~$350K start before his favorite names blew up $SIVE $AAOI $CCXI etc. So at peak (in May) he had around a ~$16M port And currently he sits around a ~$2.4M now with the 594% YTD. We also know he was in specific names that are generally down about 75-85%. We know the peak his 5900% was May 2025. And the timing of the posts seem to make sense with all the numbers and math. Couple of things: 1. Very impressive to take a port from $350k to $16M. Its like taking a $3.5M port to $160M so very high IQ stuff. 2. The drawdown is something anyone trading has witnessed, endured, experience, so cant blame him, and he is not alone. He is just transparent about it. 3. Hes still positive 600% YTD. 4. Respect for the ability to call these names, and respect for the transparency. 5. This brother has 1M people watching his trades daily. The stress is high, but his consistency, and ability to show up through the drawdown is respectable. 6. He will bounce back 7. Doing hedge fund moves for free on twitter. Wild. @aleabitoreddit , if I'm right'ish can I get a retweet/comment? 🙏
今年收益率仍有 +594.4% YTD,不过巴菲特大概睡得比我安稳。 我那套面向2027–2028年的光电子和存储投资逻辑,看来大家已经等不及了,提前一年就叫我交卷打分。 不过,巴菲特的投资组合规模大得多,他那19%的收益率,我还真比不了。
37
47
6
1,231
331,857
Ry Chang retweeted
NEW: We are short Nvidia $NVDA. We believe Nvidia has a significant China problem. Please see our full report and disclosures now available on our website, link in bio.
871
2,463
1,588
17,253
5,866,370
$TSM released earnings today. I was more curious about commentary given recent the market panic. For analyst Q&A rounds: Q: Citi explicitly challenged TSMC on AI financing, HBM, CoWoS, lasers and the risk of overbuilding capacity A: $TSM assessed CSP capex, DC construction conditions, customer financial conditions, and supply of critical components. Management's conclusion is still that "demand exceeds supply". As well as that AI accelerator revenue growth could exceed its previous 50%+ CAGR projection through 2029. This is a more meaningful statement since it addresses some concerns regarding recent selloff. Q: JPM asked if wafer supply could be tighter in 2027 than in 2026 (demand visibility) A: Wei (Chairman/CEO) stated $TSM even sends staff to verify DC construction progress on site. And after assessing all the factors, TSMC believes gap between supply and demand remains substantial. It could not determine whether the extent of the gap given customers continues to increase demand. Q: On pricing power: A: Wei responded that there is "no ceiling on TSMC's profitability" but that profits will not surge dramatically overnight. (Source: NextApple, Taiwanese press) Regarding earnings, very strong: as usual - TSMC record profits of $23.6B (68% Y/Y growth), and revenue jumping to $46.72B, beating expectations. - Q4 Guidance was $52.3-$54.3B revenue off 65-67% gross margins. TLDR: - No signs of any slowdown or collapse as AI demand far exceeds supply for foreseeable future. $TSM is probably the fundamental proxy/canary to how the overall AI semi trade is doing. - On financing concerns and fear of overcapacity showing up in semiconductor demand: $TSM specifically examined the risks investors are concerned about now and assessment still points to unmet demand. During the Q&A session, Wei said AI demand was "very, extremely strong," with absolutely no signs of slowing. Underlying AI semi supply chain growth is healthier than recent market panic suggest.
