Remember when GPUs were the dominant tech to mine crypto, then ASICs and other spepcialized purpose-built hardware came out and made GPUs irrelevant for mining, wont that also happpen in regards to ai training/inference. like aws is attemppting to do with their trainium/inferentia chips. Sure they suck now, but won't they get better and thus lose the dependency on GPUs?
It's already happened hasn't it. Google's TPUs are apparently much more efficient at inference workloads. However the hardware isn't available to the public like Nvidia chips are.<p>I'm unsure about the energy efficiency of Tenstorrent cards, but they are at the moment much more cost effective than Nvidia. But software support is far behind, I suspect if it wasn't the cost would be more equal.
Google has two different types of TPUs now. One for inference and one for training.<p>Do we actually know they're more efficient per watt or for the power used and heat produced ? Or are they just easier to scale (and manufacturer) en masse ?<p>There's also the CUDA moat which is hard to get around
Inference has two steps, prefill and decode, with very different requirements on the compute to memory-bandwidth ratio.<p>Already these are run on different hardware, at least as I understand, and the computational needs could further drive differentiation.
there's also stuff like Talaas, where the ASIC is the actual model weights
Which is now owned by AMD<p><a href="https://www.forbes.com/sites/jonmarkman/2026/08/09/amd-buys-taalas-the-startup-that-carves-ai-models-into-silicon/" rel="nofollow">https://www.forbes.com/sites/jonmarkman/2026/08/09/amd-buys-...</a>
Are they modifiable after creation? If not that seems dangerously rigid until we reach a much more stable place with LLMs. It could be great for well worn architectures and models but building one now for LLMs seems like it'd be foolish when it'll be outdated in a few months.
Why would it be dangerous? It'd effectively be a stamped model at some point in time. The bleeding edge moves on every few months, but even now it seems to me that there's only a tenuous value proposition that's progressing for me. I'd rather have an instantaneous GPT 6, Opus 5.3, or Gemini Hydrogen that I own, rather than always leasing models I have no practical application for.
no, and you're right that it's rigid. they call them "hardened models".
I'd gladly buy a $50 Opus chip pulling 5 watts.
I think it would be on the order of ~1000 of them to fit opus.
I doubt they'd ever offer that until Opus was massively out of date if it's even possible to squeeze Opus into a chip at that power and budget.
The better question is, would someone be able to turn a profit offering a deepseek chip for a few hundred dollars? I wouldn't even care about the energy draw (within reason) having that running entirely locally at better than a few hundred TPS without going bankrupt to obtain VRAM would be a game changer.
Is 5 watt achievable?
Doesnt google themselves use nvidia chips. why do they need nvdia chips then?
Basically capacity issues. NVidia + Apple have bought up a significant portion of TSMC's manufacturing capacity. The primary supplier for Google's TPUs is Broadcom, with TSMC as a second source, and they're trying to ink contracts with other suppliers like MediaTek, but there're only so many semiconductor fabs in the world and TPUs need a fairly modern process.<p>Within Google itself the TPUs are definitely preferred for AI workloads, you just can't always get them.
Crazy that a company can use a contract to buy up all the supply to build an inferior product that’s less energy efficient and then sell it at a higher price than the purpose built counterpart. Market failure.
They're not really the same markets. NVidia is selling chips that AI runs on, Google is selling services that use AI. Google's play is to make better services than the competition can because they control the full infrastructure stack, thus preventing their markets from being commoditized. NVidia's play is to commoditize the AI providers so that the number of services expands rapidly, with NVidia chips powering all of them. Arguably it's Google's strategy that is the market failure, as it's designed to keep competition out, but...<p>Google's used this strategy a few times before, and it frequently doesn't work out. They kept MapReduce and GFS proprietary so they could build better big-data services than everyone else, but Doug Cutting from Yahoo clean-roomed it, and then eventually you had interviewees going "MapReduce? That's like Hadoop, right?" They had a long string of internal ML frameworks well before the rest of the industry (I remember in the '00s and early '10s Google was by far the leader in large-scale machine-learning) but had to open-source TensorFlow in the face of external competition, and then lost in the marketplace to PyTorch because TensorFlow was filled with internal WTFs and footguns needed to support specific Google projects. They had a big lead in datacenter technology over everyone else in the industry in the early 2010s, but kept it internal and proprietary until they were forced to sell it as a service by the huge growth of AWS and Azure, then had to play catch-up as a distant 3rd because they lacked the customer relationships and developer mindshare.
