Ive been really enjoying my dual epyc 64 core workstation. Looking forward to when I can upgrade to 128 cores. :)
what do you do with all those cores? if i may ask
Simulations and randomized testing.
AMD’s Rome is indeed a monster
81–90 of 104 posts
Re: AMD’s Rome is indeed a monster
#82Earlier quoted context omitted.
HBM2 does have dramatically more bandwidth than DDR4.
It has more bandwidth but about the same latency. CPUs fundamentally attack memory latency differently than GPUs do, and raw bandwidth is less important.
It is extremely important when you have 64 cores orif you use wide SIMD instructions. Intel's Xeon Phi have 16GB HMC memory with 500GB/s bandwidth to prevent the CPUs from being starved by main memory.
Imagine you have a cache miss on literally every memory access and have to pay a 100 nano second penalty to access main memory. SMT allows the CPU to switch to a different thread during a cache miss. Effectively you can have 128 memory loads at any given time. 100ns amortized over 128 threads is less than 1 nano second. Since the minimum size you can read is a 64 byte cache line this means that program will require at least 64GB/s of memory bandwidth even though your CPU is stuck on cache misses.
Re: AMD’s Rome is indeed a monster
#83Not particularly related to Rome, but what you can buy right now for a single socket system, AMD is far ahead of Intel for reasonably priced workstations or small servers with a huge amount of I/O. One threadripper CPU has 64 pci express 3.0 lanes. One lane is 985MB/s. Working with a $399 threadripper motherboard that has four x16 physical slots, it can accommodate four Intel x710-4 10GbE four port NICs (each electri…
I looked into this in the past but had issues finding information about processing overhead for routing (or even switching) that much data. Could a, say, 2950 handle routing 40 gigabit worth of data over a network, assuming no crypto?
Easily. One modern core can route 10 Gbps or more using efficient software like VPP.
Re: AMD’s Rome is indeed a monster
#84Oh man, the licensing costs for any “per core” software are going to be insane :D
A few years ago many companies changed it to "per socket". Now with 64 cores in one socket things got even weirder, but then again, those licenses always have been in some parallel universe.
There's still a disturbing amount of software that operates on a per-core pricing model. Some of them include Hyperthread in that core count.
When we launched Oracle Cloud Infrastructure, (as Bare Metal Cloud, before we launched our VM product), our standard bare metal instance had 36 cores, 72 threads with HT enabled. One interested customer couldn't use our platform because just licensing their standard software on our platform was going to set them back some $250,000 due to all those 72 "cores". The product manager at the time mentioned there were several other customers facing that issue with their various bits of enterprise software.
Re: AMD’s Rome is indeed a monster
#85Question, why is DRAM still packed separately?
A different question is why is DRAM still colocated with the CPUs, and not centrally located over a very fast network? There was a company doing this (RNA Networks, founded 2006) which was sold to Dell and then apparently disappeared without trace. It might make more sense now that there are more types of memory-like storage which isn't quite as fast as DDR4, like Optane.
Attempts to do this are being made in academia. Some recent work tries to quantify the network requirements of this, and decides that it's far more feasible to use some local memory as a caches than to have everything be remote.
Network requirements: https://www.usenix.org/conference/osdi16/technical-sessions/...
Two systems that use remote memory: https://www.usenix.org/conference/nsdi17/technical-sessions/... https://www.usenix.org/conference/osdi18/presentation/shan
Re: AMD’s Rome is indeed a monster
#86Ive been really enjoying my dual epyc 64 core workstation. Looking forward to when I can upgrade to 128 cores. :)
Did you build that yourself? Do you know of any good vendors for this type of workstation? I know building it isn't that hard, but typically an employer doesn't want to go that route.
Silicon Mechanics can build just about anything with your choice of Supermicro motherboard.
Re: AMD’s Rome is indeed a monster
#87Earlier quoted context omitted.
They don't like Intel or Nvidia, but their facts tend to be mostly accurate. I wouldn't listen to any of their financial or market-share predictions since the market has remarkable inertia that disconnects it from technical factors.
Accurate facts, the best kind of facts.
I’m feeling almost nostalgic.
Re: AMD’s Rome is indeed a monster
#88Earlier quoted context omitted.
I looked into it for a server but I wasn't able to find a motherboard with remote management features. I suspect there might never be one so as not to muddy the waters of Epyc's intended market.
That's so frustrating. We ran into the same roadblock.
I’m sorry, you’re saying there’s no boards available with IPMI?
Re: AMD’s Rome is indeed a monster
#89I’ve been wondering lately if the reason for the Mac Pro replacement not coming out until 2019 may be due to Apple switching to the 7nm refresh of the AMD ThreadRipper. That would certainly make for some interesting news.
Apple can't use Thunderbolt with AMD, and Intel has been delaying its opening of Thunderbolt standard for whatever reason. At this point I think Intel has been so disappointing in the past 3 years Apple should just dump them and takes matter into their own hands.
Re: AMD’s Rome is indeed a monster
#90Every time I see SemiAccurate I remember their love for AMD and their constant, unending hate for Nvidia: https://semiaccurate.com/2009/10/01/nvidia-fakes-fermi-board... I don't have any dog in this fight, SA is just a terrible news source.
>SA is just a terrible news source.
Yes and does not belong to HN.