tyoma

17 ms·

151.
by tyoma 7y ago
Desktop and workstations will be the last place to find non-x86 chips. Power use doesn’t matter, and critical consumer applications and raw performance are a must. Its much more exciting to see what will happen with laptops and (in cloud)
152.
by tyoma 7y ago
Same day appointments, booked via usable app or website. Doctors who took their time with you. No waiting after you show up at the right time. I could go on.
153.
by tyoma 7y ago
Is that tuition for foreign students too? Some countries offer international students a subsidized rate (Germany comes to mind) but others do not.
154.
by tyoma 7y ago
This is my dad too. I thought he was nuts, but it was very effective. Looking at your name I am going to assume that like me you are from the former USSR. The explanation my dad said is the following: a lifetime of dealing with soviet burea
155.
by tyoma 7y ago
The big deal with AVX-512 is the scatter-gather instructions, which make it possible to efficiently vectorize different classes of problems. In this particular use case those did not really matter, but when you can use them they really help
156.
by tyoma 7y ago
Concur on pthreads and Linux syscalls, but the intrinsics should be cross-platform.
157.
by tyoma 7y ago
I know, that was one of the reasons I wanted to use OpenCL. My time for this was limited, and I saw I could get a basic implementation working much faster in CUDA. All the major cloud providers offer (only) NVIDIA cards too, so it was not a
158.
by tyoma 7y ago
I originally wanted to do just that, but my laptop is a MacBook and OpenCL is deprecated there. With CUDA I could just ssh to an old box under my desk.
159.
by tyoma 7y ago
Yes. AWS has support for ARM instances, and I have an RPI3 :).
160.
by tyoma 7y ago
There have been a few OpenCL versus CUDA posts lately and I wanted to add my take on it after writing the GPU portion of sixtyfour, which was my first GPGPU application. OpenCL needs more examples, tutorials, and documentation. I really wan
161.
by tyoma 7y ago
So the goal wasn’t to find a new and more expensive method for figuring out an answer when you can examine the oracle. The idea was to show that an “impossible” problem isn’t really that impossible given some thought about the algorithm and
162.
by tyoma 7y ago
So even though GPU time is more expensive and GPUs need specialized programming, for this particular problem they can’t be beat with raw CPU power, either in absolute or in $/ops measures. Just putting a problem on a ton of cores is an
163.
by tyoma 7y ago
There is an associated github repo with code for those who want to reproduce results ( https://github.com/trailofbits/sixtyfour ). Or if you’d like to contribute an ARM version :).
164.
by tyoma 7y ago
The accompanying source code is at: https://github.com/trailofbits/sixtyfour
165.
by tyoma 7y ago
Every time Intel has tried to more away from x86 (i960? Itanium? Maybe others...) they end up coming back. The years of backwards compatibility are a big selling point.
166.
by tyoma 7y ago
Apple Pay generates a token that can only be used by the merchant for the authorized amount. More details: https://squareup.com/us/en/townsquare/what-does-tokenization...
167.
by tyoma 7y ago
I’m certain its so you spend time wandering around and buy more stuff.
168.
by tyoma 7y ago
Had the same experience in Chicago and San Diego. The Chicago one was bad enough that I didn't come back to a Costco for about two years.
169.
by tyoma 7y ago
The Costco shopping experience is atrocious. Huge lines, people running over you with giant carts, completely unmarked aisles and you have to run around trying to find an employee to ask where something is. And then there is the checkout ga
170.
by tyoma 7y ago
I am not sure, but in my experience the driver is not present and not loadable. Regardless, the max cpu frequency is not equal to tsc frequency on newer Intel CPUs, so the number is meaningless.
171.
by tyoma 7y ago
The device is populated by the cpu_freq driver, which is used to control frequency scaling. Most cloud environments have shared tenancy and disable the feature to ensure one user can’t scale down the CPU for other tenants.
172.
by tyoma 7y ago
Hopefully something similar to this can be upstreamed into the mainline Linux kernel — ideally whatever Google is using internally since it probably works well and is battle tested.
173.
TSC Frequency for All: Better Profiling and Benchmarking(blog.trailofbits.com)
36 points by tyoma 7y ago|9 comments
174.
by tyoma 7y ago
Run for county/city elected office. Or even neighborhood zoning commission. A friend recently won a spot on his neighborhood’s planning commission in Columbus, OH. The vote counts were in the hundreds.
175.
Fuzzing in the Year 2000(blog.trailofbits.com)
1 points by tyoma 8y ago|0 comments
176.
by tyoma 8y ago
The original poster specifically mentioned plastic bottles, which, I imagine, are hard to make from paper. Refillable glass allows for exactly the kind of cross contamination that makes otherwise safe products inedible.
177.
by tyoma 8y ago
As a counterpoint, single use packaging is a godsend to people with food allergies.
178.
by tyoma 8y ago
Could someone from Sweden talk about how a year or more ofleave works in practice? I am curious about: * what percentage of parents take the full year or more of leave? * how much of a salary reduction happens on leave? * how much of the le
179.
by tyoma 8y ago
While that does seem like a buried lede, I felt it wasn't my place to raise alarms about a 2 year old bug that may have been fixed. The bug description certainly makes it sound very serious. If its not fixed in upstream, it at least se
180.
by tyoma 8y ago
This was on Ubuntu 14.04 -- I know 14.04 is dated, but its the last image I had running from the previous tests :). I wrote some stupidly simple bash to just pipe output of fuzz to every executable in /usr/bin/, with a 10 sec

More ›