138
122
36
1,426
354,213
$AAOI OCP is providing favorable industry evidence for two of $AAOI's product markets—high-speed pluggable optics and external lasers for CPO. The ELSFP opportunity is particularly interesting because it gives AAOI potential exposure to CPO even if traditional optical modules eventually lose share in certain deployments. FIBERSTAMP is showcasing a 1.6T silicon photonics transceiver with a half-DSP architecture which directly targets $AAOI. It claims maximum power consumption of 18W, a 500-meter reach, and 28–35% less power than a comparable full-DSP design. This is a relevant competitive benchmark for AAOI's 1.6T products. The claimed power improvement hasn't been independently established across comparable configurations, and a demonstration isn't equivalent to hyperscaler qualification or large-volume deliveries. $AAOI not only qualifications they are scaling it right now. Also $CLS mentioned today hardware generations are moving faster, with the qualification window for 1.6T compressing toward 18 months. $AAOI claimed in March 2026 it has compressed its own time to market to nine months. So this is definitely s huge edge for them, I wonder if they are even more efficient now. I still think $AAOI has edge in several cases here. First, its 800G and 1.6T products can use the same manufacturing lines. AAOI management says the manufacturing processes are compatible, although 1.6T requires different final testing. This could allow the company to shift capacity between generations with limited incremental manufacturing investment. Second, $AAOI fabricates its own laser chips in Texas. That gives it more direct control over a potentially constrained upstream component. However, $COHR also has substantial vertically integrated laser manufacturing, so this isn't unique to $AAOI. Third, $AAOI has evidence of customer traction. It announced an initial hyperscaler 1.6T order exceeding $200 million in March, followed by meaningful 800G orders. Those are more persuasive indicators of competitiveness than an isolated engineering-efficiency statistic. Overall bullish for $AAOI and if China restrictions get implementer $AAOI will stay relevant for many years to come.
2
3
24
3,867
Ry Chang retweeted
"It’s so fucked up how you gotta treat the women you want like shit so they’ll like you" Dominance is the missing piece. The attraction isn't coming from being an asshole. It's coming from the fact that you're no longer asking her to decide who you're allowed to be. Men often think they have two options: be nice and get ignored, or treat her like shit and get desired. But dominance isn't mistreating her. It's not surrendering your frame just because you want her. You show interest because you want her. You lead because you know what you want. You set boundaries because her approval doesn't override your standards. You can be affectionate without becoming submissive to her reactions.
This quoted post is unavailable.
6
19
1
230
22,487
Ry Chang retweeted
大家看到 CPO 良率只有兩成,第一個反應是:光學要涼了。但我覺得剛好相反,良率問題不會讓光進機櫃延後,只會換一條路進去。 數字很簡單。台積電 COUPE 每顆光引擎良率 95%,一台交換器有 32 顆,連乘下來整台只剩約 19%。問題出在焊接之後光纖偏移,引擎焊死在板子上,一根光纖歪掉,整台報廢。鴻海第四季只打算出約 1 萬台 CPO 交換器,2027 年 3 萬台,佔營收很小。 所以業界開始轉向 NPO(近封裝光學)。光引擎改成插座式,光纖先接好、測好再插上去,壞了只換一個模組,不用整台丟掉。這就像燈泡從焊死在燈座上改成用旋的,壞一顆換一顆。業界說 NPO 最快 2027 年就能量產。 重點是:不管 CPO 還是 NPO,都要雷射光源。銅線到了更高速率就撐不住,光進機櫃是早晚的事。所以真正卡住的,還是上一篇講的磷化銦(InP)雷射晶片。Lumentum 說產能賣到 2029 年初,有些產品缺口高達七成。 但這篇也講了兩個真正的風險,我覺得比 CPO 良率更值得看: 第一,中國的 CW 雷射據報導已經供過於求,就算不直接進美國,也會壓到全球價格。 第二,低功率光源像 VCSEL 陣列,可能繞過高功率 InP 雷射的護城河,連 Lumentum 自己都在做。 長線基金現在多半還在觀望,因為沒人能算準 InP 短缺會持續多久。 推薦 @OptionKing666 這篇,是我最近看過把正反兩面整理最完整的光通訊筆記 👇
