TCNs are "secret" in the UK
hence the WaPo breaking this - whoever leaded it knew that Technical Capability Notices are not discussed in UK media...
80 publicly visible posts • joined 7 Dec 2020
Given all the major cloud providers are struggling to expand capacity at the rate their customers want, this is going to be key. Designating DCs as CNI will hopefully make planning for SMR deployment to power DCs easier given at the moment it is not clearly defined how a private enterprise could leverage SMRs to power DCs as MS/Amazon/Google all want.
There's probably a revenue stream for the UK Gov by providing SMRs as a service to private enterprise in scenarios like this.
Given MS, Google and Amazon have all expressed interest in using SMR reactors for powering DCs, and the UK has in Rolls Royce, Amentum et al a manufacturer that can make them but no clear regulation about deploying them, this seems to be an area where some clear regulation could help.
I wonder if part of the "less regulation" approach, rather than being aimed at things like less AI regulation, might include making it easier for private SMR nuclear power for DCs. Currently I'm unclear on what the regs are around this and MS, Google and AWS have all expressed intent to power DCs with SMRs. After a chat with a friend who works on these for a major engineering firm current regs are fairly unclear on how these could be legally deployed by private companies and where.
this is 100% true for Azure DCs. I suspect the future will be SMR powered DCs and Oracle have just announced their first.
Also when siting DCs you need to take into account latency. All the big players are starting to create new regions due to capacity issues and encouraging customers to go multi region where possible, as well as dissuading new customers from deploying to existing resource-constrained regions.
When you site a DC, you do a whole lot of worst-case thinking. This includes things like not building near fault lines, taking into account geopolitical potential in the area etc etc. Building next door to a nuclear power station adds a whole lot of potential worst case, whether from accident or military conflict. Not saying it won't happen, but you don't just plonk these things down next to the nearest cheap power if you're a major player.
...as broadcom raise customer prices massively (like 8x in some cases) and customers are coming to us with a "screw VMware, we want off ASAP" attitude. This is a problem entirely of Broadcom's making and they should not be surprised. They publicly said that effectively they're gonna rinse their top customers who will take a while to migrate off whilst letting all the small customers churn off the platform, make out like bandits for a couple of years and then let the product die. It's a damn shame.
If an AI company were to crawl a site's contents against the robots.txt policy, I presume that site could then press legally for the removal of their content from the AI. I suspect this would be non-trivial and expensive for the AI company question. Perhaps it needs a test case to sharpen the issue.
No idea why everyone's acting surprised. They did this before with Symantec. The investors will make out like bandits for the couple of years that customers take to migrate off this dead platform. VMware will die. Shame that such a good platform has been deliberately killed, but this is the business plan of Broadcom.
All other cloud providers are scrambling to help customers off VMware - for a large org this isn't a quick or easy process though and could easily take a year or two.
So from an MSFT side what we're hearing from customers is those top 600 are facing licence increases of 17x in some cases. Moving a LORG off VMware is potentially a 2-3 year project so they're banking on, well, making bank from them for that period, whilst the angry customer offboards to a competitor. They don't care, they know it'll cause the death of VMware, they want a short term profit from those customers who are locked in. Feels almost like piracy.
Lots of customers going to AVS in Azure for easiest migration path then reconsidering their options to go either Azure IaaS or Stack HCI, neither of which are an easy or quick option. I'm sure GCP and AWS are seeing the same.
Plus for those top 600 customers on VMWare you've got to ask what the support is going to be like given all their team are going to be jumping ship at the earliest opportunity and there's obviously going to be no investment in the product going forward.
It was a solid platform, and investment capitalism has killed it and hurt thousands of companies dependent upon them.
Company cars are almost always leased from leasing companies. It would be rare for a large company to do this direct from manufacturer, they would go through a leasing provider. I can understand why they might be reticent to allow employees to spec Teslas though given long repair times and uncertain resale value due to constant random discounting which tends to make leases expensive on these vehicles...
This. Microsoft has always supported technical debt, and it harms everyone ultimately. It means that cruft stays in OS and apps, it hampers development and it slows progress and improvements to security and reliability. Historically the thing that killed old PCs is that the new OS just runs slow/badly on it and I can see why a line in the sand needs to be drawn by MS at some point. This dependency on hardware support for security features is relatively new in that it's a binary yes or no for W11. It's going to annoy a lot of people, but what's the alternative? Allow customers to run old OSs forever or force people into the current? I think if I were controlling it i'd allow consumers ESUs for free/low charge for a period of time, which is probably what they'll do.
Working 2nd line support for a large pharma company. Got sent into a restricted area with animal testing - negative air pressure zone so nothing outside could get in to contaminate experiments, which also meant that nothing in there with moving parts like CPU fans could leave without going thru an autoclave which would obviously kill any PCs - so fixing a lab machine had to be done in situ. Spent an unpleasant hour reimaging a lab PC whilst a worker gutted mice and chucked them in a bin a few feet away.
