Some interesting changes.
Next year will see some truly monstrous compute projects get underway as the AI boom enters its third year. Among the largest disclosed so far is xAIâs plan to expand its Colossus AI supercomputer from an already impressive 100,000 GPUs to a cool million.
Such a figure seemingly defies logic. Even if you could source enough GPUs for this new Colossus, the power and cooling â not to mention capital â required to support it would be immense.
At $30,000 to $40,000 a pop, adding another 900,000 GPUs would set xAI back $27 to $36 billion. Even with a generous bulk discount, it still wonât be cheap regardless of whether theyâre deployed over the course of several years. Oh, and thatâs not even taking into account the cost of the building, cooling, and the electrical infrastructure to support all those accelerators.
Speaking of power, depending on what generation of accelerators xAI plans to deploy, the GPU nodes alone would require roughly 1.2 to 1.5 gigawatts of generation. Thatâs more than the typical nuclear reactor â and the big ones, no less.
Lenovo may launch a Legion gaming handheld powered by SteamOS! There have been image leaks of the console with a Steam logo button. They are expected to announce it on 7th January at the CES. They also have SteamOS and Steam Deck co-designer as their special guest for the event.
âAlthough LLMs struggle to create malware from scratch, criminals can easily use them to rewrite or obfuscate existing malware, making it harder to detect,â Palo Alto Networks Unit 42 researchers said in a new analysis. âCriminals can prompt LLMs to perform transformations that are much more natural-looking, which makes detecting this malware more challenging.â
Great revisionist history there @Colin. First it was the PC Junior that got trounced by Apple. Second after Jobs and companie dropped the ball on a windowed system Gates ran with the idea after giving Jobs opportunity after opportunity to utilize the tech.
The daily IT madness - today from the category âdumber is always possibleâ:
Everyone gets the Christmas present they deserve - including Asus and Windows users 
GEE makes me very grateful I switched to Gigabyte for this build.
I wouldnât call it revisionist historyâŚ
The most surprising takeaway from todayâs end of year exploration was seeing the Linux kernel hitting a decade low for the number of new commits this year. But not all is bad as on a line count the annual metric is comparable to more recent years.
If the device has reached end-of-life and no longer receives security updates, it should be replaced with a new model.
A a general advice, you should replace default admin credentials with unique and strong passwords and disable remote access interfaces if not needed.
they are doing so much to drive us away, why are we still here??
Good!
I was talking about an interesting thing about AI the other day. How much pollution in Academia is there knowing that AI is pumping out fake documents? Before 2020, you never seen any discussion or policies changed in Universities (at least Canadian universities) till now. Now Chat GPT can come out with what ever you ask of it.
This makes me wonder, if you look at all the military technology thatâs being developed behind closed doors, there is no telling how advanced the technology really is. We in the public get the table scraps once they are done with what ever they are doing. So imagine something like Chat GPT behind closed doors, way more advanced then it is today writing up academic papers. It would be possible to go through the so called âpeer reviewed processâ without anyone so much as blinking an eye that it could be a faked document.
Sure, the peer reviewed process isnât something to write home about. Itâs not immune to political interference or financial self serving interest. But, if you have AI making ligament documents, with proper references to other documents and books, it could be so easy to pass a fake document through the proper approval. To the untrained eye, all the flaws could go completely unnoticed. To me, this begs the question, how much AI pollution is really out there that we donât know about? How many people have written papers, research studies, thesisâs or books based on work that was complete nonsense right from the beginning that nobody ever thought to ask? And, if you have bad research data out in the world, what is to say this doesnât have a part to play in inflating the egoâs of people so heavily involved in the profession they studied ( for example, such as physics, or anthropology to name a few) which holds back proper advancements in the field? Like a form of collective mind manipulation.
EXACTLY why ALL the big companies are pushing so hard for AI cause an intelligent workforce isnât a slave workforce.
this sounds like something my boomer would say