DeepSWE is slightly flawed in the sense that is uses only its own harness and that causes issues on models that are not correctly supported by it. There's huge amount of evidence that the harness plays a big role in how these models work and yet DeepSWE entirely removes that (and has probably only tested that it works fine with some favourite model of them).
There's also issues with cost calculation (as that harness doesn't use caches) and so on as reported on their github issues.
None of the benchmarks are perfect, but that does explain a lot of the variations between benchmarks.
It can run and the token generation is fast enough, but the prompt processing is so slow that it makes them next to useless. That is the case with my M3 Pro at least, compared to the RTX I have on my Windows machine.
This is why I'm personally waiting for M5/M6 to finally have some decent prompt processing performance, it makes a huge difference in all the agentic tools.
By what standard is it superior to WoW on Windows? Starting from the available hardware, the Windows world is far superior for WoW resulting in better visuals and more consistent fps which is a good start of having enjoyable experience.
Also I really can't think of the way Mac's mouse settings in WoW work better than the Windows ones. Scrolling, mouse movement are all having more settings under Windows to customize them compared to the experience on Mac (even with all the tools such as Mos to help with the bad external mouse handling in Mac).
So what exactly is so superior? I play WoW on both platforms and when I have a choice (that is, when playing at my home desktop - on my summer cabin I only have the Mac option) it's always the Windows one I select. It's nice that it works in Mac so I can play everywhere as I only have Mac laptop, but the sometimes weird graphics bugs, worse performance and odd issues with rendering displacing my WAs and other stuff just doesn't inspire confidence in Mac gaming compared to using Windows. And the mouse movement.. ugh.
If you look at the map you will see that there isn't multiple tracks coming from Russia to Finland. Some of them were even designed to be blown up if necessary (such as the Salla rail tracks).
Finnish rail roads are mostly north-south bound (or west of Helsinki) which are not helpful to Russian advances. The only way for them to transport weaponry would be through east-west bound (near the border) and there isn't many. It's easy to take such out and they would not impact our infrastructure at all as they're not heavily used (if at all since eastern part of Finland is economically the weakest link anyway).
It's quick and easy in the end to destroy. Rebuilding them under artillery fire isn't easy.
Far easier is to just destroy the train tracks with explosives that connect between Finland and Russia (or demolish them like done in Salla after letting them rot).
There's no defensive reason for this other than in the cabinet talks.
But for similar application, you could use MobilePay (Vipps?). That works across Finland, Sweden, Norway and Denmark.
So although only Finland uses Euro (and rest have their own currency), you can easily transfer money between persons using just their mobile number as an example.
World of Warcraft has never been without issues. When it launched it was full of bugs, servers kept crashing (but they did refund a lot of gametime back then) and so on.
And every expansion was just a nightmare start, without being able to get to the new zone, servers again crashing.
You just have golden memories of a state that never happened. Game wise, WoW has gone forward a lot since DF (the disaster of SL taught them something) and is actually in a lot better state than before. Sure, it has bugs, but it's also a massive game. And they do keep fixing a lot of bugs with quite good response time these days instead of what it used to be.
Where would you draw a line for "any real product" ? ARM for example doesn't actually produce any "real" (physical) product, but they certainly do research and produce technology for other companies to build products on.
There's a lot of "on paper" companies around the world who actually do produce novel technologies even if they don't themselves create the end product, but instead sell their inventions to other parties.
The 8600G/8700G are installed to normal AM5 socket and can be swapped. The laptop parts are socketed and are not going to be available in the AM5 market.
We've even had word processors for over 40 years, why would we suddenly need mouse support to write rich text?
Editing rich text without mouse has always been a thing, Markdown isn't something that suddenly created some fancy new rich text without mouse experience.
I wonder if you've considered (or if there's already) an integration to Git to assist in reviewing pull requests? I often find myself jumping around in VSCode trying to understand some subtle changes in the PRs I'm reviewing which might have small one line changes to many functions.
