CalmStorm

Huawei claims that they can reach 1.4nm by 2031. Does that mean EUV is no longer required for this process?

For the first time, it introduced native sparse attention into the full training process, achieving up to 11× inference speedup while maintaining model performance.

LLaSA is a simple framework for speech synthesis that employs a single-layer vector quantizer (VQ) codec and a single Transformer architecture to fully align with standard LLMs such as LLaMA.

I don’t quite understand why OpenAI would pay so much when there’s a solid open-source alternative like Cline. I tried both, and feel that Cline with DeepSeek v3 is comparable to Cursor and more cost-effective.

If it achieves its goal of both “safe” and “super intelligence”, it may have commercial value. e.g. enterprises may want to use it instead of, e.g. OpenAI’s ChatGPT. Think of recent AirCanada’s chatbot law suit.

One of the most popular HSM is Thales Luna Network HSM, which can perform 20,000 ECC operations per second [1]. Even with the size of Azure AD, Microsoft may not need a lot of HSMs for signing purpose. HSMs are not particularly easy to manage though, maybe that is one of reasons they are not used as much as they should be.

[1] https://cpl.thalesgroup.com/encryption/hardware-security-mod...

For modern CPU, is microcode still used for division or there are better and faster alternatives?

Here is the URL: https://www.jpost.com/business-and-innovation/tech-and-start...

The article doesn’t mention whether SQLite WAL mode is on. That could make SQLite significantly faster: https://sqlite.org/wal.html

Oracle’s ownership of MySQL AB seems to have negative impact on its development. My concern, which is also raised in the article, is whether similar things will happen to Postgresql.

What programming languages does it support? I couldn’t find out from the press release.

It is helpful to know the internals of SHA-2. It helps to make sense of the "Length Extension Attack" [1] on this kind of hash function construction.

[1] https://en.wikipedia.org/wiki/Length_extension_attack

The social media walled garden - quite a few people I know rarely use a browser or don't even know how to use it. They get their news mostly from their social media apps, and they have a hard time spotting fake news. It is much easier to censor content in a social network than the open internet. People live in their bubble and are not exposed to different opinions and ideas.

Trying to appreciate mathematics without understanding its inner workings is like reading a description of Beethoven’s Fifth Symphony instead of hearing it.

I personally feel J.S. Bach would be a better metaphor here.

This article [1] describes the features in 1.15. The new linker is quite impressive in reducing executable size because it reduces more unused code:

For more realistic programs, binary sizes go down by 3.5% or as much as 22%.

The original linker was written by Ken Thompson, who is now retired.

[1] https://lwn.net/Articles/820217/

In Unix:

    a; b    # Execute command b after a. The result of a is not used by b.

    a >> b  # Run function b after a. The result of a is not used by b.

    a | b    # Execute command b after a. The result of a is used by b.

    a >>= b  # Run function b after a. The result of a is used by b.

Unix pipelines are indeed beautiful, especially when you consider its similarity to Haskell's monadic I/O: http://okmij.org/ftp/Computation/monadic-shell.html

Unix pipelines actually helped me make sense of Haskell's monad.

Preventive measures may be cheaper than enforcement. UBI could reduce the crime rate and therefore save money on law and order. I think it targets the root cause rather than fixing the symptoms through more policing.

Deno's sandbox security is somewhat similar to Mandatory Access Control (MAC) implemented by SELinux and AppArmor. But it looks like not as fine-grained as MAC. In the example:

deno run --allow-net myWebserver.ts

With SELinux, one can specify the port range and network interface that the application is allowed to access. It also provides audit log that can be examined by the admin. Maybe there is no need to reinvent the wheel but just use some form of MAC if you really care about security.

The other common misconception is that the price of bitcoin is determined by the hashrate. It is actually the other way around, a higher price drives up the hashrate, because a higher price means more demand for bitcoin.

Some founders of altcoins apparently don't understand this and try to buy hashpower to artificially increase the hashrate, hoping to increase the price of their coins. If there is no demand for their coins, an ASIC farm is no different from a Rasberry Pi.

Swift does not allow null pointer, and therefore there is no null pointer exception. This is a very nice feature, especially compared with other server-side compiled languages like Golang and Java. Swift is modern and concise, I hope it takes off on server-side.

One important weakness of zkSnarks is that it requires a trusted setup, for example [1]. A new alternative is called zk-STARK [2], which doesn't require the trusted setup, and is post-quantum secure. However, it significantly increases the size of the proof (around ~50KB). In general, hash-based post-quantum algorithms require bigger size and it would be interesting to watch the progress made in this regard.

[1]https://filecoin.io/blog/participate-in-our-trusted-setup-ce...

[2] https://eprint.iacr.org/2018/046.pdf

If there are two valid, signed transactions that spend the same unspent input transaction (UTXO), only one of them will be included in the block. This is indeed guaranteed by the protocol [1].

[1] https://bitcoin.stackexchange.com/questions/81624/can-two-or...

In case there are multiple conflicting transactions (for double-spending purpose), the network will only accept the first one and reject the rest. So the micropayment processor needs to wait for a few seconds when it receives the first transaction in the mempool. If no other conflicting transactions arrive, the first transaction will be acccepted. These few seconds give the first transaction big advantage over the double-spending ones, as it will most likely be included in the block.

If the attacker sends the legit transaction and double-spending one at the same time, it will be detected due to the few seconds wait by the payment processor.

Since it is a micropayment, it is possible (or acceptable) to accept payment with zero confirmation (in less than 10 seconds), as pointed out by Satoshi in the "Bitcoin snack machine" post: https://bitcointalk.org/index.php?topic=423.msg3819#msg3819

Yes, there is a risk of double-spending, but since this is a small amount, the cost of the attack may be higher than the benefit. The payment processor can also deploy more full nodes in different geographic locations to monitor and detect potential double-spending transactions in the transaction memory pool.

I have been working on this decentralized key-value database: https://github.com/kevacoin-project/kevacoin Together with W3C's draft Decentralized Identifiers (DID: https://www.w3.org/TR/did-core/), it could provide a decentralized alternative.

Not sure what is the best way to verify Twitter/Github account though. This has to be managed by users themselves. E.g. one user posts a proof in the Twitter account, the other user verifies the proof by checking the proof against the public key posted in the database.

I have been working on this decentralized key-value database: https://github.com/kevacoin-project/kevacoin

Together with W3C's draft Decentralized Identifiers (DID: https://www.w3.org/TR/did-core/), it could provide a decentralized alternative.

Not sure what is the best way to verify Twitter/Github account though. This has to be managed by users themselves. E.g. one user posts a proof in the Twitter account, the other user verifies the proof by checking the proof against the public key posted in the database.

Edit: updated description.

It is mentioned in the article that the digital currency supports offline transactions. How can this be done with blockchain, which required up-to-date information to validate transactions?

I am working on a key-value decentralized datastore on blockchain called Kevacoin. Some people are using the datastore to record the events and news related to COVID-19.

But you also need to be careful. The slow password hashing can be exploited for Denial of Service attack. You may want to rate-limit the APIs that use the hash algorithm.