7
25
7
240
56,375
OPTICS LEDGER @Leoskie_L's arithmetic is the clearest case for optical test we've seen this week, and he's right that low CPO yield changes the route into the rack, not the destination. At 95% per engine and 32 engines, a co-packaged switch comes out whole about 19% of the time. Raise each engine to 99% before it's bonded and the same switch yields 72%. That's a power law, so the money sits in the per-engine number. A point of yield added before bonding is worth far more than anything done after. SemiAnalysis put the optics on one 36-engine switch at $36,000 to $40,000, with no way to pull a bad engine out. That's why test keeps moving up the chain toward the wafer. On our Part 2 map, a pluggable module has two main checkpoints and a CPO switch has five: laser test and burn-in, the photonic chip on the wafer, again after the electronics are bonded, the finished engine, and the full package. The NPO route he describes changes the math without removing it. A socketed engine that fails becomes a swap instead of a scrapped switch, and each engine gets tested with its fiber attached before it's plugged in. Either way, every engine gets tested more. Optical test is the thread we've been following. Part 1 on Oct. 4 walked the test flow, Part 2 on Oct. 6 mapped 28 test-equipment companies across six layers, and Part 3 on Oct. 10 set six questions for OCP. Meta's Wednesday 8:00 am Pacific talk covers its 51.2T CPO validation and its 102.4T integrated-optics switch. Any field failure data there tells us what per-engine yield looks like in practice. $TER $FORM $AEHR $KEYS $TSM
大家看到 CPO 良率只有兩成,第一個反應是:光學要涼了。但我覺得剛好相反,良率問題不會讓光進機櫃延後,只會換一條路進去。 數字很簡單。台積電 COUPE 每顆光引擎良率 95%,一台交換器有 32 顆,連乘下來整台只剩約 19%。問題出在焊接之後光纖偏移,引擎焊死在板子上,一根光纖歪掉,整台報廢。鴻海第四季只打算出約 1 萬台 CPO 交換器,2027 年 3 萬台,佔營收很小。 所以業界開始轉向 NPO(近封裝光學)。光引擎改成插座式,光纖先接好、測好再插上去,壞了只換一個模組,不用整台丟掉。這就像燈泡從焊死在燈座上改成用旋的,壞一顆換一顆。業界說 NPO 最快 2027 年就能量產。 重點是:不管 CPO 還是 NPO,都要雷射光源。銅線到了更高速率就撐不住,光進機櫃是早晚的事。所以真正卡住的,還是上一篇講的磷化銦(InP)雷射晶片。Lumentum 說產能賣到 2029 年初,有些產品缺口高達七成。 但這篇也講了兩個真正的風險,我覺得比 CPO 良率更值得看: 第一,中國的 CW 雷射據報導已經供過於求,就算不直接進美國,也會壓到全球價格。 第二,低功率光源像 VCSEL 陣列,可能繞過高功率 InP 雷射的護城河,連 Lumentum 自己都在做。 長線基金現在多半還在觀望,因為沒人能算準 InP 短缺會持續多久。 推薦 @OptionKing666 這篇,是我最近看過把正反兩面整理最完整的光通訊筆記 👇
3
2
17
1,389
NEW: My conversation with Cem Karsan about the Big Print that's coming: the government stepping in and buying stocks outright, which he thinks could be months away. Let us know your thoughts in the comments. TIMESTAMPS: 00:00 Preview 1:02 There's a major player "bullying" the market 4:33 The five pressures backing Washington into a corner 8:58 How a few stocks are hiding how weak the SP500 looks 16:00 Is buying stocks the government's only move left? 21:59 Why they need a crash, and why it hasn't happened yet 27:32 Who really controls the world's oil right now 34:00 The debt wall nobody is talking about 43:14 His views on Bitcoin, and where he thinks it hits a ceiling 54:25 How much will the market crash? 58:22 The AI story that made him start watching for a trigger 1:05:41 What he'd actually do with his money right now Follow Cem at @jam_croissant
42
70
8
