Posts by Fido

Struggling to heat your home? How about 500 Raspberry Pi units?

Re: Most stupid use of RPs from A to Z

I had a friend who for a few years heated his home doing contract work involving image processing of flyover data using a cluster of x86 servers. It was effective because the heat and computing were both needed in the same place.

I'm a Raspberry Pi enthusiast since the first model was released more than 10 years ago. I think the questions are interesting what sort of cloud computing could be done with a cluster of 500 compute modules, how is the network fabric is arranged inside the boiler, what are the failure modes, redundancy and what kind of maintenance is possible.

As an experiment I set up a tiny cluster of 12 Pi computers each with no local storage. They mount their filesystems over iSCSI from an x86 server. It works well, but I have trouble imagining that scaling to 500 devices would result in anything but reliability problems.

On the other hand, maybe some magic is possible with compute modules where one can use more of the PCIe on the SOC for high-speed communications than possible on the standalone Pi. From an engineering perspective it would be interesting to know how much local storage is available inside the boiler and what kind of bandwidth there is in and out of a single node.

LockBit's new variant is 'most dangerous yet,' hitting Windows, Linux and VMware ESXi

Whack-a-mole Gone Wrong

It's possible taking down the malware servers without catching or punishing the criminals may encourage rather than discourage criminal behaviour.

In related news, crypto-locker victims are surprised to discover their backups were intentionally compromised by the baddies before the ransom demanded.

While I'm very much in favour of frequent incremental backups using the snapshot features of modern copy-on-write filesystems, keeping administrative access to the backup server separate from all the other IT infrastructure is impossible unless the only way to log in to the backup server is through a physical VT52 terminal. Unfortunately, VT52s were in short supply when in full production during the 70's and they're even more difficult to find these days.

Careless engineer stored recovery codes in plaintext, got whole org pwned

The graphic image for this article shows a paper notepad in which the recovery passwords are stored in plain text offline. When the only time one needs the recovery password is in a disaster scenario where the online systems are unavailable is that best practice?

Personally, I'm in favour of locking the box with the key inside.

I'm out, says OpenSUSE: We're dropping bcachefs support from next kernel version

For me the main point of an enterprise-level Linux distribution is that it provides a stable maintained environment upon which to build long-lived applications that are too valuable to constantly port to the latest software fashion.

According to the late dog developer, such an operating system might prove so popular as to risk future antitrust proceedings, so all providers of enterprise software either monetise such software to the point it is unpopular or destabilise it with disruptive changes.

From this point of view VMware's price increases and resulting loss of customers can be seen as a clever ploy to avoid future antitrust action under US laws, but that's a different story.

Could it be possible the dog developer was right and IBM is monetising Redhat while Suse is desperately trying to avoid a rapid rise in popularity that would risk it becoming a monopoly?

Fork that: Three alternative kernels show devs don't need Linux

DragonflyBSD

The fact that Linux started out as a hobby operating system and became useful is an exception not an expectation.

In my opinion a similar case in history is the DragonflyBSD fork of FreeBSD. The main result was another hobby operating system that occupied the time of some excellent developers whose efforts may have been appreciated more without the fork.

On the other hand, it seems likely good developers might have a need to write some code just for fun.

VMware to lose 35 percent of workloads in three years – some to its friends at ‘proper clouds’

Re: Cloud Hosted Versus On Premises

It's still renting if you own the hardware and pay a subscription for the software.

If open source didn't lack those business dinners with the sales team, wine and cheese, then nobody would consider moving to another vendor's lock-in solution.

Junk is the new punk: Why we're falling back in love with retro tech

I recently acquired a new car. I miss many things, but most of all I miss being able to turn the volume of the radio down without having to wait for the entertainment computer to boot up first.

When the car is started the radio will immediately begin playing at whatever volume the previous driver left it at and the volume can't be changed until the computer has finished booting.

Boy riding bubble realizes what he's on, asks for more air

People are impressed that computers can now have interactive conversations like in the movies.

Online retail didn't go away when the dot-com bubble burst. The chatbots and code pilots will not go away when the AI bubble bursts. The technology is real while a bubble is a financial concept related to get rich quick schemes. It's easy to agree with Sam Altman on this obvious point.

