AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review • Hacker News Board

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#41

Post by Hasknewbie » Mon, Aug 13, 2018, 2:47 PM UTC

MrUnderhill wrote: post #20

karangoeluw wrote: post #2

Honest q: What home-use workstation would use 32 cores? (Excluding home labs or servers).

For compiling, having many cores is fantastic. Granted, on a workstation, compilation normally just involves a few files (the ones that have changed since the previous build and their dependencies), but when you have to do a full rebuild, it is fantastic to be able to do `make -j16` and watch it chug through 16 files simultaneously. Interestingly, the benchmark in this review shows that the 16-core 2950X compiles Chr…

"this review shows that the 16-core 2950X compiles Chromium faster than the 32-core 2990WX, presumably this means something other than the thread count becomes a bottleneck after 16 threads"

The article mentions that, due to the die packaging, only 16 of the cores have direct access to RAM. So for the 32-core version, half the cores are memory-starved and have to go through the 'connected' cores (also impacting these), while the 16-core version doesn't have that problem and can be at 100% for all process loads.

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#42

Post by elorant » Mon, Aug 13, 2018, 2:52 PM UTC

karangoeluw wrote: post #2

Honest q: What home-use workstation would use 32 cores? (Excluding home labs or servers).

I'm thinking of buying one to run multiple instances of Selenium with headless Firefox for crawling reasons. Pair it with 128GB RAM and I could easily run 40-50 instances simultaneously.

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#43

Post by Eochei5h » Mon, Aug 13, 2018, 2:53 PM UTC

blattimwind wrote: post #12

> After core counts, the next battle will be on the interconnect. Low power, scalable, and high performance: process node scaling will mean nothing if the interconnect becomes 90% of the total chip power. I think this means that scaling processors to more cores (nodes) in a lightweight-NUMA scheme just isn't going to work. My guess going forward is that the core counts will stay at the current level for quite a while…

Any wins they make in interconnect efficiency could either be used to boost the power budget of existing cores or add even more cores.

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#44

Post by deng » Mon, Aug 13, 2018, 2:53 PM UTC

srcmap wrote: post #40

Two additional benchmarks I like to see from Anandtech: 1) some VM base tests for these kind of CPU - 32, 64, 128 VM all running some kind of web/db/redis benchmarks inside. 2) Some compilation testing - time the clean build of AOSP, BSD, some very complex linux app - use max jobs setting for parallel compile and time how long it take to finish the jobs. ( measure the over CPU usages at the same time. )

They do a compilation of Chrome in the "Office Tests" section. The 32Core takes about an hour per compile, but the 16Core is actually a bit faster (my guess would be that linking takes longer on the 32 core machine as it's very memory intensive).

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#45

Post by barkingcat » Mon, Aug 13, 2018, 2:53 PM UTC

drewg123 wrote: post #24

It is interested that the review is not yet finished. The author writes in the comments section: Hey everyone, sorry for leaving a few pages blank right now. Jet lag hit me hard over the weekend from Flash Memory Summit. Will be filling in the blanks and the analysis throughout today. I am disappointed in that, as I was looking forward to reading the test setup and power draw sections, as I have a 2990WX on order, an…

Shouldn't they not publish until the review is done? This is really unprofessional - even bloggers have the concept of "only publish finished drafts" or use the scheduled publishing feature that all blogging platforms have nowadays.

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#46

Post by MikusR » Mon, Aug 13, 2018, 2:53 PM UTC

srcmap wrote: post #40

Two additional benchmarks I like to see from Anandtech: 1) some VM base tests for these kind of CPU - 32, 64, 128 VM all running some kind of web/db/redis benchmarks inside. 2) Some compilation testing - time the clean build of AOSP, BSD, some very complex linux app - use max jobs setting for parallel compile and time how long it take to finish the jobs. ( measure the over CPU usages at the same time. )

There are Chromium compile tests. https://www.anandtech.com/show/13124/the-amd-threadripper-29...

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#47

Post by Mankaninen » Mon, Aug 13, 2018, 2:55 PM UTC

karangoeluw wrote: post #2

Honest q: What home-use workstation would use 32 cores? (Excluding home labs or servers).

I think more CPU performance is attractive if you edit photos or movies - something that is typical for a home computer. The reason AMD provides CPU performance through a high core count is because it is power efficient, otherwise a single-core is easier to program.

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#48

Post by snovv_crash » Mon, Aug 13, 2018, 2:55 PM UTC

ekianjo wrote: post #32

nv-vn wrote: post #16

Earlier quoted context omitted.

The 2990wx blew everything out of the water in rendering. It's like 37% faster than the 7980xe in the Blender benchmark.

Faster with 2x more cores and twice more power consumption. Hardly a win in my book - it means each core is way weaker and power hungry. Intel will easily match that sometimes soon without sweating.

How much power would an Intel based system need to get that 40% performance boost?

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#49

Post by CarVac » Mon, Aug 13, 2018, 2:55 PM UTC

ekianjo wrote: post #32

nv-vn wrote: post #16

Earlier quoted context omitted.

The 2990wx blew everything out of the water in rendering. It's like 37% faster than the 7980xe in the Blender benchmark.

Faster with 2x more cores and twice more power consumption. Hardly a win in my book - it means each core is way weaker and power hungry. Intel will easily match that sometimes soon without sweating.

If you check Phoronix, they found it doesn't actually take much more power than the 7980XE.

Re: AMD Threadripper 2990WX 32-Core and 2950X 16-Core Review

#50

Post by CorpOverreach » Mon, Aug 13, 2018, 2:56 PM UTC

ekianjo wrote: post #29

zaarn wrote: post #22

Earlier quoted context omitted.

VMs, lots of them. Atm I'm running about 15 VMs on about 8 cores in a dedicated box somewhere and it's definitely noticable. I would love to shove some core services at home and have 32 cores to play with to give some more headroom

Why not use containers instead of VMs? You can run about 10x more Docker instances than VMs on the same hardware.

Because all of the containers may not be of the same operating system? Networking on containers is also a bit different.

There are also reasons for having some more isolation between guest OSes.

On my ESXi box at home I have:

* A VM that hosts my NAS shares. This does nothing other than host the NAS shares, as I want to be sure no silly experiment of mine interferes with that.

* A general-purpose VM, where I do run some containers out of (UniFi controller, Plex, etc)

* A VM running Windows Server for my Domain Controller

* A secondary vSwtich with isolated no uplink to the rest of the network. This is my mini malware testing lab.

* A VM running pfSense that I'll sometimes use to allow selective access out of the isolated vSwtich out to the internet, but not to the rest of the network.

Can't do all that with containers.