Vsora Jotunn-8 5nm European inference chip
56 points
6 hours ago
| 6 comments
| vsora.com
| HN
leo_e
1 hour ago
[-]
Impressive numbers on paper, but looking at their site, this feels dangerously close to vaporware.

The bottleneck for inference right now isn't just raw FLOPS or even memory bandwidth—it's the compiler stack. The graveyard of AI hardware startups is filled with chips that beat NVIDIA on specs but couldn't run a standard PyTorch graph without segfaulting or requiring six months of manual kernel tuning.

Until I see a dev board and a working graph compiler that accepts ONNX out of the box, this is just a very expensive CGI render.

reply
mg
39 minutes ago
[-]
Six months of one developer tuning the kernel?

That seems like not much compared to the hundreds of billions of dollars US companies currently invest into their AI stack? OpenAI pays thousands of engineers and researchers full time.

reply
pclmulqdq
5 hours ago
[-]
It needs a "buy a card" link and a lot more architectural details. Tenstorrent is selling chips that are pretty weak, but will beat these guys if they don't get serious about sharing.

Edit: It kind of looks like there's no silicon anywhere near production yet. Probably vaporware.

reply
layer8
5 hours ago
[-]
Tapeout apparently completed last month, dev boards in early 2026: https://www.eetimes.eu/vsora-tapes-out-ai-inference-chip-for...
reply
embedding-shape
5 hours ago
[-]
Nice wave they've been able to ride if it's vaporware, considering they're been at it for five years. Any guesses to why no one else seemingly see the obvious you see?
reply
pclmulqdq
5 hours ago
[-]
Look at the CGI graphics and indications in their published material that all they have is a simulation. A It's all there without disclosing an anticipated release date. Even their product pages and their news page don't seem to have indications of this.

Also, the 3D graphic of their chip on a circuit board is missing some obvious support pieces, so it's clearly not from a CAD model.

Lots of chip startups start as this kind of vaporware, but very few of them obfuscate their chip timelines and anticipated release dates this much. 5 years is a bit long to tapeout, but not unreasonable.

reply
embedding-shape
4 hours ago
[-]
> Even their product pages and their news page don't seem to have indications of this.

This seems indicative enough for me, give or take a quarter or two probably, from the latest news post on their website:

> VSORA is now preparing for full-scale deployment, with development boards, reference designs, and servers expected in early 2026.

https://vsora.com/vsora-announces-tape-out-of-game-changing-...

Seems they have partners as well, who describe working together with a Taiwanese company as well.

You never know, guess they could have gotten others to fall for their illusions too, it's not unheard of. But considering how long time something like this takes to bring to market, that they have dev-boards ready is months rather than years at least gives me enough to wait until then to judge them too harshly.

reply
lukan
2 hours ago
[-]
"that they have dev-boards ready is months rather than years at least gives me enough to wait until then to judge them too harshly."

So far, they just talk about it.

reply
bangaladore
2 hours ago
[-]
I love that the JS loads so slow on first load that it just says "The magic number: 0 /tflops"
reply
Ethan312
1 hour ago
[-]
Always good to see more competition in the inference chip space, especially from Europe. The specs look solid, but the real test will be how mature the software stack is and whether teams can get models running without a lot of friction. If they can make that part smooth, it could become a practical option for workloads that want local control.
reply
N_Lens
4 hours ago
[-]
An FP8 performance of 3200TFLOPS is impressive, could be used for training as well as inference. "Close to theory efficiency" is a bold statement. Most accelerators achieve 60-80% of theoretical peak; if they're genuinely hitting 90%+, that's impressive. Now let's see the price.
reply
all2
5 hours ago
[-]
288GB RAM on board, and RISC V processors to enable the option for offloading inference from the host machine entirely.

It sounds nice, but how much is it?

reply
rq1
5 hours ago
[-]
The next generation will include another processor to offload the inference from the RISC V processors used to offload inference from the host machine.
reply