On the other hand I do not think lack of GPUs will precipitate the crash but rather unplanned for increases in the cost of electricity and shortages. If battery powered cars cease being a rich fashion statement and go mainstream, then the bubble-popping electricity shortages will happen sooner.

Did the tulip market ever recover?

Linux is about to lose a feature – over a personality clash

The cow that won't eat your data

I found the slogan "The COW filesystem for Linux that won't eat your data" funny and brilliantly juvenile at the same time. Yes, I know not everyone has the same sense of humour.

For me "The COW filesystem for Linux that won't eat your data" is a pun based on how much a real cow eats every day in contrast to how difficult it is to create a bug-free copy-on-write filesystem with live snapshots, compression, redundancy and data integrity features.

Humour is so difficult many people don't even try; yet a smile and laughter are so essential to human health that in spite of a cancel culture aiming to eliminate people for any mistake, there are people who still try to make jokes. Maybe that is part human nature.

Anecdotally I've had more trouble with thinly-provisioned LVM volumes than BTRFS. Of course everything was backed up and the failures can be attributed more to my misunderstanding than bugs in the code.

I can understand the need for a team that gets along and works well together. At the same time I hope the development of Linux remains based on technical merit.

Beijing doesn't want Nvidia's H20s anywhere near sensitive government workloads

In response to "nerfed H20-style GPUs must feel like an insult to them (a dish of someone else's leftovers), when they have access to homegrown CloudMatrix 384" my understanding is DeepSeek overcame many of the communication bottlenecks caused by lack of capable NVLink by creating a background communications task on the H20 using PTX assembler. Note also that even with the most capable hardware there are so many other bottlenecks that reduced TOPS and FLOPS with the H20 may not be very significant.

If the H20 were widely available and noticeably cheaper than the H200, I think it would attract a large number of customers worldwide--including universities for teaching and for use in budget-limited research projects. Unfortunately, if the cost of manufacturing an H20 is not much less than an H200, then the price will not be much cheaper.

Perplexity takes a shine to Chrome, offers Google $34.5 billion

Whether Perplexity or some other company buys Chrome, the result will be similar to when Oracle bought Solaris (by way of the Sun acquisition) or when IBM bought RedHat: What was open source ceases to be so within a couple years.

In my option the benefits resulting from companies who have built products based on the open source web rendering and JavaScript engines created by Google far exceed the advantage of letting a third party monitise Chrome by making its code closed source.

Air Force buying two Tesla Cybertrucks so it can learn to destroy them

In my opinion, terrorists and smugglers using actual Cybertrucks are probably not the main focus.

To me the fact that electric vehicles have different storage, maintenance and stealth characteristics compared to internal combustion engines suggests they might be useful in actual conflicts. Since hostile foreign governments are unlikely to loan their secret battery-powered military vehicles for testing by the US Air Force, a Cybertruck is the next closest thing.

Rampant emoji use suggests crypto-stealing NPM package was written by AI

If the code was written by Claude, could the police arrest Claude? Alternatively ask Claude who done it or at least check the log files.

UK VPN demand soars after debut of Online Safety Act

Not anyone has said it, but the content being censored for children is not good for adults either.

Encouraging private companies to scoop up a large number of personal identifications could go wrong. The problem is not fake information being used to obtain access to content but real information later being used for identity theft. For example, consider a foreign government that prints fake ID cards based on the information uploaded for age verification in order to scam democratic elections.

Congress tries to outlaw AI that jacks up prices based on what it knows about you

In places where produce is bought from vendors on the street setting a different price for each customer, discrimination based on how you look and talk is common. This contradicts the ideal of equal opportunity for every person independent of age, race, ethnicity, religion and gender. If an AI does the same thing online, the result is the same except at a much wider scale.

From what I can tell the use of personal information to guage how much value a buyer places on an item is likely equivalent to the outcome of a VCG auction for a divisible good. Each buyer pays a different price so that their respective marginal valuations at the transacted quantities are the same. Thus, someone with greater need is charged more.

This optimises the value of the market; however, common sense says people are being charged unfairly. At the corporate level it's been happening for a long time. That's why corporations have purchasing departments. With AI the same value-optimised pricing can be implemented at an individual level.

