-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Intel's long stagnation at 14nm
- Thread starter Quartz
- Start date
-
- Tags
- 14nm hardware unboxed intel
Soldato
At least we will hopefully get decent gains which means upgrading every 1-2 gens will be worth it now, providing we get 15-20% gains per gen.
Intel were so far ahead for a while, AMD came back and made decent gains but still took them years of new Ryzen chips to catch Intel in gaming.
Intel were so far ahead for a while, AMD came back and made decent gains but still took them years of new Ryzen chips to catch Intel in gaming.
Decent gains? AMD hammered Intel into dust. Gaming performance was pretty even.
This is a warning for the future. Once we reach 3nm ( or whatever the actual physical limit is ) we will see a stagnation in performance from smaller processes.
AMD have done an amazing job but part of their success is to do with the fact that their chips are manufactured on a smaller node that is far superior.
AMD have done an amazing job but part of their success is to do with the fact that their chips are manufactured on a smaller node that is far superior.
Caporegime
I haven't watched it yet but why would they make this video now? When Intel are about to launch some pretty good CPU's on 10nm, its a bit clickbait and about to become irrelevant.
The counter to that is the one where Intel people say their 10nm mode is approx amd's 7nm in terms of density or something. They can't have it both ways either that's right and therefore AMD have done better, or it's as you say and AMD have an advantage in which case that's pretty poor from Intel in not getting their ass in gear and coasting to a losing position from being a couple of laps clear in the race. Either way AMD done good for once
GloFlo 14lp Trolls you hard
While that is true, I had long considered that if the time ever comes where all foundries are no longer able to advance (at costs which are at least somewhat viable), then the playing field would be truly level.
At which stage with no more new extra transistors to use every 1-2 years, designs would truly have to focus on making every transistor count.
Maybe they had been considering for a while and had dug out a few datapoints, started a write-up etc. and decided that if they don't publish now they can throw all that away?
You never know, it always kinda looks like stagnation is just around the corner but never seems to happen
Soldato
- Joined
- 6 Jan 2012
- Posts
- 5,504
Same here, will only upgrade once DDR5 is sensible price anyway maybe longer. I would jump if i5/i7 8xxx or older though.
Probably a while after AMD move to it’s servers to DDR5. The market will ramp production then.
Soldato
Welcome to YouTube.
Hardware Unboxed's channel is very limited in scope so between new hardware launches they have absolutely nothing to do with so they continue to rehash the same video they've made 10 times before as filler content so they can get money from YouTube - it's really quite a boring channel
That doesn't prevent them making the odd interesting video.
Caporegime
Yeah i know.....
I think the stagnation has approachiing a while. In the 90s and early 2000s clockspeed increments were enough to drive performance. We then saw a move to 64bit and multithreading. Then came more of a focus on IPC improvements, better cache and finally out-of-the-box 'overclocking' / boosting (I'm not saying these things were sequential, they overlapped but those are the general themes I've observed). The days of big gains, where upgrading every two years would double performance levels seem to be long gone. Admittedly we're not at a point of true stagnation yet but the number of levers that can be pulled is shrinking - turbo/boost clocks/PBO/whatever you want to call it isn't new per se, but it's very much something manufacturers seem to be relying on these days to drive higher performance, and that's arguably a warning sign that other options are drying up (notwithstanding some potential energy consumption shenanigans)
Soldato
Literally not true.
Care to elaborate on what is incorrect?
Have AMD not shown amazing progress in the past 5 years? AMD share price seems to show improvement.
Or are you saying Intel's 14nm chips are as good as TSMC 7nm? Someone should tell Apple they are making a mistake.
Soldato
The notion that AMD's success comes from a superior node is incorrect. It's a long-standing myth perpetuated by the Intel circle jerk to explain just how their beloved company can get battered around the place by somebody a fraction of their size. It's also a notion conveniently forgotten when Intel have even the slightest lead over AMD.
AMD's resurgence has been through sheer brilliance, helped in no small part by Intel's arrogance, complacency and incompetence. Then you could argue TSMC's 7nm has played a factor. But if TSMC 7nm is truly some god tier **** powering AMD's success, why doesn't RDNA 2 smash Ampere across the face in every metric, rather than being merely competitive?
AMD's resurgence has been through sheer brilliance, helped in no small part by Intel's arrogance, complacency and incompetence. Then you could argue TSMC's 7nm has played a factor. But if TSMC 7nm is truly some god tier **** powering AMD's success, why doesn't RDNA 2 smash Ampere across the face in every metric, rather than being merely competitive?