coreteks (@coreteks) header
coreteks (@coreteks) profile picture

Youtube: https://teapawt.org/t.co/AGqevEiUXC Patreon: https://teapawt.org/t.co/IkVmWNIBYI

Planet Earth
Joined February 2016
Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)Photo by coreteks (@coreteks)
I've cracked it, Vega is back 😛
6
1
55
1,562
It's crazy how most people out there don't understand how much better the semiconductor industry will be without Jensen. His exit came 33 years too late.
Over the past week, I’ve been overwhelmed by the love and support of friends, industry colleagues, and people I've never met. Thank you to everyone who has shown kindness to my family and me. It means the world to me. I came to the United States 53 years ago, and becoming an American citizen is one of the greatest gifts and proudest moments of my life. NVIDIA has been my life’s work. Building it into the company it is today alongside my colleagues has been a privilege and an honor. Nowhere but in America could a company like NVIDIA be possible. I love our country, and I love NVIDIA. I did not live up to the standards expected of everyone at NVIDIA, and I have made serious mistakes in my personal life. I created problems for the Board that I deeply regret. I own my mistakes completely, and I am committed to cooperating with the Board’s independent investigation. Since my departure, I’ve seen an outpouring of calls from NVIDIA’s employees, investors, business partners, and customers for me to return as CEO. Many continue to reach out to me directly as the company navigates complex technology, market, competitive, and geopolitical dynamics. NVIDIA is a once-in-a-generation company. If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO. Thank you again to everyone who has shared their good wishes. I am profoundly humbled by your support. To my family: I love you, and I will spend the rest of my life being worthy of yours. Jensen
43
19
2
218
14,273
coreteks retweeted
I have a major result to share. I’ve found conclusive forensic evidence that the top 4 Chinese AI labs—DeepSeek, Moonshot, MiniMax and Ant Group—have built their newest models on NVIDIA’s most advanced Blackwell chips, which are barred from export to China. This sounds insane, but I discovered that almost every AI model carries a hidden signature of the GPU that trained it. When a model is first set up, the starting values of its weights are laid out in a pattern that depends on the exact GPU. Each combination of GPU and seed number produces a unique pattern—a sort of fingerprint of the hardware used to create the model. Traces of that pattern survive training into the published model. Again, this evidence isn’t based on any insider information or leaking. It’s based on a forensic analysis of the model weights released by the companies themselves. I searched more than six trillion combinations of chip layout and seed number and matched them to open-weights models. In doing so, I found conclusive evidence that six Chinese AI models—including DeepSeek-V4-Pro, Kimi K3, and MiniMax-M3—were initialized on NVIDIA Blackwell systems. Four of the models match Grace Blackwell, the GB200 and GB300 parts in NVIDIA’s flagship liquid-cooled rack-scale systems. I then verified these results on real hardware. Running unmodified PyTorch with the seeds I’d recovered, a Blackwell B300 reproduced DeepSeek-V4-Flash’s untouched starting weights bit for bit, and a rented GB200 reproduced Kimi K3’s to a correlation of 0.999995. A Hopper H100 reproduced none of DeepSeek-V4-Flash’s, and changing the seed by one broke the match. (DeepSeek’s seed is 953049583 by the way.) The method also was verified to match 24 known open-weights models to the hardware that developers said they used. It is exceedingly unlikely that these results are a coincidence. Across six trillion combinations, the best accidental match should reach only about 8 sigma (i.e., eight standard deviations above what random chance would produce). The real matches range from 49 to 1,385 sigma, and even the weakest single test hits 20. These were data-center-scale training runs. With some analysis of release schedules and model details, I estimate that the largest, Kimi K3 and DeepSeek-V4-Pro, used 2,000 to 3,000 Blackwell GPUs each. Clusters of this size cost on the order of $150 to $200 million each to buy, and across these labs alone would add up to more than half a billion dollars of export-controlled GPUs, not counting experimental compute or otherwise. It’s quite likely that we see multi-billion dollar expenditure on export-controlled GPUs. I am highly confident in this result, and it has been independently verified by senior researchers at OpenAI and Anthropic. You can read the full report at unemploymentlabs.com/checkpo…. There is much more I’ve identified about these models through other forms of checkpoint forensics like this, and will have more to share soon.
1,820
6,414
2,194
41,186
6,673,532
Who should be the next NVidia CEO?
54%Jim Keller
6%Bill Dally
11%Ian Buck
29%Other (comment who)
112 votes • Final results
14
8
1,089
- Keller is one of the few who would know how to tackle the impending switch from GPUs to specialized in-house hardware at hyperscalers - Dally has been lowkey NV's actual bread winner for a decade+ - Buck invented CUDA, would prob maintain the status quo (which would be bad)
5
286
Now let's wait for AMD's official announcement that Lisa Su has resigned. I mean, they copy NVidia in everything right?
4
1
17
735
Jensen's "resignation" (he got sacked, let's not kid ourselves) could be the catalyst that bursts the bubble. Let's go 💪
5
3
36
902
New Video: youtu.be/BWGSCROaYSk The end of the line for Jensen Huang
2
1
10
338
Turns out you CAN carry TWO TONS in your pocket or backpack!
