[will-burner]

This is the first I've heard of Cerebras Systems.

From the article

>Cerebras had a net loss of $66.6 million in the first six months of 2024 on $136.4 million in sales, according to the filing.

That doesn't sound very good.

What makes them think they can compete with Nvidia, and why IPO right now?

Are they trying to get government money to make chip fabs like Intel or something?

[hn_throwaway_99]

You seemed surprised that this company is having an IPO to actually raise funds for operations and expansion, vs as just an "exit" where VCs and other insiders can dump their shares onto the broader public.

I might be a bit suspicious if a company in some low-capital-intensive industry was IPOing while unprofitable, but this is chip making. Even if they're not making their own fabs this is still an industry with high capital requirements.

We should be thrilled at a company actually using an IPO for its original intended purpose as opposed to some financialization scheme.

[knowitnone]

they don't make chips. they design and contract TSMC to fab the chips. The high capital is in design tools and engineers.

[hn_throwaway_99]

Thanks - I said that in my comment, but then just realized I had a typo of "fans" where it should have said "Even if they're not making their own fabs..."

[theogravity]

Does this mean that they couldn't find VCs to raise more cash?

[fakedang]

Cerebras is currently heavily backed by the Emirati government's sovereign wealth fund.

[ketzo]

VCs offer cash on different terms than the public does. This just means Cerebras believes it can get capital more cheaply (or on otherwise better terms) than it can from VCs.

That might mean VCs are turning them down, yeah, but that’s just one of many possible factors into “where do we raise money”

[chapelier]

I don’t think that’s been a problem. I’ve been following the pre IPO market on them for a while and pretty much any shares at a 7b valuation have been snapped up pretty much same day

[est31]

Nvidia's moat is real but not big enough that one can't surpass it with a lot of engineering. It's not the only company making AI accelerators, and this has been the case for many years already. The first TPU was introduced in 2015. Nvidia has just managed to get a leader position in the race.

[AlotOfReading]

Saying it's "just" a lot of engineering effort to catch up isn't wrong, but it understates the reality. There are very few organizations on earth that have the technical and financial resources to meaningfully compete with even small parts of Nvidia's portfolio. Nvidia's products benefit from that breadth of strengths and the volumes they ship.

They don't just make accelerators, they'll sell you the hardware too (unlike TPUs). They don't just sell you the hardware, the software ecosystem will work too (unlike AMD or Intel). That hardware won't just do a lot of computations, it'll also have a lot of off-chip memory bandwidth (vs Cerebras or others). Need to embed those capabilities in a device that can't fit a wafer cabinet or a server rack's worth of compute, Nvidia will sell you similar hardware that uses a similar stack, certified for your industry (e.g. automotive). Take any of that away and you're left with a significantly weaker offering.

Also they benefit from the priority of paying fabs a lot of money and placing a lot of orders.

If anything, Nvidia is less dominant than they should be because they've managed to ensure absolutely no one wants to buy from them when there are viable alternatives.

[kortilla]

People said the same about Cisco, Intel, IBM etc. It will only be a matter of time for companies to eat into the high margin stuff for specific use-cases and grow from there.

[skeptrune]

There's something weird about the market right now in that all the AI budgets being used to by GPUs are loss-leading. Orgs are treating the spend as a waste anyways, so I suspect they aren't going to be looking to cut costs. Make Cerebras a hard sell imo.

[talldayo]

> Nvidia's moat is real but not big enough that one can't surpass it with a lot of engineering.

Yes, but you also need a lot of capital if you want node parity with them. Nvidia (supposedly) spent an estimated $9 billion dollars getting onto TSMC's 4nm node. https://www.techspot.com/news/93490-nvidia-reportedly-spent-...

[will-burner]

> Taiwan Semiconductor Manufacturing Company makes the Cerebras chips. Cerebrus warned investors that any possible supply chain disruptions may hurt the company.

They get their chips from the same company that Nvidia does.

[ajb]

Virtually any competitors to Nvidia would be in the same position.

It's not necessarily to TSMC's advantage for Nvidia to become a monopolist either, although they wouldn't be totally dependent on Nvidia even if they did because TSMC serves every chip market.

[alephnerd]

They both contract TSMC to fabricate their chips.

The actual design and R&D is still done by Nvidia, Cerebras, AMD, Groq, etc.

Think of TSMC like Kinko's - they do printing and fabrication which is very low margins.

The main PMF for Cerebras is in simulations, drug discovery, and ofc ML.

As I've mentioned before on HN, Public-Private Drug Discovery and NatLab research has been a major driver for HPC over the past 20 years.

[est31]

TSMC has a market cap of 0.9T USD. It would be the 7 largest US company by market cap if it were one. Manufacturing chips is extremely profitable, at least in the current climate. It used to be that software is more profitable than hardware, which is more commoditized, but AI gave hardware companies a renaissance of sorts.

It's not a simple process at all but requires a lot of engineering and engineers to do it.

https://companiesmarketcap.com/usa/largest-companies-in-the-... https://companiesmarketcap.com/tsmc/marketcap/

[alephnerd]

> Manufacturing chips is extremely profitable

It only became profitable NOW in the last 2-3 years.

Before that, foundry after foundry was shutting down or merging.

TSMC, UMC, Samsung, Intel Foundry Services, and GloFlo are the last men standing after the severe contraction in the foundry model in the 2000s-2010s due to it's extremely high upfront costs and lack of moat to prevent commodification.

[extesy]

TSMC margins are over 30% and growing [1] - that's very far from "low".

[1] https://www.macrotrends.net/stocks/charts/TSM/taiwan-semicon...

[alephnerd]

30% net due to a near monopoly and a recent upswing due to Nvidia.

Almost every other foundry system died because of low net margins.

Software (and fabless hardware like chip design) is expected to have 60-70% gross margins or the ability to reach that.

Semiconductors is part of TMT just like Software or Telecom, and this has an impact on available liquidity.

This is why TSMC is heavily subsidized by the Taiwanese government.

[extesy]

TSMC is neither software nor fabless. I'm not sure we are talking about the same company, there seems to be some disconnect here. For hardware business 30% margins are high, Apple is one of the most famous exceptions.

[alephnerd]

> For hardware business

When a foundry wishes to raise capital from the private or public markets, it's bucketed under TMT - which includes software and fabless hardware as well.

This means it's almost impossible to raise capital without a near monopoly and/or government support and intervention - which is what Taiwan did for TSMC and UMC - because the upfront costs are too high and the margins are much lower compared to other subsegments in the same sector.

This is why industrial subsidizes like the CHIPS act are enacted - to minimize the upfront cost of some very CapEx heavy projects (which almost everything Foundry related is).

[typon]

Kinko's is not the pinnacle of human engineering - TSMC is. A slight difference there.

[Cyph0n]

> Think of TSMC like Kinko’s

What an amazingly reductive analogy :)

[ericd]

Compare it to the same period last year ($8.7M in sales). That’s a pretty solid growth rate.

[blurbleblurble]

Their tech is very impressive, look it up.

[pldpb]

It's a deadend. SRAM doesn't scale on advanced nodes.

Similar to Tenstorrent who chose GDDR instead of HBM, they throught production AI models won't get bigger than GPT3.5 due to cost.

[txyx303]

I don't think they rely on SRAM very much for training. https://cerebras.ai/blog/the-complete-guide-to-scale-out-on-... outlines the memory architecture but it seems like they are able to keep most of the storage off wafer which is how they scale to 100s of GB of parameters with "only" 10s of GB of SRAM.