lhl

63 ms·

631.
by lhl 7y ago
That benchmark page I referenced shows scaling on a SkylakeX at 1, 26, and 52 cores, on a Haswell at 1, 12, and 24 cores, and on Zen1 at 1, 32, and 64 cores so you can probably map out a ballpark scaling co-efficent from that. I'm sure
632.
by lhl 7y ago
Oh, just for fun, I manually replaced the BLIS libs w/ the AMD BLIS 2.0 fork (Ubuntu MT binary on an Arch system: https://developer.amd.com/amd-aocl/blas-library/ ) and it seemed to work ok. Slightly faster th
633.
by lhl 7y ago
Everything on that benchmark page is listed as GFLOPS/core so at 30GFLOPS that Epyc 7551 should be pushing around 2 TFLOPS. You can take a look at reproducing them if you want if you have a similar system: https://github.com
634.
by lhl 7y ago
You can actually see in my pasted command that I created a new venv for the blis install. It's as simple as "activate" or "deactivate" to switch around. That's basically all there is to it, but here are some do
635.
by lhl 7y ago
It's certainly not an "obscene edge case" - MKL cripples AMD on every single matrix operation tested here: https://github.com/flame/blis/blob/master/docs/graphs/large/...
636.
by lhl 7y ago
The article was specific to AMD's BLIS fork, but just as an FYI, using standard BLIS w/ numpy (which I believe would still outperform OpenBLAS) is pretty straightforward: conda create -c conda-forge -n numpy-blis numpy "b
637.
by lhl 7y ago
Based on Zen1 Epyc performance it would seem that BLIS might be significantly more performant even without AMD-specific optimizations (I know AMD has their own fork, but check out the 64 core results for "vanilla" BLIS: https:&#x
638.
by lhl 7y ago
As mentioned, plenty (maybe even most!) vegetables are not very high in digestible carbs - certainly not the ones that are high in fiber and nutrients (you can probably skip the potatoes unless you're on a mono-diet). But this is where
639.
by lhl 7y ago
Very few vegetables (starchy tubers are the main exception) contain much sugars (and specifically fructose, which is uniquely bad for your liver), so I think it's worth separating out fruit vs vegetables, but yes, there's also a v
640.
by lhl 7y ago
The body of work, while still emergent, is fairly robust IMO if that's what you're asking. I've cherry picked just a few out of hundreds of papers in my personal archive related to this topic, but if you have an interest you
641.
by lhl 7y ago
The top papers I included are purposefully recent as they are recent reviews on an emerging area of research. Having read a lot of papers this year, my conclusion btw is actually that recent papers are much better than older papers. One, ne
642.
by lhl 7y ago
Sleep (and related, circadian health) are some super interesting topics of emerging research. For those that like books, there are two recent ones written by researchers in the field. Matthew Walker's Why We Sleep, and Satchin Panda&#x
643.
by lhl 7y ago
There's a huge difference between the acute inflammatory response (obviously good and necessary for life) and chronic inflammation (very bad). The biology is hideously complex and not well understood. This past year I've gone thro
644.
by lhl 7y ago
Alternatively, you can quite dramatically reduce inflammation in the brain via some simple lifestyle modifications (cutting out sugar, intermittent fasting): Mattson, Mark P., Keelin Moehl, Nathaniel Ghena, Maggie Schmaedick, and Aiwu Cheng
645.
by lhl 7y ago
The issue with the press backlash was that originally the press embargoes for both the AMD and Intel HEDT parts were both set for the same day and time (Mon Nov 25 @ 9AM Eastern) but then a few days before embargo lift, Intel shifted their
646.
by lhl 7y ago
One of my minor goals for 2020 is to switch all the gear I carry around to being USB-C PD. I've been doing some research on the subject, so just thought I'd throw my 2 cents in (some of this stuff mentioned in other comments). The
647.
by lhl 7y ago
I agree that carbon footprint should be something that people keep in mind when they're eating. What's worth noting though is that regenerative agriculture (where the goal is to be sustainable and can be carbon net negative) requi
648.
by lhl 7y ago
For those not clicking through on the paper (it came up in discussion in one of my metabolomics groups the other day) the paper you cited shows that (previously unknown) that there would be microbiome adaptations that could mitigate previou
649.
by lhl 7y ago
Last year the NYTimes had a nice writeup about her https://www.nytimes.com/2018/12/05/sports/courtney-dauwalter... - I was following up on that and found an interview she did with Joe Rogan in 2017 after
650.
by lhl 7y ago
Even for melanomas, the dangers of sun exposure have been misguided. It turns out getting regular sun exposure (not burning) appears to actually lower your risk: "Overall the clearest relationship between reported sun exposure and ri
651.
by lhl 7y ago
I actually skimmed through the paper but wasn't able to determine if it did take the weight of the exo into account. Do you have a reference for how it specifies that? Also, I saw the BOM in the supplementary material (using 2 x 500g 3
652.
by lhl 7y ago
For those looking to learn more the full study is linked from the NPR report: https://science.sciencemag.org/content/365/6454/668 "We show that a portable exosuit that assists hip extension can reduce th
653.
by lhl 7y ago
The point is that conventional agriculture (eg, with industrial inputs) is literally unsustainable - there is a 56% food gap in the coming decades. [1] Also, there's plenty of evidence that small, sustainable farms are both more effici
654.
by lhl 7y ago
Almost all conventional/industrial agriculture (plant or animal based) is unsustainable due to the external inputs required (70% from fossil fuels for growing corn! [1]), soil depletion incurred [2], and just plain shipping things arou
655.
by lhl 7y ago
Yes, although it's not entirely straightforward and there are many sharp edges, so it's maybe not for the faint of heart. Since I just did it, here's a guide I wrote up with my specific hardware and some of the issues I'
656.
by lhl 7y ago
Actually, it's AMD Vega cards that have a PCI reset issue that doesn't work well with VM cycling. There's a new kernel patch just the other week that fixes this: https://forum.level1techs.com/t/vega-10-an
657.
by lhl 7y ago
ECC support is built into the IMC (on the CPU) so de facto, every single Zen chip (except the non-PRO APUs) usually have support, but official support is still on a per motherboard basis (and you'd probably want to check the memory QVL
658.
by lhl 7y ago
I think STH's writeup does a particularly good analysis here: https://www.servethehome.com/amd-epyc-7002-series-rome-deliv... The second is important. Customers need to adopt AMD EPYC. To our readers, it is important
659.
by lhl 7y ago
AMD has uProf [1], which while maybe not at full parity, definitely does performance counters (along with energy and power profiling, call graphs, and the other basic things you'd expect). I use ECC on a (consumer) Ryzen chip/boar
660.
by lhl 7y ago
While the headline is a bit clickbaity, for the retail CPU market, I don't think it's actually that unrepresentative. Here's reports from Japan of AMD overtaking Intel in retail CPU market share in July for the first time eve

More ›