Untitled
Other urls found in this thread:
DELET
SHITEL POZZED HOUSEFIRES BTFOREVER
sir do the needful pls delet
INTELBROS THIS WAS NOT SUPPOSED TO HAPPEN
OH NOONONONONONONO
NOOOOOOOO! THIS CAN'T BE TRUE!
Where were you when
>INTLEL IS FINISHED AND BANKRUPT
>a bigger nm is worse and slower than smaller nm
Color me surprised.
Intel's 10nm is slower than their 14nm
The only way it was meant to be Jewed.
Great, when can I fucking pre-order this?
You can buy it in October. And THIRDRIPPER in December.
Kek
u wot
the 3950X goes on sale in September
>u wot
3950X is coming out in October and THIRDRIPPER is slated for a release in late December of 2019.
AMD said the 3950X will go on sale in September and the only thing they've said about Threadripper is that it'll be Q4.
Hmmm...I guess my info is old on 3950X. So, it's actually earlier by one month, eh.
>3950X will go on sale in September
i am cum
DELID*
CIA
>pure despair
Its over
>spend half a decade refining 14nm
>change back end metal pitches and decrease density without telling investors that their touted figures are no longer accurate
>tweak things to get clocks as high as possible
>half a decade delayed 10nm can't compete
>10nm process also had metal pitches changed to decrease density without any public announcement
Don't worry, intel bros. By 2024 10nm will be just as good.
The real question is when investors are going to sue intel for lying to them constantly
>its only delayed a quarter!
>its ramping up faster than 14!
>we're industry leaders!
lol
Reminder.
Wait until they get SMT4 oh boy
Gaymernexus on suicide watch
This was a super old rumor and its not really based on anything except vague Linkdin job postings that a schizo named Seronx put together.
Supporting more threads would require making the core even wider. Look how wide IBM's POWER9 is compared to Zen.
Even Seronx in his speculation thought that Zen3 would have to split and differentiate between enterprise and consumer SKUs, having different cores for each market, because a core wide enough for SMT4 would just waste power and clock worse per watt.
Surely all this anti-intel bias in the media is a direct relationship to anti-semitism
Stage 1
>This was a super old rumor
youtube.com
>May 8 2019
>super old
Yes, user. Super old. The guy who originated this claim was spreading it around Anandtech and SemiAccurate in 2016.
Someone taking the same unfounded rumor and putting it in a video does not give it any more credibility.
I just discovered semi-accurate. What do you think of the website?
>just discovered S|A
Imagine being this new
You gotta start somewhere.
Watch the video, you dumb reality-denying piece of fagshit. It's pretty much confirmed.
>watch this speculation video from a total nobody!
Stop posting your own videos and begging for views.
Intel 10nm supershitter 2020
What reason do people have to still use Intel processors?
Seriously, I don't get it.
Just how much better does AMD have to be before people switch?
Or is it just illegal trade practices from Intel that are keeping them alive?
Fanboy-ism.
I really hope this is true, but we should be sceptical. It could be faked or it could be ultra cooled with LN2 or something.
Either way, fuck Intel.
So your claim is Intel is still in the lead?
I need proof. Proof required.
>It could be faked
>Official database of geekbench
Now you're going to saw HWbot can be faked too, kek.
*say
How does GeekBench stop people just sending numbers to their servers, for example? I'm sure there must be some system - maybe your computer sends an encrypted string of the results or something if it's a genuine test. But surely someone could just monitor this outbound traffic and figure out how to spoof it.
Anyway, I'm just saying, as much as I do want AMD to beat Intel, we should be sceptical until we see the actual part for real.
>How does GeekBench stop people just sending numbers to their servers
Geekbench revises all the content, it's impossible to pull a fake through.
>as much as I do want AMD to beat Intel
>want to
They already did.
What do you mean by revises all the content? Revises it in what way?
If the benchmark is legit then yeah. But we don't know necessarily if it is.
>Revises it in what way?
Nothing gets onto database until it gets confirmed to not be fake. 3dMark works same way.
>If the benchmark is legit
No, they've already beat Inturd quite a while back, actually. See this:
>US Dollars (higher is better)
hmm
How do they "confirm" it though? Do you mean they run the same test on the same hardware? But then Geekbench would need a 3950X and they're not released yet. Also the benchmark in OP's link isn't explicitly identified as a 3950X. It probably is, or at least is a manufacturing sample of one, because it's using the X470 chipset (which is AM4 socket) and it's got 16 cores, and the only chip of that description is the 3950X. Although you'll notice in pic related that the reported clock speeds are lower than the 3950X is advertised as. Maybe this manufacturing sample was just underclocked or something.
Anyway, I will still reserve final judgement for later.
They've beaten Intel on price/performance, yes, but outright performance is something they haven't quite yet got - apart from when they have chips with more cores that get better multi-threaded performance than equivalently priced Intel chips (due to more cores), but they suffer on single-threaded performance in those scenarios.
>but outright performance is something they haven't quite yet got
That picture is interesting given that the 9400 does better on single-threaded on Geekbench.
Maybe people are overclocking their 2600Xs? And maybe that's why it's getting better scores on that website?
HIGHER=BETTER