Intel is dead

Imagine seeing this and being on suicide watch

Attached: chrome_LhgYZHB7VC.png (252x171, 98K)

Other urls found in this thread:

youtube.com/watch?v=M2LOMTpCtLA
twitter.com/SFWRedditGifs

I'm guying Intel simply because it's blue, that's all I need, don't care about anything else.

Attached: 1547449668785.png (500x600, 210K)

What's that?

traces for another chiplet making it a 16 core

>Intel is dead
>no amd laptops in dell, hp or lenovo store
very much alive in brazil i guess.

AMD demoed an unnamed 8c Zen 2 Ryzen chip that ostensibly was pretty much tied with an i9-9900K for some cherry picked benchmarks, at a much lower (~65W?) power.
The organic substrate had a conspicuous space for a 2nd 8c compute chiplet to be placed. Even though Lisa Su said "moar coars" and deliberately didn't commit to a full 16, that's what everybody is acting like is guaranteed now.

The next Coffee Lake revamp (Comet Lake) is purportedly capping out at 10c, so even 12c Ryzen would beat Intel.
I have a Threadripper 1900X and will be upgrading to >=24c late this year, so I do not give two shits about whether this gen of AM4 Ryzen maxes out at 8c, 12c, or 16c.

Attached: AMD-Zen-3.jpg (1452x949, 80K)

delid

Won't the possible 16 core variant be starved to death by dual channel ram? Seems the IF isn't going to like that to much.

Why do people fan boy over fucking pc parts? If there wasn't competing brands, the prices would be terrible and the performance would be garbage.

nope, it is proven false. it will be enough.

I cheer for innovation, not brands. Ryzen 3xxx looks like it's going to be a significant step forward in CPU performance. That's a good thing. If Intel manages to actually fix their 10nm node and develop a new uarch to compete after Ryzen 3xxx is launched, that will also be good.

So I bought a R5 2600 and.. its literally nothing. Literally the same as my 2500k, not a gaymer btw.

What I am getting at is obviously that moar corez = doesn't matter.

Wasn’t the demo against two i9s though?

until it does

What the fuck do you guys do all day on your computers that's entirely single core

Even web browsing is multithreaded now

AMD got good deals with oems now so i expect that to change

You are thinking of the EPYC 64c demo which at one socket is equivalent to a two socket Intel server

The ryzen 3000 demo was 8 core vs 8 core

>Comet Lake
you mean Cope Lake

Threadripper doesn't suffer any negative effects with 32c/64t with just quad channel memory. AM4 should be problem free with 16c/32t and dual channel. Zen2's IMC should handle faster DIMMs than Zen1 also.

>New arch
>Node shrink
>Still can't beat 14nm++++++++++++++++ intel's housefires
LMFAO

More like coma lake lmao

Attached: 1505209564672.jpg (1500x1054, 122K)

hello sirs please do the needful and buy ryzen cpu and radeon gpu my son rakeesh needs to eat

hello sirs please do the needful and buy intel cpu and nvidia gpu my son rakeesh needs to eat

nope, proven to be a software issue in the windows scheduler not an issue with infinity fabric memes or non-uniform memory access
see: youtube.com/watch?v=M2LOMTpCtLA

Double the shit now

Attached: 1472569081370.jpg (770x711, 102K)

The current R7 2700 scores 1561 in Cinebench multi-threaded and 168 in single threaded. The Ryzen 3000 65w ES at CES scored 2050, and based on the single-threaded to multi-threaded score ratios of the 2700 and 2700x, must have scored around 200 in single threaded.

That's ~30% faster in multi-threaded and ~20% faster in single threaded (at least). Now imagine how fast the 105w x version will be.

Imagine having the best 7nm process available and only managing to match a 14nm+++ 9900k core to core.
Ryzen gets slower past 8 cores due to CCC latency.
So why would 16cores be any faster?

No, that was the demo of a single Rome beating two Xeon Platinums.

Attached: 1517599466043.jpg (800x450, 45K)

>Pretending not to realize the Ryzen 3000 at CES was a 65w chip and that the 105w x version will be much faster.
>Actually being too stupid to know that Ryzen's gaming performance was hampered by memory latency, not core to core latency.
(lol

More like Zion lake

cope pajeet ayymd still behind in games

>and based on the single-threaded to multi-threaded score ratios of the 2700 and 2700x, must have scored around 200 in single threaded.
that's where you are wrong and the entire zen2 series gets pozzed

>muh 4% in gaymes

Attached: 1547419788322.jpg (588x823, 117K)

It's a blue waffle.
Google it.

>3%
More like 25%. But still, who cares about gaymen.

>being this angry you started your own thread

>More like 25%
Been living under a rock for the last year, mate? Ryzen is within margin-of-error percentages away from Intel nowadays.

>Even though Lisa Su said "moar coars" and deliberately didn't commit to a full 16, that's what everybody is acting like is guaranteed now.
Because it is.

It's a reasonable hope, but there's good cause for AMD to be dancing around making the promise already.
Besides the 105W AM4 TDP limit, they would probably prefer to use all the top-clocking fully intact compute chiplets in more profitable Threadripper and Epyc parts.

12 and 16 core parts won't have the same high clocks that the 8 core parts can reach. They might have the same peak XFR and single core, but sustained multicore pstates won't be equal.
They're not putting top tier dies in any consumer desktop part.

preach it

Attached: Screen-Shot-2017-11-06-at-3.04.25-pm.png (677x423, 344K)

Attached: relative-performance-games-2560-1440.png (500x1490, 93K)

DELID

Attached: mayo_i7_guy.png (864x1006, 229K)

>2019
>Not being a blood

it's a reflection
amdrones are hoping it's something big, lmao

>Now imagine how fast the 105w x version will be.

Attached: 1547697586525.jpg (218x250, 7K)

>ES of a midrange CPU
>top tier silicon and top tier CPU
>tied
>haha r5 is tied with this 700$CPU, AYYMD BTFO
You must've won a gold medal at mental gymnastics

>it's a reflection
A reflection that just happens to have the exact layout of a 7nm chiplet? Is this bait? This is bait, right?