From a political perspective the result is similar to "tax the rich" but with an unpredictable AI-identified category in place of rich. Ethically it's wrong to "do bad things" to any identifiable group.

Nvidia extends CUDA support to RISC-V just in time for next wave of Chinese CPUs

Does Nvidia not support IBM Power anymore?

Not too long ago Summit and Sierra were top 10 supercomputers. Maybe that's why open-source ROCm looked so attractive during the last procurement.

AI can spew code, but kids should still suffer like we did, says Raspberry Pi

I don't see mentioned in the comments or article the fact that the company which makes Raspberry Pi computers is separate and independent from the Raspberry Pi Foundation being reported on. This may be obvious to many people, but I mention it anyway.

From my point of view the job market for programmers and software engineers has historically expanded since the 60's due to the everywhere increasing use of computers. Then suddenly AI makes writing software so efficient that even though the everywhere use computers still expands, the job market for programmers and software engineers contracts.

The question boils down to how fast the need for new software continues to increase versus how fast developer productivity increases due to AI. From my point of view, productivity increases will be so great the job market is likely to contact.

People make analogies about how the job market for skilled machinists didn't grow with the automobile market due to assembly-line productivity increases.

Here’s what it’ll take for Nvidia and other US chipmakers to flog AI chips in China

Re: Newegg pricing today

AMD and Nvidia pricing for datacenter accelerators is similar. Anything with 100GB or more is super expensive.

Right now Nvidia is the primary development target for all accelerated computing. Conversion efforts are needed for other hardware.

Software developer mindset is so important to the success of the hardware that AMD open sourced their software stack while Nvidia is willing to sell poorly-performing hardware in China at whatever price is needed to maintain marketshare.

I suspect the resulting supply of cheap Nvidia hardware in China will increase availability of GPU accelerators to software developers and allow the Chinese AI software industry to move even faster than the hardware.

Intel's latest CEO Lip Bu Tan: 'You deserve better'

Maybe 1TB RAM is better.

To clarify, what I'd like is for Intel to create a competitive offering in the niche characterised by the AMD MI300A unified CPU GPU. That requires more than a 4-bit TPU. However, a fully scaled up variable-precision matrix unit sounds good.

At least the new CEO is not talking about zetascale supercomputing next year. At the same time the world does not need any more low-performance high-core-count CPUs for running web servers. Instead it would be nice to have an affordable processor that could run a 500 billion parameter LLM locally at reasonable cost and speed.

To me it seems likely AI combined with traditional numerical computation will form a sum greater than the individual parts. For this reason I'd look forward to huge on-package matrix accelerators with unified HBM memory and selectable precision from 4 to 64 bits. If the promise is for Intel to listen, then one can dream that the voice of one user might be heard.

VMware distributor Arrow says minimum software subs set to jump from 16 to 72 cores

At the university here we repurposed the old scientific computing cluster as a cloud using VMware. The nodes each have a total of 32 cores so I'd naively expect a 72/32=2.25 increase in licensing costs.

Since we're also looking at a possible 10 percent general budget reduction due to lack of planning, it seems possible the solution will be to turn everything off but not back on again. For educational institutions Microsoft offers extended Windows 10 support at $1 per seat for the first year doubling each year after that. Will Broadcom offer similar discounts on legacy VMware for schools.

I'm not involved in the IT decisions, so my only problem is as a user of the cloud. Hopefully the discounts are worked out and operations will continue normally.

Signalgate storm intensifies as journalist releases full secret Houthi airstrike chat

I'm interested only in the technology aspects of Signal, not the specifics of this case.

If a hostile foreign government were able to compromise the mobile phone networks of a targeted nation, would it then be possible to secretly install remote access software on a government issue phone?

To avoid giving away the fact that end-to-end encryption is compromised, a foreign spy might use the remote access software to add a third party to a private chat who is likely to leak the information. When the third party leaks the private information, the foreign government does not get blamed yet benefits from the leaked information.

I don't know how real spy operations work. However, it seems to me that this technique could be effective at circumventing end-to-end encryption without revealing who did it.