> They're not really the same markets.<p>Both are fabless customers of TSMC, which is the choke point and center of the high-efficiency chip market.<p>Unlike gp, I do not believe its a market failures as much as the cutting-edge process node being a natural monopoly. Once TSMC processes got really good, TSMC started capturing most of revenue dollars and reinvesting them, which made TSMC products even better - a positive feedback loop that makes it impossible for Samsung or Texas Instruments to catch up. Intel rode on a similar wave for decades
By the time Hadoop came out, Google had already moved on to Big Table.
BigTable and MapReduce are differennt use-cases. The former is a database; the latter is a framework for running batch jobs. The (internal) successor to MapReduce is Flume (2010), which is quite a bit more capable, but didn't come out until Hadoop was already well-established. The external equivalent is Apache Spark (2009), but the gap between internal and external technologies had already shrunk to be nearly indistinguishable by then.
> Crazy that a company can use a contract to buy up all the supply to build an inferior product that’s less energy efficient and then sell it at a higher price than the purpose built counterpart. Market failure.<p>They're not buying up "all the supply to build an inferior product". They're buying it up build a <i>different, high-demand</i> product.
It's logical that company like TSMC sells capacity to the highest bidder.<p>The main thing is that chip manufacturing is the most highly capitalized industry in the world. Each manufacturing step has gargantuan costs and so no company wants to manage (and pay for) all the steps - so TSMC buys chip manufacturing equipment from ASML and sets up a system where corporations buy capacity. It's essentially a very, very, very expensive print shop. The print shop doesn't make its own printers nor it print a newspaper.<p>And naturally, with very, very objects like leading-edge chips, "inferior" or "superior" is a ridiculous simplification. Nvidia chips are incredibly powerful and useful. Nvidia has been more committed than any other company to general purpose GPU programming.<p>But Nvidia also is doing it's best to sell at high, monopoly prices. That's why they buy a lot of capacity. The thing about the situation is these companies did spend a lot of money reaching this monopoly situation so from their viewpoint, this monopoly pricing might be fair. But I personally wouldn't care and hope some regulation stops it. But that kind of regulation would still have to get these companies significant profit on these investments, so they'd keep making them.
TPUs are a systolic array. It's not comparable to the hardware architecture that backstops CUDA, nor do they serve the same markets.<p>The writing was on the wall for TSMC ever since the 5nm node; Apple proved that anyone with enough liquid cash could block competitors from fab access.
Interesting. What manufacturing process is Broadcom producing them at ? I'm guessing they're not newly as small (lithography wise) as the best of the best yet but probably good enough
Broadcom isn't a chip foundry so calling it Google's supplier doesn't make sense. TSMC has been the primary manufacturer of TPUs but Google's hoping to use Samsun and Intel also going forward (and these are the only leading edge foundries in the world). Broadcom is co-developer of the chip.
What node process do they need to be made on ?
Presenting broadcom as an alternative to tsmc is like wet roads cause rain. The average knowledge of computer hardware development cycle in hn comment sections is frustratingly low.
google doesn't use them, they buy them to make them available for google cloud customers that require CUDA or other nvidia-only workloads. All training and inference at Google is done on TPUs.
> <i>However the hardware isn't available to the public</i><p>It is available to Oracle.<p><a href="https://www.oracle.com/news/announcement/ai-world-oracle-brings-google-tpus-to-oracle-cloud-infrastructure-2026-10-27/" rel="nofollow">https://www.oracle.com/news/announcement/ai-world-oracle-bri...</a>
This may be, but look at Alphabet's share price:
> However the hardware isn't available to the public like Nvidia chips are.<p>Why? Don't they want to add another $3-5T of market cap?
You have basically managed to sum up why NVDA keeps dominating, no on can "just buy a few TPUs" and that has been the case since the first TPUs were announced years ago (yes yes GPU shortages and so on, but no one questions general availability of GPUs), and the current Tenstorrent model support page looks like an invitation to beta test HW that might (or might not) be able to run current and future models at some point.<p>But the next gen of Tenstorrent HW looks great on paper - problem is, I could have written that before the current gen was released.<p>All of the allegedly competing HW looks like that and has looked like that for years, either a lack of availability, capability, or both.
TPUs aren’t as effective at training.<p>Inference has already been commoditised.
