This is the classic EU thing - the much superior most rigorous program - that no one has ever heard of, that turns out products that are rebadges or clones of other stuff. China has been getting more and more impressive. These EU "experts" are literally rebadging Chinese open models. 15 years ago that would have been a wild concept. China is totally OK fast following with no shame. SpaceX clone, they are getting to re-usability faster than the EU etc.
Except that no, it doesn't churn "clones"; MVC is quite a serious company and working on far more interesting stuff than whatever shit the SaaS or ads company you're statistically likely working at is.
EIC churns out DeepTech, invests 2.5M nondilutive with funding horizons of 10+ years and larger equity-based programmes too. Multiverse is a quantum computing company working on energy-efficient AI algorithms. The EIC is the reason companies such as ICEYE (https://www.iceye.com/) can exist. The chips running all that inference going in those US datacenters being built? 99% certainly european chips, almost certainly EIC-funded. (Eg. Hailo, Axelera, ..)
You haven't heard about the EIC because they don't care about marketing to americans; their target market is European companies, not you. But sorry to say, the talent is in Europe; the only reason you've been getting that talent in the US is a better business ecosystem, and your current administration has been utterly destroying this while our leaders have been actually fixing our issues. So keep shitting on Europe from the top of that crumbling hill, we'll see you when you emigrate.
Which states that this is a 40 TOPS int4 PCIe device, not what you'd put in a datacenter and certainly not what can run a frontier class model (or the model above). Embedded devices for inference are really cool! But that's the opposite end of scale for what goes into a data center.
I was sharing the EIC backed ones, but it's fair that in a datacenter it's more likely to be Axelera (or vsora/semidynamics if you start digging into earlier-stage competitors). Nvidia is the only American company selling AI chips to datacenters at scale but they're a huge supply chain risk so it's most DCs don't do single-supplier either way.
You'd need to compare by volume. And when you do that - unfortunately - NV is by far the largest supplier. 'Most DCs' are effectively single-supplier. And I wish it weren't so. This is the hard one to crack, NV needs credible competition. AMD and Intel are the only two majors that potentially could do this but they're both way behind where they should have been today.
I remember doing my first bit of 'shader repurposing' a long time ago thinking 'wow, I wonder whether Intel realizes this graphics card is the more powerful processor in my computer'. Back then (in the pre-CUDA) days you had to contort yourself in the worst possible ways to get any kind of unintended computation done but when it worked it was amazing. Since those days NV has been pumping enormous effort into building out that miniscule head start that they had in the days of 'Brook' and the subsequent release of CUDA. It is going to cost a lot of money to duplicate that, and even more to compete with NV on what is now their home turf. But I really wished someone with very deep pockets would do just that.
- Return experience is TERRIBLE. I'm not kidding, with Amazon you've got one click to a QR code and a UPS store dropoff. Some of these mfgs you are jumping hoops for a week plus! And then have to box, buy a label, often pay for shipping and more. IF you can even get in touch with someone.
- Shipping experience is TERRIBLE. How does FedEx stay in business? I'm serious - the express port of their name is a joke. Stuff will randomly get stuck in a warehouse for a week. I've had their call center tell me that for SURE it will be delivered x date (because the online tool shows that date) but the package is still out of state at 9PM. So they'd need to get it in state, then to distribution center and then to a truck to my house by midnight - surprise surprise that didn't happen.
- I've gotten used products from the mfg? Do they get amazon returns back and then try to ship direct with those? How does this work that the new product is under the amazon.com seller and the mfg has the USED?
- You get on more mailing lists going direct. ULINE and friends now ship me these huge catalogs following tiny tiny purchases. Catalogs are still a thing!
I recently bought a camping tent from Naturehike, a budget but respected Chinese brand. I ordered on their website and, long story short, my parcel got stuck at a warehouse for weeks, presumably lost. They said they can't refund me before they retrieve the tent and "inspect the package", as if I somehow went to a warehouse in a different country, found it there, and damaged it.
It took me well over a month of back and forth to get a refund, and only after I repeatedly threatened them with a chargeback. I ordered the exact same tent on Amazon and had it the next day. As a bonus, it was now discounted on Amazon but not on the manufacturer's website.
I only order from Amazon a handful of times a year when I can't find an item elsewhere, but manufacturers are really doing their best to push me towards it.