Strap in, get ready for more Rust drivers in Linux kernel

Somehow the disadvantages of having two distinct programming languages in a single code base has to be outweighed by one of them being Rust. Having to maintain separate bindings for each language and synchronise them by hand sounds more error prone to me than the danger C code in production for more than a decade possesses a hidden use after free, bounds error or null pointer dereferences. At the same time, I'd get very excited to see a nontrivial GPU, network or USB device driver written in Rust that avoids the unsafe keyword.

A year ago there was a US government directive to stop using C and switch to memory safe languages for essentially everything. While the directive looks as well thought out as the long since abandoned idea of using ADA for everything, it's possible adding Rust to the Linux kernel is based primarily on the need to demonstrate progress being made towards use of memory safe programming practices.

Such a benefit, though nontechnical in nature, could be enough to offset the interoperability and maintenance disadvantages of two languages in the same project.

The same considerations could be the main motivation behind both Microsoft's and Google's reported progress on the adoption of Rust.

Linus Torvalds forgot to release Linux 6.14 for a whole day

As someone who has been enjoying Linux since 1993, I just want to say thanks to Linus.

SoftBank buys server-grade Arm silicon designer Ampere Computing

Re: What

My impression is the Altra, Altra Max and One processor designs are slow compared to recent Graviton, Intel, Epyc and likely the next generation Loongson processors. Although slow is not an ideal feature during the transition to compute intensive AI, the conjecture I've heard is Ampere has an engineering team with enough experience to quickly integrate a massive TPU into existing ARM designs.

I think at this point it's clear AI is not a fad that will fade like skinny jeans, but a tool which greatly increases productivity for workers in certain technical and creative fields.

When the dot-com bubble burst, online shopping didn't go away. Neither will the need for economical AI inference go away if the AI bubble breaks. Whether Ampere turns out to be a valuable acquisition for SoftBank is a different matter.

Photoshop FOSS alternative GIMP wakes up from 7-year coma with version 3.0

I'm hoping everyone else's 10 percent doesn't make it even harder to find the 10 percent that I use.

Ubuntu 25.10 plans to swap GNU coreutils for Rust

I think the latency--time between pressing the key and a program recieving a character--was greater with a USB keyboard. Since it makes no difference when playing Nethack or Dwarf Fortress, never mind.

I find it an interesting question whether a non-GNU licensed version of GNU core utilities would eventually become a closed-source proprietary component of Ubuntu Linux. I don't have any insight into Ubuntu's roadmap, but even if I did, the management can change or the company be sold.

It's been a few years since I experimented with Rust. Back then it did not have a shared library runtime. If that's still the case, each command in Rust core utilities might load hundreds of thousands of extra bytes. The resulting slowdown may be difficult to measure with micro-benchmarks; however, the large executable sizes might thrash both disk and CPU caches.

HP Inc settles printer toner lockout lawsuit with a promise to make firmware updates optional

Where I work the solution was a policy that prevents nearly anyone from buying any type of office printer. Now there are a couple networked copiers in the mail room for everyone to use. Thanks HP for the additional exercise.

Apple's alleged UK encryption battle sparks political and privacy backlash

If I understand what end-to-end encryption is, there is no way to add a backdoor without an observable change to the software running on the user's device. As criminals and spy agencies constantly reverse engineer software updates to look for security related patches, actually adding a backdoor to UK phones would be a much more noticeable canary compared to making no changes in the code that implements the encryption.

From this point of view the least malicious form of compliance could be disabling the feature rather than spelling out in code exactly what was asked for.

Apple has locked me in the same monopolistic cage Microsoft's built for Windows 10 users

Re: So why does Apple get less stick than Microsoft for their planned obsolescence blackmail?

My impression is people who buy Apple trend not to be as concerned about money compared to those who buy a Windows PC. Since saving money is not an objective, the Apple customer is not unhappy when older kit is obsoleted because that provides a reason to get the latest greatest.

Here the university has been purchasing new iPads, iPencils and portfolio keyboards for every incoming student since the epidemic. I tried to figure out how students could use this technology for programming tasks related to numerical computing and computer science. The Apple point of contact recommended Swift playgrounds.

Swift playgrounds features a 3D rendered version of Karol the Robot for toddlers but is actually much more: One can also write, compile and run real programs. Unfortunately the stack size is small, no optimiser is available and I could not find any scientific libraries.