Understanding how all of those are tangled together would simplify reviews a lot. Especially if the code in question touches places that I haven't visited in a long time (or written by someone else who might have slightly different style of architecting).
I could already see myself using the tool for pair coding / explaining my own PR to others, this looks seriously nice for any sharing.
Can't help but I feel slightly disappointed. This is essentially a weaker version of RK3588 based SBCs, but lacks both smaller cores as well as NPU. And those boards are available with 16GB of memory. And we're talking about a very old core at this point, A76 really isn't new (~5 years old core).
That isn't to said it's useless, the Ampere Altra is using basically the same arch as this one and it's perfectly usable. But if I wanted a small machine, an Alder Lake-N is a lot faster and at the same price - and runs perfectly with upstream Linux. I feel like the only nice selling point is now the formfactor and hopeful future "neat-little-addons" to this one.
RK3588 can't use upstream and I assume this one will not either for now (Phoronix left that out of their article - curious). I would hope this gets better support than the RK3588, but when? And since this is actively cooled in any case, the ADL-N looks far more interesting as cheap computing/server solution.
Isn't iron in spinach insanely difficult for humans to actually absorb? Quick googling says it's around ~2% of total iron in spinach which we can absorb, so instead of eating 100g to get 2.7g you'd need to eat 5kg of spinach per day.
> Depends on the latitude and these numbers seem to be for very high ones close to the polar circles.
No, those values are far from polar circle. I'm guessing closer to central Europe, since for example in Finland the PV produces 0% during the winter months.
10-15% would be insane to get here, but there simply isn't any energy in the sun (and closer to the polar circle you get - there's no sun at all during winter) and the panels are often covered in snow in any case. And I'm not even talking about cloudy days now, but "sunny" ones.
March/October are already approaching those 10-15% levels. Nov-Feb is closer to 0% in most of the Finland.
Because Windows UI is better than anything MacOS or Linux provides. And because everything just works (tm) when it comes to drivers. And hibernate, sleep etc are always there since I don't need to dual boot. And I don't have to deal with buggy Cinnamon scaling experience that just occasionally decides "maybe I won't render things correctly today or just crash the UI when Firefox is launched".
I've ran Windows as the host for my Linux VM for years because it always provided me nicer experience than the other way around. Not to mention games are of course native in that context also (and there's no issues with anticheats).
Wouldn't that depend on the sink? If they can cool down to the water, then running a water-water-heatexchanger would require less power (due to better heat transfer) than simply cooling it to the air.
Same I guess if they could use evaporation to do some of the cooling.
In the end, living on northern region (Finland), my PC power costs in reality are meaningless. My gaming probably eats what.. few hundred kWh per year? Lets say 1000kWh to make something meaningful. 1000kWh is less than the yearly fluctuation of electricity consumption of the house in heating costs. It's less than my Tesla's winter consumption fluctuation depending on the amount of heating the battery / car requires for me to open the doors.
See the insanity of optimizing couple of watts from PC? I don't play 12 hours a day, I play after work in the evenings - at best few hours a day, but not every day.
My workstation's power usage does not have any impact on my life. Even if doubled, tripled, quadruppled, it would be lost in margins of everything else where I use electricity.
PS5 eats about the same amount of power to be fair (or my XSX) so there's not much to save here. Also, they're a bit slower than my desktop and do not provide the flexibility.
Proper gaming performance requires power as no one has invented any golden goose to reduce it (not AMD, not NVIDIA, not Apple) at the same performance. I'd even wonder if Apple's usage of GPU in M2 Ultra is that power efficient when looking at the real world games as the fps counters are so low.
Theoretical performance is pointless in the end if it doesn't actually crunch anything that fast in reality. Maybe the glue architecture is only nice on paper and that's why Apple after years of ignoring specs is now only advertising specs on their GPU parts, not what it could do.
There's also issues with cost calculation (as that harness doesn't use caches) and so on as reported on their github issues.
None of the benchmarks are perfect, but that does explain a lot of the variations between benchmarks.