0xbear

61 ms·

601.
by 0xbear 9y ago
That actually sounds pretty badass to me. Legally naive, but cool idea and execution. Kudos to Elaine.
602.
by 0xbear 9y ago
I actually think it just loses your choices completely on its own and local news thing is unrelated. If you add them a few more times, they stick.
603.
by 0xbear 9y ago
This is not very well known in the US, although the Russians were sore about it for quite a while. This was back in early 90s when the US was universally perceived as a "friend" by freshly ex-Soviet people for the first time in th
604.
by 0xbear 9y ago
Let's not get too cocky here. Copying went both ways: https://arstechnica.com/science/2016/01/nasas-newest-cargo-s... And the pork barrel project extraordinaire: http://aviationintel.com/
605.
by 0xbear 9y ago
I remember reading somewhere that those shuttles were in an old hangar that collapsed on top of them. Looks like that wasn't true, this was definitely shot fairly recently, drones and all.
606.
by 0xbear 9y ago
Protip: identify the news sources you don't like and ban them wholesale under "Manage Sections / Sources" in Google News. Banning just a few improves the experience tremendously. HuffPo and Buzzfeed are prime candidates
607.
by 0xbear 9y ago
Not really, no. Once you give up Slack it's a much better place to work at. Thankfully the use of Slack is not mandatory here.
608.
by 0xbear 9y ago
Missing from the list: it encourages group conversation and therefore pointless (and often damaging) flame wars where opinions differ. At one point or another some otherwise decent human being starts using it as a platform to spread their p
609.
by 0xbear 9y ago
Hopefully there isn't much to do to make it run on H5. H5 is a much better CPU
610.
by 0xbear 9y ago
You need to be 62 as well.
611.
by 0xbear 9y ago
Increasing cache has the net effect of increasing hit ratio, sometimes substantially. With 20MB per die this may change the calculation of where things drop off. I have found that I can't reliably predict how a chip will perform, so I
612.
by 0xbear 9y ago
I have never encountered such an abnormal workload. This is also less likely to happen in Broadwell Xeon and up, where last level cache can be partitioned. And this is also less likely to happen on Google Cloud in particular, because Google
613.
by 0xbear 9y ago
Most other distribution methods for those artists have also been shown to net vanishingly little money. If you can't make money with a live performance, you sure as heck aren't making any selling CDs through a record company.
614.
by 0xbear 9y ago
For training, that's likely to be true. For large scale inference it's not possible to beat CPUs right now if cost is a factor. You might be able to beat them once you can buy TPU access in cloud, depending on how steep a premium
615.
by 0xbear 9y ago
Also to add to the article: I have also discovered that for our deep learning workloads 8 core VMs are the sweet spot in terms of cost/perf. This is on Google Cloud, which in the particular zone I tested uses $5k apiece high end Broadw
616.
by 0xbear 9y ago
Hyper threading is not harmful per se. It lets your CPU make forward progress when it would otherwise be stalled waiting for something. My issue is that they call hyperthreads "vCPU" which makes it seem like you're getting a
617.
by 0xbear 9y ago
This is especially important if most of your workload is matrix multiplication. Those workloads heavily benefit from vectorization. It might also help to enable Intel MKL, because Eigen, which TF uses by default is not the fastest thing out
618.
by 0xbear 9y ago
Yep. But they compensate for that by charging a lot more and using lower end CPU SKUs with less cache. And GPUs are still per die.
619.
by 0xbear 9y ago
FYI, y'all: cloud "cores" are actually hyperthreads. Cloud GPUs are single dies on multi-die card. If you use GPUs 24x7, just buy a few 1080 Ti cards and forego the cloud entirely. If you must use TF in cloud with CPU, compil
620.
by 0xbear 9y ago
Taibbi has published "Griftopia", a book which I heartily recommend. Perhaps the least partisan of his works, because everyone in it gets it: republicans and democrats alike. Even Obama, believe it or not. He has a knack for expl
621.
by 0xbear 9y ago
If this is levied at me, I will concede the conspiracy theory charge. I have no evidence that this is "paid for by Soros" or anything like that. But to claim misogyny or victim blaming you need to point to some kind of verbiage wh
622.
by 0xbear 9y ago
That's pretty much what I did, modulo the physiotherapist. I would lift, get punished for it with brutal back pain, wait for a few days for it to subside, then do it again. I was free of back pain within a month.
623.
by 0xbear 9y ago
Same here. Went from being unable to lift even 100 lbs without pain to 495lbs. I don’t recommend going quite that far to everyone, but 315lbs should not be a problem at all for men.
624.
by 0xbear 9y ago
Your muscles are locked in spasm because they’re insufficient to carry the load. The long term solution to that is not acupuncture, it’s building up those muscles through exercise, and in particular through training with free weights.
625.
by 0xbear 9y ago
Dude is getting fired in 3, 2, 1...
626.
by 0xbear 9y ago
People who work 70 hours a week on a sustained basis are just bad at saying "no". Not saying "no" often is the surest way I know of to disappoint all those people you're saying "yes" to. People's atte
627.
by 0xbear 9y ago
I’ve been in this situation. You should leave and get a pay raise that way. Next time when aiming for a promo remember that results alone won’t get you there. You have to fiercely and constantly beat your chest about your achievements, and
628.
by 0xbear 9y ago
Unless your _blank_ is food or water, your startup will have a heck of a time in Ghana.
629.
by 0xbear 9y ago
Anything discretionary that is. Ghana’s per capita PPP GDP is $1381 and falling. That’s a bit over $100/mo
630.
by 0xbear 9y ago
That would be a really weak argument though. 100% of VCs are in it for the money.

More ›