500
141,010
卧槽!TradingView 上那些复杂的买卖信号,居然有人直接把底层 Pine Script 源码开源了! 连多空力量、趋势反转、动量分析和交易策略,都能拆开研究。 这个开源项目叫 TradingView-Proprietary-Indicators。 它不是普通的指标合集,而是一套涵盖趋势识别、量价分析、K 线形态和交易策略的源码库,做股票量化的可以直接拿来研究。 核心亮点: 1️⃣ QQC 量价动量:结合多空 K 线数量、实体影线和成交量,分析市场背后的多空力量。 2️⃣ MACD-V:对传统 MACD 进行波动率标准化,研究不同市场环境下的动量强弱。 3️⃣ Ichimoku Cloud:一目均衡表,结合云层、转换线和基准线判断趋势与交叉信号。 4️⃣ ADX 趋势过滤:区分趋势与震荡行情,辅助过滤无趋势环境中的交易信号。 5️⃣ K 线形态识别:支持 Engulfing、Inside Bar、Kicker 等形态,识别潜在反转与突破机会。 6️⃣ Trading Strategies:提供 Kijun Arrow、Bollinger Bands Force、HOLP 等策略源码,研究具体进出场规则。 我比较感兴趣的是 QQC 和 MACD-V。 一个研究量价背后的多空力量,一个通过波动率判断动量强弱,再结合 ADX 过滤震荡,就能尝试搭建自己的量化交易策略。 最关键的是,这些不是只能看买卖箭头的黑盒指标,源码可以直接阅读、修改和研究。 经常使用 TradingView、研究技术指标,或者想开发股票量化策略的,这套源码值得收藏。 github.com/harryguiacorn/Tra…
好家伙!TradingView 上那些常用的技术指标,Python 居然有一整套现成的 RSI、MACD、ATR、布林带,连 K 线形态识别都有,150 多种算法直接调用,做股票量化的可以看看! 项目叫 TA-Lib Python,12K+ Stars,BSD-2-Clause 协议。 它是经典 TA-Lib 技术分析库的 Python 封装,底层采用 C 语言计算,通过 Cython 提供高性能接口。 以前开发量化策略,光是写指标公式、处理行情数据、验证计算结果,就得折腾不少时间。 现在几行 Python 代码,就能把这些指标接进自己的量化系统。 几个比较硬核的能力: 1️⃣ 150+ 技术指标:RSI、MACD、ADX、ATR、布林带等全覆盖,趋势、动量、波动率和成交量分析都有。 2️⃣ K 线形态识别:自动识别吞没形态、锤头线、十字星等,把原本靠肉眼判断的形态转换成量化信号。 3️⃣ Cython 高性能:底层调用 C 语言库,官方称相比旧版 SWIG 接口快 2–4 倍,适合批量指标计算。 4️⃣ Python 数据集成:支持 NumPy、Pandas、Polars,可以直接处理行情数据,接入自己的量化策略。 5️⃣ Streaming API:支持快速获取最新指标值,方便接入实时行情监控和交易信号系统。 我觉得它真正厉害的,不是指标数量多,而是把量化开发中最重复的计算工作给解决了。 比如用 yfinance 获取美股行情,TA-Lib 计算 RSI、MACD 和 ATR,再交给 vectorbt 批量回测。 行情数据 → 技术指标 → 交易信号 → 策略回测,整个流程就串起来了。 做 Python 股票量化、自动化选股或者策略研究的,这个项目值得收藏。 仓库:github.com/TA-Lib/ta-lib-pyt…
1
35
4
201
26,866
Ry Chang retweeted
Replying to @BenBajarin
$GFS deal with TSMC, $2 B for 5 years for CPO collab. I'd say Transceivers will dominate for sure beyond 2028 teapawt.org/AnalysisOp/status/2109…
$GFS 'll participate in the juicy part of integrating the OE in the packaging? "Today's pluggable Transceivers carry their own small board with a DSP, lasers and the photonic chip In CPO, the Optical Engine moves to the interposer;it shares the silicon with the GPU and the HBM"
1
3
202
Ry Chang retweeted
摩根士丹利 OCP大會與800V之爭 800VDC預期過早,但「全面延期」敘事過度 - 市場對800V直流供電的預期跑在實際部署前面,但後續「延期」說法又把變化講得太廣。預期已回歸合理,下週OCP大會可能成為相關股票的正面催化劑。 原生800V機架確實推遲 - 原生800V機架交付時點向後移動。 - Rubin Ultra世代仍使用54VDC,首個明確原生800V機架預計推到2028年左右(約1MW的Feynman世代)。 - 原計劃中的Kyber機架形態不再用於Rubin Ultra。 其他部分進展正常,未全面停滯 - 800V側掛電源(sidecars)已在出貨。 - 2027年Power Centers與VRM可先行擴量。 - 整體電源架構過渡方向未變,只是「原生」部分爬坡形狀有變。 對半導體的影響 - 受益股:電源管理/功率分立相關股票( $ADI 、 $TXN、 $ON、 $WOLF、 $NVTS 等)仍是受益者。 - 更大機會:可能推高明年後緣利用率,甚至出現短缺(目前後緣利用率已很高,僅中國與TI例外)。 - OCP催化:下週全球OCP峰會有較大電源議程(含專門800V/HVDC場次),有望強化「趨勢仍在」的敘事。
摩根士丹利:OCP 大会应厘清 800V 之争——原生 800V 机架推迟至2028年、Rubin Ultra 仍用54VDC;板块估值回落至约6.5x 报告认为,市场对 800V 直流供电架构(800VDC)的预期此前已跑在实际部署节奏前面,但随后关于「800V 延期」的叙事又变得过于宽泛。
1
13
107
10,062
Ry Chang retweeted
🚨 Tech isn’t selling off. Semis are. At 1:05 pm ET, $QQQ was down ~0.7% from Friday’s close vs. ~0.1% for $SPY. The chip ETF $SMH was down ~1.8% (Alva intraday prices). It looks like a semiconductor-led selloff, amplified by oil and inflation worries. The main catalysts: 1/ Oil and rates: Brent rose above $106 after the weekend strike on Riyadh’s airport. Higher energy prices revive inflation concerns ahead of this week’s CPI, which is particularly uncomfortable for growth-stock valuations. Reuters identifies oil as the broad-market pressure. The Treasury market is closed today, so there isn’t a fresh bond-yield move to point to. 2/ Memory-chip