It's time to lube up, boys. Dual socket 64C/128T setup,

where are the RGB leds?

>>59593208
http://www.oracle.com/us/products/servers-storage/servers/sparc/oracle-sparc/t5-8/sparc-t5-8-ds-1922874.pdf

>>59593208
Why are there so little pci slots and sata ports?

>>59593268
>SPARC

How about an architecture that's good at more than just handling databases and being a fun Fujitsu project

>>59593305
sorry, cant hear your poorfag bitch whinging over the 4TB and 1024cores

>>59593296
Not relevant for the market or might just be a engineering board.

Though it has Infinityband support so it's probably not a eng board, so possibly not needed for this rollout

>>59593208
>only supports DDR3 RAM

Gayyyy, it's like they want my ARMA3 server to be slow.

>>59593401
Are you mentally deranged?

>>59593401
kys

File: IMG_0417.png (594KB, 600x449px) Image search: [Google]

594KB, 600x449px

>red pcb

>B0 stepping

So we can expect higher clocks?
Though with 32 cores I wasn't really expecting them going over 2.4-2.6GHz

>>59593208
how much would the electric bill be, tho

does anyone else remember that faggot that was happy he got some server really cheap off ebay or something, and them came back a month later whining his electric was like 700 bucks?

>>59593635
naples will have lower tdp than those 16 cores prosumer ryzens(that are rated at 180W) will have, so not that much even when not accounting the performance

>>59594185
MOAR CORES won.

>>59593208
if moar coars are so import in the server industry
why aren't GPUs more widely used?

>>59594843
Because you can't replace a truck with a Ford GT

>>59594843
They are. I met a guy who maintained a nvidia cluster for gas and oil industry

>>59594843
GPU's are shit at branching and seriously memory limited
But they're seeing a lot of adoption wherever they can be used

>>59594893
makes a lot more sense desu.

>>59594891
cpus are supposed to be about IPC and floating point calcs, for parallel computing cpus don't come close to gpus. you analogy is retarded.
>hur dur you wouldn't download a currr

>>59594843
try transcoding a video on the gpu and then on the cpu, tell me which one looks better

>>59594843
Gpus are able to handle about 30% of the work that a server would be used for, cpus cover the other 70% and the gpu load is inflated with phi being a gap between needs smarter then a gpu but also needs fuck load of cores.

There may come a time when gpus find a sweet spot between we are smarter then old gpus but dumber then good cpus, but the many core design lets them take over completely.

>>59593296
It's a server motherboard.

I have a server with similar specs in my data center. It's a esxi host for several database servers.