Intel's plan for the 2018/early 2019

No Icelake in late 2018/early 2019.

Other urls found in this thread:

archive.fo/ZxHMs
twitter.com/SFWRedditVideos

Posting other tables.

...

...

So yeah, their plan is basically sell Skylake for the period of 2015-2019. Perhaps add more cores here and there.

...

>mfw just bought 8700k
feeling comfy desu

Btw, the images weren't taken by me so sorry for the huge size and bad quality.

Here's one of the sources:
archive.fo/ZxHMs
The original source is some German web site: Computer Base.de

Because as said before their 10nm is fucked up.

why would they glue 4 cores together?
isn't it bad like amd level bad?

so they released the coffee lake earlier instead of the original Q2 2018 schedule. Intel is on panic mode.

"earlier"
Covfefe lake was 14 months late due to "complications" with their 10nm node.

wow that's fucked up than I imagined.

JUST FUCK MY 10NM NODE UP SENPAI

Dead on arrival
Zen 2 is going to be out in q1 2019 and shit all over it. At least they still have mobile

panic mode why though?

Where they concerned about amd, or was it because they wanted to get something out untainted by meltdown and specter.

? Coffee lake is still affected. Ice lake might, might not be

They only area Intel lead in now is in clock speeds, but that comes at the cost of TDP and power draw.

AMD are more efficient, with a nigh identical IPC, and significantly lower power draw and thermals. Where AMD can get an octa core into a 65w TDP package, Intel struggle to keep their quadcores inside of 95w

DDR5 when?

everything intel is affected by meltdown

specter is a bit harder to test, but it should still effect everything intel, and possibly amd.

>stuck on kabylake/coffeelake for consumer lines
>b-b-but muh 7nm by 2019
ROFL intel been liars since 2016...

the 1.0.0.6 patch for amd should have actually had amd pull ahead in ipc clock for clock across the board.

intel has a process advantage and clocks, which may be tied to process, amds refresh may close the gap enough to be intels equal, and 7nm will be when amd has the process advantage regardless of what intel does (shy of buying gf and kicking amd out)

Intel's mobile platform power management is still like 2 generations ahead. Race to idle and power state behavior is FAR superior to RR's. Right now the RR APUs still use like 2 to 3x the power for light workloads compared to kaby lake mobile. Idle is a little higher too iirc but thats probably from i/o stuff. I remember some review pointing out using a nvme ssd uses more power than a hdd because it stresses the pcie channels

this, at least you won't have to buy a space heater during this harsh winter.

You've been mentally retarded since you were born.

intel has a slight advantage in idle or very light load, not 2-3 times, amd then takes the throne when you actually use the cpu. but we area talking a single sub 20% difference when its better and when its worse, not a 200-300% difference .

Looks like late 2019 or sometime in 2020 for servers, probably a generation after that for desktops.

It's not as apparent because most of the reviewed Intel laptops have a discrete mx150. Even then there are laptops that have 50-60% higher efficiency in light workloads even with a dgpu. 3x is an exaggeration but its close to 2x. Full load on battery doesn't strike me as a realistic use case. Most people are going to plug in for something intensive. Then again they plug in when they have an option anyway

avx2 benchmarks are Intel's best friend

They needed it to release before the vulnerabilities went public, or it could hurt sales.

>At least they still have mobile
Until samshit or sk shitnyx reach 3Gt/s pin speeds on HBM.
Then, well, it ends even there.

Yeah seems like Intel 10nm is fucked. I actually picked up some amd stock during recent market dip hoping for big gains if this all happens. Don't let me down streetshitters. I think refresh will be solid and 7nm will be a turning point

before the dip this was the first quarter report where amd didn't have a massive sell off, it seems the masses of people who sold at the drop of a hat are gone, expect quarter reports from here on out to gain instead of lose.