AMD Ryzen Threadripper 3000 32-core CPU is more bad news for Intel • Hacker News Board

Re: AMD Ryzen Threadripper 3000 32-core CPU is more bad news for Intel

#161

Post by rasz » Mon, Sep 23, 2019, 8:31 PM UTC

astrodust wrote: post #159

rbanffy wrote: post #32

Earlier quoted context omitted.

That's true, but memory/IO is a much bigger issue in this scenario than processing power. While running a cluster in VMs, I'm frequently limited by IO (and if your VMs use more memory than it's physically available, you'll quickly descent into swap hell) well before my CPUs register a significant uptick in usage. And, mind you, my workstation is 4 year old Xeon with 64 gigs of memory and reasonably fast (but not amaz…

Consider: NVMe makes SSD look like HDD. More PCIe channels means more NVMe as most servers don't use those channels for GPU.

>NVMe makes SSD look like HDD.

why? Its not true. Program/data load times change by single digit % between SATA and NVME SSDs.

Re: AMD Ryzen Threadripper 3000 32-core CPU is more bad news for Intel

#162

Post by celrod » Wed, Sep 25, 2019, 2:11 AM UTC

0-_-0 wrote: post #153

celrod wrote: post #128

Earlier quoted context omitted.

You can look at binning statistics for non-avx/avx2/avx512 clock speeds: https://siliconlottery.com/pages/statistics For example, the worst 7980XEs do 4.1/3.8/3.6 GHz for each of these respectively. 0.5 GHz down clock isn't too bad. You can change these settings in your bios however you'd like on the unlocked CPUs (ie, the HEDT lineup + W3175X). I do find those down clocks are necessary; I've passed 100C with a 360mm…

> I've passed 100C with a 360mm radiator and roaring fans Damn, which radiator? Was there a GPU under load in your water loop?

Celsius S36. My GPU is on a separate loop and was idle at the time.

I was running benchmarks of Intel MKL's zgemm vs zgemm3m because of a Julia PR that recommended replacing the former with the latter. I don't think anything hits a CPU quite as hard as a good BLAS.

I think my thermal paste may be bad, because the CPU idles hot -- nearly 35 C. I ordered a Direct Die-X and MO-RA-420 radiator, so I'm planning on swapping the AIO for an open loop with way more radiator area and flow through the fins.

Running dgemm, that CPU would hit just a tad below 2 teraflops. I'd like to get it just over that (and run much cooler).

Re: AMD Ryzen Threadripper 3000 32-core CPU is more bad news for Intel

#163

Post by astrodust » Wed, Sep 25, 2019, 4:40 PM UTC

rasz wrote: post #161

astrodust wrote: post #159

Earlier quoted context omitted.

Consider: NVMe makes SSD look like HDD. More PCIe channels means more NVMe as most servers don't use those channels for GPU.

>NVMe makes SSD look like HDD. why? Its not true. Program/data load times change by single digit % between SATA and NVME SSDs.

2GB+/s vs. ~600MB/s is a pretty big deal for some workloads.

Re: AMD Ryzen Threadripper 3000 32-core CPU is more bad news for Intel

#164

Post by rasz » Wed, Sep 25, 2019, 4:48 PM UTC

astrodust wrote: post #163

rasz wrote: post #161

Earlier quoted context omitted.

>NVMe makes SSD look like HDD. why? Its not true. Program/data load times change by single digit % between SATA and NVME SSDs.

2GB+/s vs. ~600MB/s is a pretty big deal for some workloads.

For precisely one "lets copy a lot of data around" workload. In real world you wont feel the difference even when editing 4K movies.

Re: AMD Ryzen Threadripper 3000 32-core CPU is more bad news for Intel

#165

Post by kraig911 » Fri, Sep 27, 2019, 1:48 PM UTC

benj111 wrote: post #46

kraig911 wrote: post #38

Now if only programming multi-core work could catch up to this hardware. Honestly chips with more cores is starting to sound like cars with more wheels or I guess more valves lately. Doesn't matter the horsepower the speed limit is the same until we can make software better.

"or I guess more valves lately." Extra valves can increase efficiency, which is probably why the valve count is increasing, can't say I've noticed though. Extra wheels can improve aerodynamics also. https://en.m.wikipedia.org/wiki/Tyrrell_P34 I'll stop undermining your point now :)

Efficiency does not mean you're going faster though. My point still stands.

Multi-core processors can invariably let us do more. I get it - valid point. But we're not doing more. We're just cruising along at the same speed I think.