Phate.exe

It's like the Phenom II era in that way, although at least this time they locked you out of unlocking the CPU's.  They were making the full chips (quad cores deneb's and late 6 core thuban's) and disabling "nonfunctional" portions of the CPU to sell them as lower end.  I had a 1090T six core, so I didn't get to play around with unlocking cores until last year when I scooped a 960T black for the machine I put together for a friend, and we ended up with a 3GHz 6 core with overclocking headroom for under $50 shipped.  feelsgoodman.jpg
 
Make the "big" 2 CCX die, and really good ones become 1800X's, fully functional cores that don't like high clockspeeds as much become 1700's.  1 or 2 dead cores become 1600's and 1600X's depending on how well they clock.  Etc etc etc.  This is why the odd 8 core Ryzen 5 and 8 core non-SMT Ryzen 3 showing up doesn't surprise me at all.
 
Arguably the lower-end processors actually cost a bit more than the 1700/1800, since they have to disable portions of the chip after testing, rather than just sending them to packaging.  It's not a lot more, but it is extra production time.
 
 
I'm pretty happy with my 1600X right now.  I'm still interested in what changes we'll see in the refreshed CPU's, in particular where the new clockspeed ceiling will be.

It's like the Phenom II era in that way, although at least this time they locked you out of unlocking the CPU's.  They were making the full chips (quad cores deneb's and late 6 core thuban's) and disabling "nonfunctional" portions of the CPU to sell them as lower end.  I had a 1090T six core, so I didn't get to play around with unlocking cores until last year when I scooped a 960T black for the machine I put together for a friend, and we ended up with a 3GHz 6 core with overclocking headroom for under $50 shipped.  feelsgoodman.jpg
 
Make the "big" 2 CCX die, and really good ones become 1800X's, fully functional cores that don't like high clockspeeds as much become 1700's.  1 or 2 dead cores become 1600's and 1600X's depending on how well they clock.  Etc etc etc.  This is why the odd 8 core Ryzen 5 and 8 core non-SMT Ryzen 3 showing up doesn't surprise me at all.
 
Arguably the lower-end processors actually cost a bit more than the 1700/1800, since they have to disable portions of the chip after testing, rather than just sending them to packaging.  It's not a lot more, but it is extra production time.
 
 
I'm pretty happy with my 1600X right now.  I'm still interested in what changes we'll see in the refreshed CPU's, in particular where the new clockspeed ceiling will be.

They're not using it to fight Vega mobile, they're using it to develop their EMIB tech.

lol
Breaking News! A website that you created an account on and which recommends stuff based on your watch history and ratings knows what you watch. More in a bit.
 
I mean Netflix just looked at one of their movies and saw some unexpected viewing habits from some accs. 

It's not like they're even looking at specific user's viewing histories to get to this data.  Anything with a "recommended for you" field is obviously going to be tracking your usage data.  They probably have some sort of histogram of the number of times something was viewed vs users, then saw a weird spike, then noticed that it was one user watching something over and over again.

It's not like they're even looking at specific user's viewing histories to get to this data.  Anything with a "recommended for you" field is obviously going to be tracking your usage data.  They probably have some sort of histogram of the number of times something was viewed vs users, then saw a weird spike, then noticed that it was one user watching something over and over again.

If you're gonna just be gaming at 1080p, especially if you aren't going high-refresh, you'll be good for a LONG time on a GTX 1070.

If you're gonna just be gaming at 1080p, especially if you aren't going high-refresh, you'll be good for a LONG time on a GTX 1070.

I would certainly mess around with this on my 950.  The snapdragon 808 is still snappy, it has enough ram (3 gigs), and I've been completely spoiled by both the screen and camera to the point that I'm not especially impressed by most new phones.
 
If somebody puts together an Android ROM with a W10M or W8.1 style launcher, app list, and most importantly DARK THEME, I'd rock the hell out of it.
My 1020 was an absolute tank, just like any of the polycarbonate Lumia's.  My 950 has also taken a beating, although I like the look and feel of the polycarbonate unibody phones (920/1020/1520/etc).  Cases were not necessary on either one.
The 520 was a cheap sub-$75 phone (I saw them go as low as $30 on ATT GoPhone).  It was low spec and cheap, but managed to still feel snappy in general use.  The thing that actually got me to switch away from was the fact that in the Galaxy S4 era there really weren't many sub-$200 android phones that were actually worth using.  I bought my old Lumia 521 (T-Mobile version of the 520) for $60, and liked the low-spec Windows Phone experience enough to buy a used Lumia 1020.
 
Like I don't know if I can properly articulate just how trash budget android phones were in late 2013 early 2014.

it is; but to be fair, if some random dudes can get it even semi-functional, it really brings into question the validity of Intel's claims towards its being completely incompatible. Intel, or any of the board partners, with all their inside knowledge and extremely low-level product information/workings.... they should have no issues making a perfectly functional version of this same stuff.

3D printing isn't more popular because it takes a lot of know-how. Getting great prints with high reliability takes a LOT of work. Even if you buy a very high end printer it's not set it and forget it. Just learning the software and how the machine operates is fairly daunting, and downright frustrating for a non technically minded individual. Not to mention the actual design process for parts is a whole separate skill that is a huge time investment to learn and can never really be automated even if you had the perfect, one button printer. Resolution doesn't even crack the top 5 of what is stopping 3d printers from becoming popular. 

TFW your monopoly is threatened and you need to fight back

This.
 
The cards really just have a bad case of being poorly-tuned out of the box.  Just like the RX 480 did.  I'm interested to see if any of the aftermarket cards are tuned better, and if not at least I'd be able to start with a quiet cooler.
 