competition: China’s $CXMT says it plans to introduce denser server memory by year-end. That is a plan, but it challenges the tight-supply pricing story behind memory stocks. Micron was down 3.5% at 1:05 p.m. ET. 3/ AI-chip spending is being reassessed: Microsoft unveiled Maia 300 and plans large deployments of its own chips next year. That raises the longer-term question of how much accelerator spending stays with outside suppliers. Microsoft itself was up at 1:05 p.m. ET, while Nvidia and AMD were down. What's your move today? Are you buying the dip in chips?
4
1
4
2,283
NANYA TECH Q3: DRAM ASP +60% QOQ, GROSS MARGIN HITS 85.7% per MoneyDJ / company self-reported results; color from TrendForce • Revenue NT$133.65B, +61.9% QoQ • Shipments fell mid-single digits. This was all price. • Op margin 80%, EPS NT$25.40 • AI infra (servers, networking, storage) already >20% of sales • LTAs cover ~50-70% of the book. Old contracts rolled off in June. Customers are pushing 3-month deals out to 3 years, and some want equity stakes just to lock capacity (TrendForce / CNA) This morning Korea memory sold off on peak-cycle fears. Nanya just printed software margins on commodity DRAM with bits falling. That's not a peak. That's allocation. $MU $SKHY $SSNLF $NNYAF
SK hynix just fell for a 4th straight session in Seoul. ₩1.84M on Oct 2, ₩1.6M range today. All while HBM prices are forecast to more than double next year. That's not a demand problem. Listen to what SK's chairman Chey Tae-won said in the last few days: • "If China or Micron of the U.S. expands capacity at a faster pace, South Korea's market share shrinks... We need to move on expansion quickly." • "We moved up the construction timetable for the Yongin cluster by 12 years, but even that is expected to fall short." • The ₩400T Honam cluster may now go up in parallel with Yongin instead of after it • SK wants to more than double its memory output capacity within five years (all per Seoul Economic Daily) And Samsung is reportedly looking to pull its first P5 DRAM/HBM tool move-in forward to around May-June 2027, per Korean media. 2027 pricing already looks locked. TrendForce sees HBM average selling prices up 121% next year, and Micron has locked in most of its 2027 HBM output at higher prices. Memory cycles don't die when demand slows. They die when everyone builds at once. Korean press is blaming capex burden and peak-cycle fears for the selloff. My guess is the market's already trading 2028-29 supply, not 2027 prices. If Korea really builds two clusters in parallel, the toolmakers get paid first, no matter who wins the price war later. $SKHY $MU $SSNLF $LRCX $AMAT
2
1
8
1,682
Ry Chang retweeted
Replying to @zephyr_z9
Problem is RAM supply
325
352
193
4,249
1,120,360
Qualcomm CEO reveals AI companies want phones running 100-billion-parameter models continuously by 2028--and confirms some are building phones Cristiano Amon: “I have to be careful not to disclose names, but some AI companies—some of the companies at the forefront of AI—are telling Qualcomm, ‘By 2028, I need to make sure that I have the ability to run at least a 100-billion-parameter model in a phone, and I need to have this thing running kind of all the time.’” Alex Heath: “So you’re saying there are AI companies that are talking about—” Cristiano Amon: “Well, AI companies are starting to do phones.”