Yeah, Amazon has disadvantages (counterfeits, fake reviews, more expensive) but its advantages (insane shipping speed/cost, strong return policy) are nearly impossible for competitors to compete with. The manufacturer is not going to offer same day delivery and no-questions-asked returns.
It is also often more expensive than Amazon, which is I think is largely due to Anazon's anti-competetive, and possibly illegal "most favored nation" terms, but as a consumer it does discourage avoiding Amazon (which is why Amazon has those terms).
The security posture at AWS is different. AI startups are going to get hacked and leak data etc. All the startup webapp builder tools, vscode plugin players etc.
AWS could still be hacked, but they've taken some care to make it a bit less likely, a bit easier to track which customers affected etc. If you dig into AWS logging for example, there is a TON if you turn it on, you can really go back and see who did what to the permissions / environment etc. I imagine they've got pretty good logging of their staffs access to things as well. I had to jump through some hoops once to have their staff on my account.
You really don't understand what AWS offers if you think this is what is getting them workloads (including competitors and highly sensitive govt workloads).
Running postgresql is an order of magnitude more complicated than sqlite.
130k tps even with unlogged is not always super easy especially if getting hit concurrently. Postgresql connection overhead alone can be pretty brutal if you are setting up and tearing down connections or have 1,000 writers etc.
Postgresql generally requires good network connectivity. Folks doing sqlite distributed tend to have everything independent, you literally don't need to worry about connection / security / firewall / permissioning / internode escape or data leaking etc, can even have problems in local side networking and services can still serve.
That was per container, with 16 containers per data center, so would be a lot of DBA tickets to get something that large; SQLite scaled with the horizontal scaling of the app; and we did have a flaky network - something like one in 100,000 tcp connections would fail. And occasionally the whole network would just go away for a number seconds. And the persistent container storage was managed by the same storage team that managed storage for the DB team, so base scalability and availability high.
It seems to give the booster a real kick - what's that do to turbo's and fuel movement?
You've got hot exhaust onto cold cryo fuel tank header?
You've got to carry more mass in terms of protection for the tank?
Is doing MECO and then push and then get 100 yards apart or something before second stage / ship engines kick on a big enough penalty to justify all the extra complexity?
IIRC, it basically removes the need to do a handoff between the relatively tiny tanks that can reliably pump fuel and oxidizer in zero-g from the main tanks that struggle with that in starship, and completely removes the need for any of that on the booster.
They were originally planning to not do hot staging, aiming instead for a somewhat funny approach where they spin the ship and booster slightly so they are passively separated by centrifugal forces. A bunch of things went wrong in the first test flight, so this was never attempted, and they switched to the "simpler" hot staging in flight 2.
Does anyone know what is up with RDP support on linux? I'm trying to migrate to linux but I need to be able to RDP to a headless machine running my desktop from Windows machines. How is this not solved? Is wayland worse or better here?
Accessing a Linux machine from Linux/Windows via RDP is not fun.
Accessing a Windows machine from Linux/Windows via RDP works excellent - I use Remmina on Linux, but there are ofc lots of alternatives, as usual.
Points on accessing a Linux(Fedora/KDE Plasma) machine via RDP:
- as I understand it, you cannot open a new session, you can only access an existing one -> forget about a headless machine, it will have to render its DE into to void if you want access it via RDP. The work-flow is more like VNC than RDP.
- X11 has problems, Wayland is definitely worse. Queue the people who will tell me/you that it works fine them. My last attempt on Fedora ended with a "working" setup. Working in quotes, since I had to accept/allow every incoming connection on the host machine (in a pop-up window which auto-hides after a few seconds and did work ~60% of the time), making it useless for the intended use case. You can workaround this by SSH'ing into the machine and accepting the connection somehow, but I gave up at this moment.
- there is also some fun to be had regarding display resolution and "session passwords", but compared to the fun with Wayland "security" and portals, its manageable
If they are going to be the most broadly adopted AI platform where does that leave nvidia?
What is the AI PC platform? The experience on windows with windows 11 for just the basic UI of the start menu leaves a lot to be desired, is copilot adoption on windows that popular and does it take advantage of this AI PC platform?
Ryzen AI 400 mobile CPU chips are also releasing soon (though RocM is still blah I think)
Nvidia is still playing in the AI space despite all the noise of others on their AI offerings - and despite intel hype, Nividias margins at least recently have been incredible (ie, people still using them) so their platform hasn't yet been killed by intel's "most widely adoptoped" AI platform offering