Vega is obscenely efficient if you keep the clockspeeds sane (and adjust voltage accordingly), but when you get to a certain point they become power hungry monsters.  A Vega 64 running at R9 Fury X speeds (which happens to perform pretty similarly to a Fury) uses about as much power as an RX 470.  Performance-wise, they're extremely competitive in a world where the 1080ti doesn't exist (so basically if it didn't get delayed).
 
Because it didn't perform quite as well as they wanted it to, AMD pushed clockspeeds (and voltage) in order to get better looking benchmarks, just like they did with Polaris, and then we end up with everybody talking about ridiculous power draw.
 
Saying "They're quite good if you're willing to spend a few minutes screwing around with the card" is absolutely not an excuse for how they perform out of the box, but just dismissing the architecture completely isn't quite fair either.

also wait for the XFX or Sapphire or Powercolor version of the Vega cards for better power delivery and cooling than the reference version
 
 
Skip the Asus Strix as it suck balls and worse than the reference

Interesting that Notebookcheck had pretty much the opposite conclusions from their benchmarks. Guess we need more sources and quite frankly more devices to test with.

What? of course a single 1080 Ti is sufficient for 4k60fps gaming, at worse case scenario you lower a tiny little bit of in-game quality settings, that's it.

I don't think you understand the graphs you posted.
 
The first ones you posted appear to be at ultra settings.  Very high/ultra, all features enabled pretty much translates to "all sliders to the right, they don't always call it ultra so we won't either".  The Metro Last Light benchmarks I have no clue, you just posted a bar graph with any useful information or context cropped out.  On Rise of the Tomb Raider, Very High is the highest quality preset, with a handful of additional settings (purehair, soft shadows, etc) that have an additional notch to go up.
 
Read/watch literally any optimization guide.  There are some settings that hit the gpu HARD despite offering very little improvement in quality outside of static screenshots.  It's just silly to see a card struggle at the highest settings and extrapolate that you're gonna have to drop things to low in order to hit 60fps.
 
 

Right.  I've looked at benchmarks, they're almost always done with things turned way up past the point of diminishing returns.  I said "reasonably high settings" on purpose.  I get 45fps with a mix of medium and mostly high settings is on the Fury, a card from 2015 that barely cracks 1000MHz, and most of the time graphics settings need to be turned down due to running out of memory, which I can't do anything about.  A Vega 56 has about the same clock-for-clock performance as my card, but will clock 50% higher.  Sure, it's not gonna scale linearly, but between the clockspeed and not running out of vram 60fps is no problem.
 
You're acting like stuff is gonna be unplayable, and shit like this is why everybody thinks they need to buy a 1080ti for $700+.  If all you're gonna do is push all the sliders to the right, well yeah, of course performance is gonna tank.  But while there might be a large difference in quality between low and medium, or even medium and high, there's not a large perceptible difference between high and very high, and even less between as you get to the maximum settings.
 
Spend ten minutes tuning the settings, get your 60 fps, barely notice the difference in quality, if at all.  @Princess Cadence linked a video that goes into far more detail.

Fear not!
 

Problem is you might have to cross the line of what is street legal in some countries to achieve it, and for mass(ish, it is Tesla) production cars the spec sheet is for what is street legal and complaint in all regions it's been sold in. That's why I'm looking at the numbers with a fair amount of skepticism, because the range they are also claiming indicates to me that it's not going to be that light weight. 
 
It should still be possible in an electric car though due to the very different way they accelerate, just have to not break traction on the initial start and your 90% of the way to do it. However I'd personally only be comfortable driving in such a car if it was using tyres like the RE050A I use and not Michelin Pilot Sport Cup 2 or similar, because I value my life more than the car being able to do 0-60 in 1.9s.

I'm well aware.  And none of them are running high second quarter miles, at least not on those tires.  No need to be condescending, you're making the assumption that a tire like the Pilot Supersport (even the more custom versions that they use on something like a Koenigsegg or similar, exactly the tires your referring to) is in any way remotely comparable to a DOT-legal race tire.  A prepped surface for drag racing is more than just clean.  It's actually sticky, like you'd lose a sandal walking across it.
 
I'm saying that the numbers they are claiming are extremely difficult, even for race cars running parts and maintenance schedules intended for racing.

Yeah, I'm interested in some more detail specs for sure.

Seriously questioning those numbers for the supercar.
 
An 8.9 second quarter mile time, from the factory is literally insane.  It would also be faster than the current electric drag record (for an actual car).  A Rimac Concept One (another batshit crazily fast electric supercar with over 1000 horsepower) goes 9.9.
I have absolutely no problem with EV's and hybrids, and actually would love to build an EV of some sort.  I've driven enough RC cars to know that electric motors can be pretty nasty.
 
My problem is that most of them are just soul-crushingly boring to drive.

I love the fact they're comparing the Cd between vehicles of wildly different sizes as if it means anything.  Marketing wank at it's finest right here.
 
Don't get me wrong, their semi tractor is massively impressive, and I fully expect to see around town and short over the road routes (think distribution centers for stores like walmart/target/bestbuy/etc) handled by electric trucks in the near future.  The fact that time spent loading/unloading allows for some built-in charging is another huge advantage, along with the ability to re-capture energy that would otherwise be wasted in engine braking.
 
But Cd is only part of the drag equation (primary considerations are Cd, speed, and frontal area).  Yeah the truck is way slipperier than a normal semi, an achievement in itself, but it also likely has a frontal area of like 10+ square meters, compared to the 2-ish square meters of the Bugatti (the Veyron was 2.07).
 
So yeah, the truck has a great Cd, but it's stupid to compare the drag area, which is what actually matters (literally just the Cd multiplied by the frontal area).