> Sure they suck now,<p>Some suck. Google's Tensor is an example of one that doesn't, that some consider superior to Nvidia's offering. There are likely others.
Inference, yes, see <a href="https://taalas.com/" rel="nofollow">https://taalas.com/</a> and <a href="https://chatjimmy.ai/" rel="nofollow">https://chatjimmy.ai/</a> As expected, efficiency boost is extreme. But you've got to pick the model you want to enshrine in silicon months or maybe years before you actually get that silicon (and maybe there are additional difficulties if the model is large. Maybe in a few years everybody will just want to use the same four or five models, with no tectonic shifts on the horizon. Then it'll be a no-brainer to make ASICs for those models.<p>Training? I'm not so sure. You need something flexible so that it can be trained. Burning the weights into the board is the opposite of that.
I read into Taalas when they first were putting out info and they also are wiring up massive CPU cache to the thing. It's not just the burned in models.
As much as I love the idea of Taalas, their chatjimmy demo is a hard-wired (and probably very quantized) 8b LLM producing garbage extremely fast (17k tokens/s). Extrapolating from the picture of their PCIe board in their demo, I would expect a frontier model to require an IC not smaller than my desk.<p>At these dimensions, you are better off producing general-purpose inference ICs that can run any model, instead of baking in a single fixed LLM.<p>So, Taalas (now AMD) will probably target small models with these (maybe something like qwen3.8-27b, which seems too large already), if at all.
To be fair GPUs have already changed a lot in response to the AI demands. A lot of silicon these days are tensor cores which have very few applications outside of deep learning.
The difference is the algorithm needed to be embedded in an asic is static for Bitcoin. Models are changing too fast
Yeah, that's my thesis since the beginning of this madness. I see it as a temporary blip (albeit it's already much much longer than I would expect !) and once the bubble pops, Nvidia will be back to their normal cruising speed. Which is a good company, making great products, but their valuation will be 1/10 of what it is now.
And if the boom continues, there is no way either that their clients won't develop a product of their own. It just doesn't make sense to buy chips by the 10s of billions $ on which Nvidia makes >50% margin.
There are also the opposing architectural directions to extend toward:<p>one in the "Cerebras" direction of ever-larger scale general purpose HPC silicon running inference workloads. Cerebras may have skipped ahead to the last act of that scaling story, notwithstanding any further wafer-diameter enlargements past 300mm. Whether the ending is happy or not is still unclear. Lot of iterative room to grow into it further and they have a formidable patent moat waiting for conventional die-sized AI semi vendors when they inevitably run up against limitations of single-die designs.<p>the other is in the "Taalas" direction of freezing a specific model into something like high speed ROM / FPGA and getting orders of magnitude inference speed improvements - for that specific model. Like Cerebras, Taalas may also have jumped directly to the architectural telos of that philosophy, with only iterative improvements in cost per work unit left to grind, though like CBRS quite a lot of grinding left to do especially in getting cost per custom tapeout down to near-disposable (hopefully, recyclable) levels.
NVIDIA GPUs stopped being GPUs a while ago, that didn’t happen for crypto, but now it seems the will drop the entire customer graphics market long term they just haven’t committed themselves to it. They can’t go back to GPUs anyway there are not enough gamers to reach the revenue numbers they need now and soon gaming/graphics revenue will be irrelevant for them. Most US companies won’t maintain a boutique of products that don’t have synergy or are actively competing for internal resources long term they put all their resources where the margins and growth are best and for NVIDIA that’s not gaming anymore.
> NVIDIA GPUs stopped being GPUs a while ago, that didn’t happen for crypto<p>But Bitcoin mining moved GPUs to ASICs many many many years ago and Ethereum switched from proof-on-work using GPUs to proof-of-stake: GPUs are pointless now for Ethereum. And Ethereum was <i>the</i> big one for GPU mining. All the others cryptocurrencies are too tiny to register.<p>Ethereum switching from proof-of-work to proof-of-stake has been the death blow to GPU mining at scale. I'm not saying some have not been profitable mining other less important coins: I'm saying the market for GPU mining shrank to next to nothing.
Nvidia has multiple B2B customers that <i>require</i> the consumer graphics featureset, like Nintendo. There's sufficient motivation in keeping the GPGPU/raster capabilities on the same card even if PC gaming dies as a market.