In the MS world I remember a PFE getting a call out for UK military for a sharepoint problem. Despatched to an address where it became immediately clear that it was an airbase and no, that wasn't where the server was. Straight into a transport plane and off to Iraq greenzone, after being allowed a phone call home to tell his wife that he probably wasn't going to get back home for a few days :-(
...is that the kneejerk "why should we pay for software config" means that now BMW will have to create a new SKU - a BMW with the heated seats fitted, vs one that does not. This adds cost and complexity on the assembly line and supply chain. Say, for example, that the cost of all BMWs rises by £20 as a result, and those that pay for heated seats to be fitted pay an additional £300 (to pluck some numbers from the air). Is that a good deal, or would it be better for everyone to save £20 on their car, and only those that want heated seats pay an extra fee?
An entirely history based article, that covers precisely none of the issues of competition right now. You could have at least done some basic homework, compared market size and sales data across platforms etc etc - but no, just a "M$ is bad" article that could have been written in the 90s. Sigh.
Microsoft has been in business in the UK for 40 years, and is pretty much the largest corp in the world. Pissing off the UK whilst at the same time mouthing crap like "Unicorn Kingdom" is a stupid and shortsighted UK policy that will a) likely reduce MS investment in the UK, and b) make the UK even more of a digital irrelevance as no UK company has the power to host planetary scale AI like the big boys do.
You connect to corp resources, you get a corp mobile device management policy. This enforces things like:
* encryption
* PIN/passcode complexity
* Block/allow list of apps that can be installed
* Remote wipe (with granular control of work-only resources using Android Work Profile)
* Conditional Access to corp resources (e.g. you have to be patched and up to date on OS etc)
This is TRIVIAL to set up. There are loads of options for corp IT and if they are an M365 user they can just enable InTune in the management console and it'll do it for them.
IMHO ALL corps that care about security should be doing this now. This also should apply to gov/civil service etc.
This is FUD. MS have a Trust Center that details EXACTLY what is done with data and their model isn't to monetise it like google. It'll offer an org very granular control over where their data resides. They spend millions on this already for their existing offerings, it's a key differentiator from google
I remember a 2nd line engineer at a large pharma company in the UK once creating a gold image for (I think) an NT3.51 image. They accidentally included their personal profile in the image. Which included a *LOT* of MP3s, dubious NSFW images etc. It got deployed to a reasonably large number of people and they only realised when a scan of a network discovered things that really shouldn't be there...cue a lot of backpedalling and work to remediate without telling management...
Oxford Uni sets a LOT of essays for students. They are marked, but the marks aren't really important compared to the marks given to sessions with tutors where you are asked to discuss your essay. Kind of hard to cheat in that scenario and also protects the need for trained human educators...
No, running crypto mining in the cloud where you're paying for your compute isn't going to be cost-effective. How could it be?
So it follows that pretty much any cloud crypto mining is *probably* someone doing something they shouldn't, on someone else's dime. Folks with access to their company's Azure subscription who thought they'd do a bit of unofficial mining that they don't have to pay for, etc.
There are definitely specific use cases where you might want to do crypto mining, e.g. creating a dev environment for testing, and for those instances you ask and are granted permission.
For everything else, blocking it seems kinda sensible. MS aren't stupid, this is being done to protect people.
Microsoft's own "Project Natick" proposed a solution to this as well: containerised submarine DCs.
https://natick.research.microsoft.com/
I forget the exact quote but a very high percentage of the world's trading centers are close to water, so this offered a way of using the sea for cooling and low latency by being physically close to the trading floor. Using Azure Stack HCI you have pretty much feature parity with standard Azure and a single management plane as well for this.
At the other end of the scale, MS invested in LSEG so there's probably some commitment on the LSEG side to use Azure Cloud, you would imagine: https://www.theregister.com/2022/12/12/london_stock_exchange_ms/?td=rt-3a
See https://blogs.microsoft.com/blog/2022/12/09/microsoft-acquires-lumenisity-an-innovator-in-hollow-core-fiber-hcf-cable/#_ftn1
It's a spinoff company from Southampton Uni, which is pretty cool, and I think Southampton also worked with MS a couple years ago with Project Silica for data archiving in glass.
If you can shift bits 47% faster between HCI, then that's pretty awesome in an HCI data center, of which MS has hundreds across the globe.
Seems a very smart buy to me!
This is a dumb article trying to be clever. FOSS is great, but to argue there is NO commercial software that's worth paying anything for is plain stupid.
To think that you, with your little cute DC of a handful of servers, can compete for availability with a commercial cloud platform is also stupid.
There is nothing wrong with a professional artist paying for Photoshop, for example, and no, GIMP is not a useful alternative.
it's a case of MS having vast amounts of concrete data on reliability of its fleet, which is custom hardware. They know way more about its performance and reliability than the vendors themselves do and indeed have negotiated custom warranties with the components. When you run at planetary scale you can try things like "what if we run a DC at a slightly higher temp, can we see any difference in hardware reliability? Can we save energy on cooling?" or "what if we extend hardware life, do we see any increase in storage errors?"
It's foolish to replace functional kit when you don't need to based on an arbitary lifetime. Do it based on changing hardware requirements, or improvements in perf or reliability. A fixed "best before" date doesn't make a lot of sense, particularly when you have billions of cores...