Qualcomm’s CEO on the coming AI phone supercycle @cristianoamon on the China vs. US AI sentiment, how AI is changing smartphones and wearables, Qualcomm’s data center strategy, and 6G. We discuss why he thinks phones aren’t going away, how agents will use our personal context, and why he expects smart glasses to become the biggest category of AI wearables. We also cover differences in consumer AI adoption between China and the US, Qualcomm’s acquisition of Modular to offer developers an alternative to Nvidia’s software, and what 6G could change about the devices we use. This conversation was recorded live at Qualcomm’s recent Snapdragon Summit in Hawaii. Timestamps: 00:18 How China and the US see AI 04:33 Should AI development slow down? 07:25 Trump, China, and chip geopolitics 08:25 Why the smartphone market is shrinking 11:15 Memory prices and pent-up demand 13:15 Can cloud agents replace the phone? 18:30 What is an AI smartphone? 25:05 The state of VR and smart glasses 29:30 Is there an iPod moment for wearables? 30:45 Why Qualcomm is building data centers 34:45 Modular and the fight against CUDA lock-in 39:05 What 5G actually changed 42:00 Why 6G is built for AI 45:05 When 6G arrives Spotify: open.spotify.com/episode/3fT… YouTube: youtu.be/FcYL3h5bwlk Substack: sources.news/p/qualcomm-ceo-… Apple Podcasts: podcasts.apple.com/us/podcas… Thanks to the show’s premiere sponsors: @Atlassian, @meetgranola, and @mercury.
113
351
170
2,531
797,979
Ry Chang retweeted
I have received an extremely disturbing dm that contained a link to an alleged telegram groupchat with tons of screenshots of the Cornell 7 rapists sharing nude images of a girl and bragging about it. It was also found on 4chan and dates back to 2024 when the rapes took place. I do not know if the girl in the images is the victim or someone else. Winston, Matthew, Johnathan, Diego, and both Scott Norris and Scott Kretzschmar are seen participating in the messages and sharing imagery of the girl on Snapchat. These are some of the messages In one of the snapchats Johnathan says “I think she fell asleep” as he’s ejaculating on her face and showing a photo of what looks like a girl passed out. Winston is also encouraging them and messages “post more.” Diego says “Imma miss tonight fkkk” They appear to be trying to find another girl to invite over and Scott K says “Better not be a fatty again” Matthew then sends them a nude photo he received from a girl saying “JHU fatty sent me this btw” It seems like they often had group sex together based off these messages. They then share an image of a girl to which Jonathan says “Yuck. I wish she worked out half as hard as she sucked dick.” Winston also uses a slur and says “This ch**k slut won’t be walking tomorrow” This is some of the grossest things I’ve seen and it is clear they knew exactly what they were doing and were degrading women. These messages also prove this did not happen once and they have a history of looking for women to invite over. I do not know if the media has seen these messages but they are now public. Some of the girls faces are blurred but some aren’t. For that reason I do not feel comfortable posting a direct link to the channel. I will share a few of the messages where the girls cannot be identified as well as proof of what I posted above with the girls image blurred out.
824
6,736
1,803
43,047
28,328,893