7 ms·
Still parroting the same uninformed takes from January >DeepSeek claimed to have built its base model for about 5% of the estimated cost of GPT-4
by htrp 1y ago
Still parroting the same uninformed takes from January
>DeepSeek claimed to have built its base model for about 5% of the estimated cost of GPT-4
- juujian 1y agoWhat do we know now that we did not know in Jan? Is there some information on this that I have missed?
- fzzzy 1y agoThey didn't include the costs for developing v3, the base model. [edit] also they seem to be saying r1 is a base model, which it is not. Very sloppy.
- xnx 1y agoDidn't they also train off of ChatGPT API output?
- ipsum2 1y agoThis is a rumor that has not been confirmed by OpenAI nor DeepSeek.
- nextaccountic 1y agoI don't see the issue here. OpenAI trained ChatGPT off my own comments, and your comments, and the comments from the person you replied to.. I didn't authorize it and you probably didn't too. Meta was caught pirating over 80TB of books to train their AI, and they are claiming not only training AI on other people's stuff is legal, but piracy is also legal (well at least, piracy done by US tech giants is legal)
- xnx 1y agoFor sure. I was just pointing out that DeepSeek is not going to "beat" ChatGPT if DeepSeek relies on it.
- natrys 1y agoYou could maybe make that accusation about V3 (to the extent that it's a bad thing and not fair use, specially considering amoral origin of OpenAI's models in first place), but don't think the claim makes sense for R1 since OpenAI's o1 did not expose its CoT traces even in API. They published about GRPO (key algorithm behind R1) a full year before[1] they scaled it for R1. Given the research they do in open, it's not far-fetched to think they had the talent and technical know-how to achieve R1 on their own. [1] https://arxiv.org/abs/2402.03300 https://arxiv.org/abs/2402.03300
- joejoo 1y ago[dead]
- irjustin 1y agoYeah apparently they parked the cost of hardware, 50k GPUs and model development underneath another entity, high-flyer because it was "shared resource".
- AustinCarrBW 1y agoYou're misreading. The article is referring to V3 when it cites the base model behind R1. It does not say R1 is the base model.
- Analemma_ 1y agoBloomberg still has not retracted (or even really commented on) the Supermicro spy chip story, preferring to hope people just forget about it if they maintain total silence. They're fine if you need to look up where the Nasdaq closed yesterday, but don't expect serious tech reporting from them.
- mannyv 1y agoIt's interesting because servethehome.com found some oddly labeled chips at one point: https://www.servethehome.com/dude-dell-hpe-ami-american-megatrands/ https://www.servethehome.com/dude-dell-hpe-ami-american-mega... But yeah, saying the chips are everywhere is BS.
- protimewaster 1y ago> not retracted (or even really commented on) the Supermicro spy chip story They doubled down on it. They did a follow-up claiming that a cyber security researcher from a US-based firm had been called in to investigate suspicious traffic at a US telecom. The investigators claimed to have logs and a bunch of other evidence. The investigators also claimed that Bloomberg was misleading people by focusing on SuperMicro, as they'd reportedly seen to with other manufacturers too.
- NicoJuicy 1y agoHikivision literally sends hidden packages from their cameras to China. Discovered by our security team where I work. It's the reason our VMS doesn't have support for Hikivision cameras.
- ta20240528 1y ago"Discovered by our security team where I work" Any published where?
- NicoJuicy 1y agoInternal audit in the company for customers ( security Company). A couple of years later, our US customer had the same conclusion. Probably that one was published ( US government)
- zeroq 1y agoI'll put my tinfoil hat on and say it plays to the current US vs China "propaganda" tune, that US is winning on all fronts, but the ice thin and have to support local tech behemoths to full extent to secure our position in this world defining struggle.
- cookiemonsieur 1y agoIt's not US vs China. It increasingly looks like China VS a conglomerate of multi national companies with foreign born billionaire CEOs whose HQs happen to be located in the USA.
- cedws 1y agoChina vs Chinese migrants in the US
- tempeler 1y agoThe accuracy of this comparison is highly speculative. One should not ignore the possibility that dominant firms in the market might be inflating their cost figures to block new entrants and extract more capital from investors through such narratives. When you compare electricity prices in China with those in the U.S., such a large gap would require a truly extraordinary breakthrough to be justified.After all, these are privately held commercial firms, and they are not obligated to disclose their financials accurately.
- SV_BubbleTime 1y agoAgreed. For all anyone really knows, it was 5% for OpenAI compared to Deepseek. If you believe that Deepseek was released to undercut US AI value (duh) it makes no sense to take the official line as the absolute truth.
- jiggawatts 1y agoElectricity is the cheapest input by far! People and large networked GPU clusters are far more expensive. Typical models are now trained on clusters of roughly 20K GPUs. Even if you get a volume discount you still need cabling, switches, etc… The minimum entry price to play in the game at this level is about 200-500 million dollars. Meta spent something like $10B on their AI compute, for comparison.
- charleshn 1y agoIndeed, see https://semianalysis.com/2025/01/31/deepseek-debates/ https://semianalysis.com/2025/01/31/deepseek-debates/ for a well researched article about the actual cost.
- oefrha 1y agoThe provided source for every concrete figure on DS in that article is "we are confident that", "we believe" or something equivalent. How is it any better researched than any other article with a conflicting set of beliefs?
- Der_Einzige 1y agoThere are times when "just trust me bro" is okay. Semianalysis articles are one of those times. You are free to pull the contrarian "source please" shit, but the reality is that they are far more accurate at most types of GPU, Cloud, or AI analysis than almost anyone on this websites or anywhere else. Just trust them bro. Unironically.
- oefrha 1y agoWe're talking about a narrative that's affecting tens of billions of investment dollars, possibly more. I'm not gonna "trust me bro" anyone on this.
- yorwba 1y agoThen you definitely shouldn't trust the unsourced, false claim that "DeepSeek claimed to have built its base model for about 5% of the estimated cost of GPT-4." DeepSeek never claimed this.
- oefrha 1y agoYes, I don’t trust that either.
- Der_Einzige 1y ago
- yorwba 1y agoWhat DeepSeek actually claimed: "DeepSeek-V3 costs only 2.788M GPU hours for its full training. Assuming the rental price of the H800 GPU is $2 per GPU hour, our total training costs amount to only $5.576M. Note that the aforementioned costs include only the official training of DeepSeek-V3, excluding the costs associated with prior research and ablation experiments on architectures, algorithms, or data." https://arxiv.org/pdf/2412.19437 https://arxiv.org/pdf/2412.19437 They don't even claim to have spent $5M (since they own their GPUs instead of renting them by the hour), it's a purely notional figure suitable for an academic paper. But when R1 got released and started generating hype, it was the only dollar figure anyone had to go on, so it got interpreted as more significant than it is.
- AustinCarrBW 1y agoThe article specifically says "it's likely this sum referred only to the final training run—a data-refinement process that transforms a model’s previous prototypes into a complete product—but many people perceived it as an insanely low budget for the entire project." The article also delves into the SemiAnalysis report, as well as denials from ex-DeepSeek employees.