Today it appears the main purpose of the iPad is to create a cyberpunk version of a medieval university where the students use iPencils to transcribe the text being lectured on as it is written in chalk on a blackboard. Traditions die hard.

On a modern note, the iPads do satisfy the digital equity objective of providing underprivileged students the same ability to TicToc, Facebook and Instagram instead of studying.

Scientists create woolly ma-mouse by looking at mean genes from the Pleistocene

Global Winter

Rather than mammoths and mice equipped to survive extreme cold, wouldn't a species adapted to global warming be more practical?

Linux royalty backs adoption of Rust for kernel code, says its rise is inevitable

Re: The problem is

Since the original design of C was to eliminate as much assembly language as possible, if C were eliminated in favour of Rust, would that imply a need for more assembler in the kernel again?

If not, could the use of unsafe in Rust be more unsafe than writing the same routine in C for those OS things where unsafe is anyway necessary?

Fear of the unknown keeps Broadcom's VMware herd captive. Don't be cowed

Deploying commercial single-vendor IT infrastructure forms a long-term business partnership with a vendor that is generally based on trust. The difficulty occurs when a takeover or buyout changes the nature of the partnership.

A similar thing happens at smaller scales with blog sites and free software. The owner sells to a third party that subsequently monetizes the readership or inserts adware.

The difficulty is that further regulations could make things worse. Consider, for example, when SoftBank went in to ARM-architecture monetisation mode after not being able to sell to Nvidia. All sorts of things can suddenly change the nature of a business partner.

Since IT embeds itself so deeply into a company's business process, relying on single-vendor software does not provide the robustness needed for long-term planning. While decision makers with golden parachutes may not worry much about the future, protection of the business from disruption caused by changes in software licencing and development is an argument for open or at least owning the source.

Said another way, switching from VMware to another commercial offering is a fool-me-twice decision.

Someone is slipping a hidden backdoor into Juniper routers across the globe, activated by a magic packet

It's not cyber war yet...

When does the prepositioning end and the first cyber war begin?

Micropatchers share 1-instruction fix for NTLM hash leak flaw in Windows 7+

Imagine if you suddenly had a return of 1,000,000 on every $ or £ deposited in your bank account. The conclusion: what separates the high flying business person, diplomat and government spy from Homer Simpson is less than a factor of a million.

Given the way cyber crime scales once the malware is written, it is not unreasonable to imagine millions could be automatically targeted by a known but unpatched vulnerability. From this point of view, it's not possible to disappear into the multitude. Since it takes more insignificant users for a significant effect, the insignificant might be more likely to be targeted.

T-Mobile US CSO: Spies jumped from one telco to another in a way 'I've not seen in my career'

Re: FIDO2

Since the telecoms carrying SMS based second factor authentication were already compromised by foreign agents, I can see how FIDO2 helped T-Mobile avoid being as easily hacked.

Said another way, if your adversary can already read your text messages, then not only do they know the second factor but you have also told them when and where you are logging in.

1,000s of Palo Alto Networks firewalls hijacked as miscreants exploit critical hole

My recollection is CrowdStrike pushed a configuration file--not even code--for their Falcon platform and put Delta Airlines out of business for a week. Others customers were affected for a total of over 5 billion direct losses.

Needless to say corporations who had opted out of automatic updates due to having their own mission critical review process for high availability weren't pleased that CrowdStrike pushed the update everywhere anyway.

As much as the security business seems overrun by clowns, if automatic updates are not supposed to happen then it's clear updates shouldn't happen automatically, even if the updates are good.

The Register takes AMD's Ryzen 9800X3D for a spin

Also for Science

While the gaming community may be larger than engineers who develop and run numerical codes on their desktops, the time stepping in a scientific simulation is quite similar to in-game dynamics but without the game. In fact, since a numerical run has no graphics-card bottlenecks, the X3D series CPUs with full-width AVX512 are astonishingly fast for small-scale computations.

Woohoo! If more science and debugging can get done faster on the desktop, that leaves leadership-class supercomputing available for larger runs.

AMD sharpens silicon swords to take on chip and AI rivals

Re: But AMD has clearly stated that they give up on GPU leadership

When I read the Tom's Hardware interview with Jack Huynh, AMD's senior vice president and general manager of the Computing and Graphics Business Group, I initially had the feeling that leadership was clueless and lacking.

