The Ryzen 6 core Linux benches. Also includes the 8 cores fo

File: AMD-Ryzen-5-UnixBench-dhrystone-2.jpg (80KB, 856x634px) Image search: [Google]

80KB, 856x634px

>>59834983
Erre's some dhrystone.

>>59834983
Also where the fuck does this guy get all that server hardware.

>>59834983

Intel just BTFO.

>>59834983
The 1600 and 1700 are really best value, come with a 130W cooler and are cheaper, but can OC as much as the expensive X chips, who come without a cooler.

File: 1485091759559.png (279KB, 500x500px) Image search: [Google]

279KB, 500x500px

>>59834983
>>59835017
>ryzen quad core beating the 7700k
Absolute horse shit.

>>59835417
There's no 7700k tested you blind goof.

>>59835451
It implies that it scored too low to appear on the list.

>>59834983
>>59835017
I'm looking at these and I see 1500X > 1800X scaling is almost linear from twice the cores, that usually isn't the case with Intel even in highly threaded applications.

This is pretty impressive, Naples will be a monster.

>>59835529
Doesn't imply anything, the guy running these tests is a server seller and he doesn't deal have a 7700k to test because a 7700k is relatively useless to his work.

Here's the guy that runs STH, good long standing friends with Ian from anandtech.

http://www.anandtech.com/show/11169/the-anandtech-podcast-episode-41

File: 1489080074975.png (247KB, 881x757px) Image search: [Google]

247KB, 881x757px

>>59835417

>>59835559
00:15 – Patrick’s 2000 cores
01:41 – Atom C2000 Avoton/Rangeley Hardware Bug
09:22 – Denverton and C3000
15:17 – Xeon D-1500 Networking CPUs
18:02 – Opportunities for AMD Naples


Wow, an actual interesting to listen podcast without phone garbage, I'm in.

>>59834983
>Xeon-D 1587
> Recommended Customer Price $1754.00


Lmao

>>59835643
Pay for 16 corz you dumb goy
its a xeon look look shinny ECC and 10GbE, dumb goy you can NEVER have that as a consumer!

>>59835696
Yes I can, it's called a 1700 and any regular 10GbE NIC you can find on ebay, which is about 5 times cheaper.

>>59835559
As if those shitty little Ryzens can run a server.

>>59835767
A dual core can run a server, doesn't mean it can run one well.

File: AMD-Naples_Infinity-Fabric.jpg (309KB, 1510x460px) Image search: [Google]

309KB, 1510x460px

>>59835533
AMD weren't joking the fuck around.

>>59834983

Every single day I see these charts (especially the newest ones), the 1700 looks like the top buy for the dollar.

>>59836207

What do you think about the 1600?

>Linux kernel 4.4.2
Didn't AMD made some contribution on recent linux release ? Can it impact the performances of ryzen ?

>>59838019
Yes, but it's a slight increase.
Zen had linux patches a while ago

>>59838019
yes; run an up-to-date kernel. these test boxes are running 4.10.1, btw.

>>59835806
they can't joke anymore, screwing up Ryzen could mean bankruptcy for them.

>>59836552
Not the quoted guy, but best performance to value chip atm. Especially if you're not shy about overclocking. All the Ryzen chips are fundamentally the same and can be pushed to at least 3.9Ghz based on most testing. (The upper end X models can usually hit 4.0 or even 4.1, not really worth the extra cost IMO.) Global Foundries (The makes of the Ryzen Dies) has a damn efficient process and there doesn't seem to much binning due to defective silicon.

*Might* want to invest in a beefier cooler though as they will get a bit warm if you do OC to 3.9.

>>59840438
Do you think a Noctua Nh-15s would be sufficent for a Ryzen chip to go 4.0?

>>59840776
Yes, that thing is overkill. Also a pain in the ass to install. But it'll do the trick.