I think you are overestimating power usage by order of magnitude or two. For a much bigger Deepseek model, you can get <0.1J/token[1], or 36M output token for 1 kWH.
Or more than 500 book worth of token for 1 hot shower.
Interesting. Is that just for inference or does that also amortise the energy used to train the model across the expected number of tokens served over its useful lifetime?
Which non US government are you talking about. China is investing heavily and making at least some progress. EU couldn't get "legacy rocketry programs" working, so the reason is incompetence not stubbornness. India is trying a bit, but they don't have capacity to spend. Russia is also trying.
DeepSWE was released on May and Deepseek v4 flash was released on April. By definition it couldn't be benchmaxxed on this particular benchmark. My guess is the mutation is not like real world and it is tripping small models, not benchmaxed model.
It's not clear which v4 flash they used. There are 3 different releases for that one.
Edit: I think that's the answer. This page lists the score as 53.3% which is close to the 54.4% reported here https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731 for the 0731 update. The original from April was called preview, doesn't list the score, but in the comparison with the update it scores 7.3%.
The actual vulnerability got very little discussion. It's clever. Because Linux networking people are obsessed with checksum performance, virtio-net plumbs checksum offload metadata through the host/guest driver interface. Guests can thus deliver the host packets along with instructions on what to checksum. The vulnerability is a missing sanity check that allows checksum instructions to instruct the host to overwrite the header length field in the IPv4 header, rather than the checksum field.
Crucially, this happens in the middle of packet processing, so the kernel code is forced to work with two different ideas of how long the IPv4 header is. All hell predictably breaks loose.
Additional color: per the commit, the bug was found by Paulos Yibelo, who runs PwnAI (the author of this piece), and worked on security at AWS for 4 years prior.
reply