1
1
6
428
Jensen Huang to his prison cell mates: "Why buy just ONE cigarette from me? Buy a whole pack! ...Trust me, the more you buy, the more you save"
1
14
585
ChatGPT's reasoning once the #AIBubble money runs out
5
533
No SOUPer for you!
en.gamegpu.com/news/zhelezo/… NVIDIA has canceled plans to release the GeForce RTX 5080 SUPER graphics card with 24GB of GDDR7 memory, which was meant to replace the 16GB baseline model and increase buffer capacity by 50% using 3GB modules. By early November, board partners still had not received the necessary design kits and technical documentation originally scheduled for October. Despite recent online speculation that the disappearance of the standard GeForce RTX 5080 from retail stores was intended to clear inventory for the new 24GB revision, the actual reason is a massive shift toward the artificial intelligence sector. All available silicon quotas, PCB supplies, and dense memory chips are being reallocated to high-margin enterprise AI accelerators. #NVIDIA #GeForce #RTX5080SUPER #RTX5080 #GPU #PCGaming #Hardware
2
10
638
This aggregator project is really cool 👌 Far from perfect but it should be said that the Windows/Lx PC survived subsidized consoles in the 90s-2000s, commoditized phones and tablets in the 2010s, and it might even survive Jensen in the 2020s No other platform is this flexible, resilient and fun 💛 #stillthemasterrace
Modders launch RecompLauncher, listing over 200 console game ports for PC and Mac 🔗videocardz.com/newz/modders-…
1
7
506
Biowho? ✅ Made good games 15+ years ago ✅ Adopted DEI ✅ Injected "non-binary" woke BS ✅ Called its customers "bigots." ✅ Self destructed Oh, those clowns. Just another pathetic leftist-Marxist ideology victim. Enjoy unemployment you retards.
2
16
542
SHOCKING: We have confirmed that multiple Neoclouds have been refused access to GTC Berlin due to them adding AMD GPUs, TPUs, and other chips to their cloud offerings Note that this doesn't apply to hyperscalers, as they have more buying power. But for neoclouds, executives feel NVIDIA has been using high-pressure tactics to keep them NVIDIA-only. As such, neocloud executives are starting to consider offering TPUs or AMD GPUs quietly to avoid getting pressured by NVIDIA and to avoid the feeling that NVIDIA will retaliate. teapawt.org/SemiAnalysis_/status/2…
SHOCKING: Many neocloud executives we spoke with feel that if they have non-NVIDIA networking gear on their cluster, or if their cloud has an AMD GPU or TPU offering, NVIDIA will retaliate. They feel that retaliation includes not giving early allocation or no longer supporting a potential IPO/VC raise.(1/3)🧵
191
166
109
1,795
588,304
Microsoft's new Surface line has SoCs from Intel, QCom and NV - but not from AMD, despite them being a long time MS partner. Why? I suspect that uncle Jensen told Satya: "you want future access to Rubin? you want to stay on good terms with 'the family'? .. then kick AMD out"
7
1
27
1,792
I wonder how much funding this got from Jensen and friends, directly or indirectly Will people pay to watch a 2h AI propaganda video at the cinema?
Today is historic MPA-approved trailer for first AI film with a traditional theatrical run Gods Don't Give Gifts is R-Rated Submitting for Best Animated Feature Oscar --> Variety article @Gossip_Goblin @fableshowrunner
1
6
672
New Video: youtu.be/6YbkHOibcsw I discuss how AMD's answer to DLSS 5 goes well beyond replacing character's faces, lighting and shadows
1
15
763
The elites and AI-Bros are redefining a threatening definition ("ARTIFICIAL") because accepting its ordinary meaning would endanger their goals, and because making the definition invalid means one need not answer the definition. This semantic evasion aims to reframe "artificial" (not reliable) with "super" (beyond human) because the "artificial" label threatens their status and morally corrupt objectives: to pocket as much money as possible while they can from a bubble that has no positive economic outcome. Refusing the "artificial" characterization in favour of "super" is a childish attempt at turning an apt insult that accurately describes what this type of intelligence is, ARTIFICIAL, into a badge of pride. It's about controlling the narrative: "AI is not doomed to collapse the economy, of course not! Because it's NOT artificial, it's actually SUPER! It's so dangerous and beyond your own comprehension - because you only have REGULAR intelligence - that if you don't give us more money you will be LEFT OUT of the SUPER revolution that we promised years ago and have continuously failed to deliver, all the while seeing no financial return, only ever-increasing unpayable levels of debt and a hardware market starved of memory." They can self-credentialize all they want with this rebranding but in the end money talks. AI has a modest place and a modest future: in LOCAL machines. AI has no future in GPU-driven mega data-centers that burn egregious amounts of energy, destroy the rest of the hardware industry while producing no economic value, and exist primarily as a circular investment platform that fills the pockets of the "arms dealers" - aka the morally corrupt memory and hardware manufacturers.
1
7
463