>yfw zen+ is ahead schedule Ryzen 2000 series End of

HES SO POOR HE USES AMD OMFG LOOK AT HIS TINY INVERTED COCK

>>61588765
>ahead of schedule
Fixed.

File: 1500517074521.png (124KB, 358x272px) Image search: [Google]

124KB, 358x272px

>>61588765
It better has the fucking 16MiB of L3

>>61589944
It won't, Raven Ridge is gonna top out at 4/8 which would mean 8MiB L3 is the max.

File: 564894.jpg (42KB, 680x399px) Image search: [Google]

42KB, 680x399px

>>61588765
That's Raven Ridge you mug. Pinnacle Ridge (Summit Ridge on an improved process) has always been Q1 2018.

>>61588765

>ryzen 5 2500u

what

>J-just wait guys! Intel is finished

It is a mobile apu the U is from the intel nomenclature.

>>61590454
the 1500X has 16MB L3 dummy

this is obviously spoofed

File: p.png (16KB, 646x230px) Image search: [Google]

16KB, 646x230px

>>61590454
>8MiB L3 is the max.
No it's not

>>61590614
Yep it does, 2 CCX's with full L3 cache is 16MiB, I dont think AMD will be doing 8/16 APU's though and we would get APU's with 1 CCX and an IGPU, that caps the CPU at 4/8 with 8MiB L3 cache.

>>61590656
>RAVEN RIDGE
You fucking moron.

>>61590669
Oh I forgot, L3 is bound per CCX.

Not having the interconnect latency could help on performance though. 1400s and 1500s are 2+2.

>>61590765
Oh it definitely will, thats why im so confident in AMD only doing 1CCX APU's for now, they will want Raven Ridge to be as good as it can possibly be so risking a performance hit because of latency would be a bad move.

File: IMG_0588-a7bcf5ec93130aef[1].jpg (30KB, 695x391px) Image search: [Google]

30KB, 695x391px

>>61590800
They've already shown Raven Ridge.
Assume 60% of the die is dedicated to GPU as it tends to be, that leaves enough room for a single CCX complex to squeeze in

>>61590800
It will be interesting to see whether a larger L3 or lower latencies net a bigger performance uplift.

>>61590833
Oh dang, didn't see that. Wonder how big the IGPU is gonna end up being?

>>61590941
>Assume 60% of the die is dedicated to GPU as it tends to be, that leaves enough room for a single CCX complex to squeeze in

>>61590941
It's 768 Vega shaders.
It's gonna be pretty fucking big.

>>61590911
Someone has probably been able to see the difference between using the R5 1500x and the 1400, one has half the cache disabled so if you clocked em the same we could find out if the larger L3 does much for Ryzen.

>>61590989
704. 11 CUs

>>61591038
No 12CUs?
Why tho.

I'm curious to how many pcie lanes it has, zepplin goes with 32 (am4 using 16gpu+4nvme+4pch) and bristol ridge with 16 (8gpu+4nvme+4pch). So in theory am4 only uses 24. but it could follow the same amount as BR and save die space/power for mobile applications, on the other hand amd says their power gating on the IO controllers is 100%.

>>61591038
That'd make a pretty damn nice SFF system right there if we can get the same 1450mhz+ outta the Igpu.

>>61591084
It's Vega shaders, so yes.

>>61591055
My guess is there are 12 on die, one disabled for power/yields and desktop edition could get 12.

File: 23156.gif (49KB, 338x200px) Image search: [Google]

49KB, 338x200px

>OEMs will pair this with single channel RAM and 768p TN panels

>>61591485
THANK YOU BASED INTEL.
PLEASE KILL THE AYYMDEAD.

>>61591485
There's already a laptop from ASUS with a full desktop R7 1700 in it. Ryzen Mobile is not just going to be relegated to craptops this time around.

http://www.pcworld.com/article/3199028/computers/asus-debuts-the-first-ever-ryzen-laptop-with-a-mobile-radeon-surprise-too.html

>>61591485
>1366x768
came here to post this

File: suicide.jpg (7KB, 250x200px) Image search: [Google]

7KB, 250x200px

>>61591485
shit timeline

>>61591485
AMD doesn't deserv better, it deserves to die!!!

>>61588765
>U nomenclature

15-19W confirmed