Yeah me too in 1982, using the Melbourne House Z80 reference, aged a young 10 years old. Working with POKE and no macro-assembler, I wrote mnemonics then translated them to machine-code by hand. A baptism of fire that to this day that I've not forgotten.
Ironically the power per cycle is decreasing - power and thermal dissipation are really the limits NVIDIA is exploring. It’s what the software does with those cycles that is leaping exponentially.
The other bottleneck they are studiously exploring and minimizing is beachfront area and networking/interconnect bandwidth.
Nvidia went all-in on infiniband serDES while AMD chose pcie/CXL. But since Pcie signaling requirements are tighter, you need bigger stronger PHYs, which means you get less actual area per beachfront. The penalty is latency/power, but who cares when gpus are latency-hiding machines anyway?
this in turn means that nvidia can implement more links or bigger links in their nvswitch networks, which means they can construct bigger systems and push the TCO down.
Two 7900X is still functionally a 7900X, but two 3090s is functionally a 48GB card. Nvidia has got the interconnect bandwidth to a point where it’s a significant enough fraction of the local bandwidth to be functionally one single gpu - this is the same argument as MI300X etc. Doesn’t matter whether the link is on-package or off-package, what matters is that it’s a significant fraction of the speed of your local memory or cache ports. Nvidia did that, with large numbers of gpus, not just a pair of chiplets.
Nvidia has been thinking about this one for a long time - nvswitch is on its third generation, and can switch literal terabytes of data per switch, times several switches. The Mellanox purchase too, but it goes back way longer.
And unlike AMD they actually have a driver that works and just trivially exposes these capabilities and gets out of the way. If you want to tinker and build the open alternative that’s fine, other people want to work.
This is shocking to many AMD fanboys but actually Jensen is a good engineer too, nvidia is mostly on top because they sell products that people want (to such a relentless degree they get furious if they don’t get faster every year etc) and cannot be trivially displaced by “just as good” Radeon drivers etc - just see the latest installment of the geohot saga. Nobody is trapped by nvidia, it is a golden cage - getting actual work done or just going and playing a game instead of spending hours playing with regedit hacks to disable dxnavi to fix DX11 shader compilation stutter is what you’re buying.
Nvidia is on top because of relentlessly competent engineering and savant-level business direction, and as much as people scoff at the idea… that’s literally the reason you hate him lol. He is a Jobs-like visionary figure that can see what the tech can be and drive the engineering and business factors to align along the long-term to get him where he wants to go, while also providing the funding and profit in the short term.
The only company with comparable parasocial negative attachment is apple and it’s for the exact same underlying reason . People are also systematically unable to understand that apple users are not “trapped” or in need of rescuing either. People buy apple because it does what they want it to really well, and they don’t care about installing Linux on their phones. And nerds resent that deeply. It’s not a coincidence there’s this axis of warfare around both Nvidia and the App Store with the EU etc. Nerds cannot abide someone choosing the “wrong” hardware. They are right and you will buy the same thing as them or they will get the EU to outlaw your product, or change the symbol licensing to prevent you running on Linux, etc. If you don't like the same filesystem as me, obviously that means I get to relicense some symbols that have been there for 20+ years and break your filesystem. Btrfs is better, the council has spoken.
It keeps happening for a reason, folks, lol. Nerds can’t tolerate others making different choices. And those users disproportionately self-select to “nerd” platforms like android and AMD.
You're right, it's astroturfing a placeholder in the market in the absence of product. The difference is probably just the target audience - feels like this one is more aimed at share-holders and internal politics.
Landlords -> lobbyists of large real estate investment corporations. One of the biggest social crimes is enabling housing ownership companies to become large investment vehicles.
Post-production QA is an industry problem in general - the difference between systems is who does the QA.
In the dealership model - after transit, they do a once over and post-production/transit repairs before the car appears on the lot. So much so that most states wrote legislation to limit repairs allowed while still being called "new". (often ~5% of the retail value of the car). Note that dealers also pay wholesale rates on the cost estimate for that, so this can be quite large repairs.
The T* direct-model should put the SC in that same spot. Bug is they're not doing the work - probably the emphasis on throughput incentivizes the wrong behavior. All too often the the SC tries to palm-off problems and/or the consumer has to do drive QA. Fixing it means the customer is on point - obviously YMMV.
IMHO this is a huge gap and flaw. Unfortunately OEM dealership behavior is so predatorily atrocious that even this flaw isn't enough to overcome the otherwise positive T* sales experience. I now know how few signatures are needed to buy a car in my state. I have zero intention of participating in the "sit outside of a finance office for an hour, no I don't need Scotchgard, interest rate manipulation" routine ever again.
Dealerships often talk about "relationships" - this BS sales talk - we're just prey to them.
Doesn't matter if you like or dislike the T-word or EM-himself. Fact is they have built one hell of an experienced/industrialized muscle in this field. That experience is priceless right now - Ford, VW, GM etc. don't just need to run-as-fast, they need to run faster. All the while not repeating the mistakes that others were allowed to make when there were less optics on the problem. This technology change is going to be a capital-intensive, painful experience for the incumbents.
Incumbents thought they were going to be bin assemblers as they’ve always been, not an engineering first org with cutting edge battery technology experience. Oops. Can’t MBA way your way to the lead, physics cannot be fooled. I hope they catch up, the space needs robust competition.
This is also why Tesla is building and selling 40GWh/year of stationary storage from their Lathrop Megapack manufacturing facility; their battery management systems are proven under the harshest conditions vehicles see. Utility storage demands are comparatively benign compared to high end powertrain duty cycles in top trims.
Doesn't mean they need to require a google account to provide that info though; It definitely costs more to maintain a good privacy portal, which is probably why the companies I've seen do it either list an email address in their privacy policy and handle it ad-hoc, or they use one of the P/S-aaS providers that handle the portal ( and the concerns with owning the data for the types validating documentation people provide )
> It definitely costs more to maintain a good privacy portal
It costs them more to do fraud and abuse prevention than to run a customer support system that happens to hook into google3 (because that's what it is - if you click 'remove result' it just puts it in a queue for review. The privacy portal at myactivity.google.com/results-about-you is just a way to check on the status of your removals).
Consider what the Union goal is - expansion of membership and increased funding/power/leverage. Uber’s goal is brand-washing. Fighting employment law is a cynical win-win for the Union and Uber in this instance - member welfare be damned. Tells you a lot about the intentions of both the union and employers.
Funny thing is this cycle seems broken. So many deep pocketed competitors have stomped in and skewed the content market by showering money on all the producers.
In turn this has drove up the pricing, enabled exploitation by subpar producers, and over-extension of streamer capital in a bid to stay competitive.
Implosion and consolidate seems inevitable, the market just isn’t big enough to support 5x Netflix sized companies.
It does seem like there should be an opportunity to fund promising content that costs more than $0 (YouTube has that covered) and less than hundreds of millions. Almost like a YC model. What happens if you find someone with a decent premise and give them a bit of money and loaner equipment to make a pilot? They could even make it into an “American Idol” style competition to get funding for more episodes/a full series.
This book was the ignition that changed my life... https://archive.org/details/z-80-reference-guide-alan-tullya...