>Xeon 8176
>Using AVX
>Still loses to Ebyn
http://www.anandtech.com/show/11544/intel-skylake-ep-vs-amd-epyc-7000-cpu-battle-of-the-decade/21
AHAHAHAHAHAHA
7700K still better than Ryzen
>>61441086
>a-at least we still got high clockspeed quad cores ..
lol. Intel is giving up on those with Coffee Lake-S anyway.
>>61441086
*In muh vidya
*In CPU bottlenecked situations
*When nothing else is running in the background at all
>>61441021
K E L L E R
E
L
L
E
R
>>61441021
BUT IT'S GLUED TOGETHER GOY, BUY OUR PREMIUM QUALITY INTEL™ XEON™ SCALABLE™ PROCESSORS™
>>61441021
HOW CAN ONE TECH GIANT BE SO BTFO???
SHINTEL ON
S U I C I D E W A T C H
>>61441021
Core count is not same idiot. Price is different story.
>>61441266
I really hope you aren't implying Intel does better on price.
>>61441266
But zen2 will have one or more of:
more ccx's
more cores per ccx
smaller dies thus more dies per chip
AVX512
Either way Intel will still be behind.
>>61441266
>much cheaper
>moar coars
>shitload of PCIe lanes
>no artificial RAM limits
>uses less power
>beats the Intel alternative in most workloads
It's a no brainer, there's no point in buying the Intel alternative.
>>61441347
Bringing more cores to dual channel socket is not gonna work well, that's why Threadripper exists.
Intel is not a toy corporation, they finally learned the lesson at least partially with their bingbus.
>>61441266
>14% more cores
>Still performs 22% better
>>61441086
>check out these 800x600 benchmarks of 8 year old games!
>>61441443
>Intel is not a toy corporation
Yes, they make good SRAM (and sometimes good NAND). Also NICs.
>>61441443
Yeah, now they're using BingMESH. It's measurably shittier in consumer applications, but at least the latency memes aren't true anymore.
>>61441898
>the latency memes aren't true anymore
Yeah, now the latency is always shitty.
>NAMD
Where is the NINTEL one?
>>61441898
>>61442192
https://www.pcper.com/reviews/Processors/Intel-Core-i9-7900X-10-core-Skylake-X-Processor-Review/Thread-Thread-Latency-and-
Yep.
>>61441021
How does Zen have such godtier floating point performance?
>>61442716
>How does Zen have such godtier floating point performance?
They focused on the best possible way to feed FPU instead of making XBAWKS HUEG vectors.
They've learned well from both Larrabee & Bulldozer mistakes.
>>61442716
I'd rather trade a 30% faster FPU for a 15% faster integer units
FPU is nice, but near 80% datacenter workloads are integer.
>>61442787
Zen is also great at INT.
It's a great all-around core, probably the best x86 core up to date.