After refection I decided trying to sell what the engineers made is better than promising zetascale compute

I think AMD unified the engineering behind Radeon video cards with the Instinct GPUs but badly presented this as deprioritising gaming.

Right now AMDs CUDA equivalent--ROCm--is not well supported on Radeon video cards. As games may soon include a nontrivial element of AI inference, getting ROCm to work on all AMD GPUs seems essential to me. Unfortunately, this unification seems to have led to the next generation Radeon cards not being competitive for high-end gaming.

Although a senior vice president can only sell what the engineers can make, I would have characterised the lack of a flagship gaming card as a necessary unification of two diverging GPU architecture so ROCm driven AI libraries can run everywhere.

Begun, the open source AI wars have

The expense of training a model is likely to go down, just as a Raspberry Pi is faster and cheaper than the original Cray supercomputer.

One important security aspect of regular open source software is verifiable binary executables. For example this makes it possible to check that unexpected changes weren't made to a program while creating the packages in a Linux distribution.

A typical software engineer does not participate much in verifiability; however, the ability to make modifications to the source goes hand in hand with the need for verifiable binaries in open source.

Given the even greater possibility of hiding things in generative AI, a class of verifiable models may be needed before people can safely use open source LLMs. While there are admittedly randomised elements in stochastic gradient descent and other aspects of training, requiring that the original training data be available is an important step towards verifiability.

Note also that the more expensive it is to train a model, the more important it is being able to verify the result.

Upgrading Linux with Rust looks like a new challenge. It's one of our oldest

The difficulty I see is that more bugs are likely when simultaneously using two languages in the kernel compared to using only one. Thus C mixed with Rust will have more security issues than only C, just because of the mixing, even if Rust might be better.

I also find it surprising that Redox OS has been in development for 9 years and still not available for production use on any cloud platform, but that's a different story.

Linux Deepin 23: A polished distro from China that Western desktops could learn from

The article states "Although some of the betas offered Wayland as an option, it's gone from the final release, which uses just X.org." If the Deepin desktop is as widely used as suggested, sticking with X11 may say something about direction of the window fracture in Linux.

As a contrast Raspberry Pi OS switched to Wayland by default last year. Though tied to a particular vendor's hardware, in my opinion that OS is also trend setting.

Broadcom ditches VMware Cloud Service Providers

Could letting Broadcom buy VMware be as bad for the industry as not letting Nvidia buy ARM?

The 15-inch MacBook Air just nails it

The article states "twenty-four gigabytes of Hynix LPDDR5 memory can't hurt" but I thought the M2 had some special kind of unified memory which was faster.

Maybe I'm confused. Is unified memory the same as LPDDR5? Is there a reference which indicates the RAM is made by Hynix?

As the Top500 celebrates its 30th year, with a $5 VM you too can get into the top 10 ... of 1993

Re: But can it run HPL?

My understanding is the Pi 4 gets around 11 GFlops while the Pi 5 reaches about 30 GFlops on the high-performance Linpack benchmark.

On the other hand, the price of a Pi is more than 5 dollars.

Police ignored the laws of datacenter climate control

Around here it is not rock dust but chalk dust which clogs the computers in the classrooms used to teach mathematics and other sciences. Things have improved but the mathematicians still make a lot of dust with all their equations.

Core blimey, Intel's answer to AMD and Ampere's cloudy chips has 288 of them

From what I understand, the efficiency core is not a performance enhancement but a setback that scales up to support more VMs while still meeting data center thermal and power requirements. As likely as not, eight to 32-core CPUs are chasing single thread performance and anyway don't have much trouble meeting the power limits of even the oldest racks.

Arm's lawyers want to check assembly expert's book for trademark missteps

Re: Scrap the book

I think a similar book written by the same author about RISC-V would be appropriate whether or not this takedown from Softbank ever happened.

In particular, RISC-V does not require any malice towards ARM to succeed. Since Linux, GCC and LLVM have reached the level where ARM is practical in the data center, so RISC-V and even Loongson are equally practical.

In the end it's about who has the better and faster hardware. ARM is causing trouble for companies with architectural licenses to innovate, see the Qualcomm Nuvia lawsuit, while RISC-V is so easy no license is required.