But why wouldn't nvidia be expected to be a major successful player in that development?
It's Nvidia's 80% profit margin that's threatened.
They didn't really get into it with Asics last time, at least. Are they looking into it this time?
So will people buy gaming GPUs fron Nvidia again after how if abused them this time ? Or will the AI obsession be their downfall & example for others what not to do and how not to alienate the customers who get you where you are now?
This is why I think their pivot/venture to general robotics was smart. However, I doesn't seem like they will become profitable on that front because they were outcompeted pretty hard in most measures I can conjure.
Yes. Half dozen companies are designing special chips to bypass/replace the GPU.<p>Bitcoin was lucky for Nvida, then AI was lucky. Third times the charm?<p>On other hand, luck favors the prepared. Beyond luck, they did make their chips accessible, easier to program than other GPU's. So bitcoin miners and then AI could utilize them.<p>There are still a lot of people that want to run regular liner algebra.
>Bitcoin was lucky for Nvida<p>Nvidia cards were the worst hashrate/purchase price of any cards. AMD were the lucky ones there if you consider anyone lucky in this situation. Consumers definitely weren't lucky because the GPU market hasn't recovered since the Etherium boom.
Maybe mojo will take off as an alternative to cuda?
Mining ASICs are bound to a single cryptographic hashing algorithm, meaning they do one thing and they do it very fast. Something similar may make sense for running models locally, but I believe the more multifaceted Nvidia chips make more sense for training.
Taalas takes a similar approach but at a much higher level of complexity - instead of an array of nnn tiny specialized cryptographic cores running the same algo (or a mix of several with interlocking dependencies), it freezes a single entire LLM on to one die, running inference both far faster and cooler than any general purpose compute but with the same one trick pony obsolescence drawback. Question remains whether high volume fab tech can be wrangled into a form that allows this approach to drop unit costs near "disposable/recyclable" levels of cheapness per chip. This also may not matter when models all start to converge on a 'good enough' level of coverage for specific domains that the chips can remain productive for 3-4 years before needing a recycle.
>This also may not matter when models all start to converge on a 'good enough' level of coverage for specific domains that the chips can remain productive for 3-4 years before needing a recycle.<p>Yeah I was thinking this will probably be the case. We may eventually settle into a slower release schedule as well, where a new model only comes every year or two.
Check out Taalas, recently acquired by AMD. They printed a model directly to silicon and it generates 16,000 tokens per second. You can even try a demo here:<p><a href="https://chatjimmy.ai/" rel="nofollow">https://chatjimmy.ai/</a>
Whats amazing is that this is around 5x faster than the top end Cerebras wafer-scale silicon, at much less power/cooling cost per token. Of course Cerebras is a general purpose design, but perhaps Taalas could license CBRS wafer-scale patents to create a single monster chip that held either a single massive trillion parameter-scale model, or could an MoE design with a half dozen smaller specialized models on the same wafer with a shared SRAM pool.<p>I wonder if cooling such a beast, with very non-uniform hot spot distribution, would be too complicated to bother with though.
They were acquired for their ROM-like function block IP not ASIC stuff.<p>That function block was worth the cost of accusation.
won't<p>those Asics have a fixed algorithm<p>Cerebras already does that an need to make a new chip for each new model<p>that's why you couldn't make an ASIC for Ethereum
Cerebras don't need to customise their chips. To my knowledge they are crazy fast but struggle to fit larger models and it's not clear they're economically favourable.<p>Ethereum's work function was designed to be memory bandwidth limited, which is something gous are already quite well optimised for. An asic would be faster but not by enough to be worth the investment, at least while gpu margins were less crazy.<p>The main thing is the actual number crunching and memory interfaces for LLMs are not crazy hard to implement, and the potential scale is huge, so NVIDIA's moat is not likely to be big enough to make their current margins sustainable in the longer run.
Not to mention, many ASICs were outclassed by a few well-tuned GPUs in terms of power-efficiency. There were only a few coins that bucked that trend, usually just altcoin/shitcoin experiments that weren't worth mining in the first place.<p>I'm kinda bewildered, it's bizarre seeing people cheer on quickly-obsolete accelerators instead of demanding their OEM to adopt GPGPU. CUDA will be relevant for much longer than any ASIC can ever be, even (or especially) if the transformer gets replaced overnight with a better architecture.
Yup. Pretty much. Any model can be a a hard wired circuit.
[dead]