mcbain

I think when I looked single instr crc32 was ~5 times faster than the pclmulqdq version.

As an aside, if they have a server in production that doesn't support CLMUL, they should junk that machine - it appeared in Westmere and Bulldozer - everything earlier is EOLed. Crc32 was in sse4.2, so Nehalem.

The crc32 instruction is hard to beat, has been around for years, and I would guess the different poly could be phased in on their systems like you would for changing a password algorithm, or just try the new poly since it is quick, and if it fails use the old poly.

Core i3's have ECC support.

That $15 vs $150 comparison is stupid - where I can buy a A9 for $15?

The A9 is a great jump forward by Apple - and I'm impressed by my iPhone 6s - but I'm yet to see the "crushing" benchmark, except for those that are skewed by SHA results.

You can ask people if they want some toast, are they as strong as Robert Förstemann (hint: no, they are not):

https://www.youtube.com/watch?v=S4O5voOCqAQ

Content blocking requires an ARMv8 processor (so 5S and later) - must be a requirement of whatever they are using for regex.

There's a few problems with that chart, but the one that jumps out is that there is no Intel Core i5-4660. Might be a typo, but makes you question the attention to detail.

The graph would be even more complex if you trace all the part acquisitions and divestitures in and out of Avago (LSI, PLX, Infineon, Axxia, SSD tech to Seagate, Emulex).

Then again, Broadcom liked to buy things, for example ServerWorks, Renesas, Netlogic (which in turn had acquired RMI, etc).

Big fish, little fish, ad infinitum.

Xeon D for starters.

http://ark.intel.com/products/family/87041/Intel-Xeon-Proces...

http://www.theregister.co.uk/2015/03/10/facebook_open_comput...

Proper fenders and racks, or clip on ones that fall off if you look at them the wrong way?

We don't need a Waze-alike - Strava heatmaps already do it. Cyclists are pretty good at finding the most efficient paths.

So many WTF, so little time.

They designed a carbon frame and fork for a commuter bike, going through the huge time and money sink that is creating their own moulds (plus need at least one per frame size and they are expensive) and it doesn't have fender or rack mounts (for aesthetics). Nice practical choice there.

Maybe it is just prototype photos, but custom designed bars/stem/fork and they couldn't route the cables for the electronics internally? Plenty of examples of that working already (look at Di2 builds using PRO parts from Shimano). And on that: integrated stem/bars might be fine, but only if they are exactly the right size for you.

And routing/directions? (Ooh, haptics, that'll work on a bumpy street.) For a short-ride/commute bike? That you'll need once to learn the way and then not again? Maybe on a touring bike to get you through a complicated all-day route, but this is not that bike.

Yes, I'm a bike snob, but I'll stick to my cheap commuter and my fleet of pricer rides, thanks.

I'd also suggest adding --track-origins=yes for uninitialized scans - it might be a bit slower, but is very useful for nailing these things quickly.

We make sure all of our unit tests run clean under valgrind, including the unit testing framework itself - it has caused some devs to grind their teeth making the tests themselves work cleanly, but it saves everyone's sanity in the long run.

The one thing that bugs me about valgrind (well, other than support for non-Linux, non-x86 targets - although that's always improving), is that it lags behind in support for newer additions to instruction sets, like AVX2 - the core are there, but not all the semi-documented forms of the instructions. (I've got a TODO to try and work through a few of the VEX errors I get to make some of the things I work on valgrind'able again - although the latest release notes look positive.)

They were in YC S12, so that makes it magical by default.

This is where they are flying:

http://mvc-apollo68.nl/ https://goo.gl/maps/hYYQg

That's somewhat uninhabited for The Netherlands.

It also doesn't work for some of the major bridges that I ride over.

Cisco, Juniper, et al have built/are building a lot of networking gear around Cavium and Broadcom multicore MIPS parts. Intel is muscling back in to the arena, though.

The IXP was an interesting family, but I recall the toolchain being less than pleasant - although at that time the MIPS vendors weren't covering themselves in glory either.

I won't be the first to ask but: Linux client, please.

I was chatting to my mid-80's grandmother only recently when she casually mentioned that she had all her (still relatively healthy) teeth removed before her wedding, essentially as a benefit to her husband-to-be as the above post suggests.

My grandmother made it sound like it was a standard thing to do here in Australia in the 1930s/40s - whereas I'm still shaking my head.

Funny he says that about the English accent - even from the other side of the (Australian-based) hiring desk, whenever I pick up the phone and hear that accent I just know what the next question will be, and I could copy&paste my answer.

No, we aren't looking for anyone right now, and certainly not with those skills (we've never looked for those skills); yes, that's the same answer I gave you last time you called.

And if they could remove the phrase "a quick heads up" from their lexicon, it'd make my year.</gripe>

Coincidently, Keith Bostic (co-creator of BDB) announced his new project in the last few weeks: http://wiredtiger.com/

I don't know all that much about it, but given the background of those involved, should be worth keeping an eye on.

Similar - 1 GB VM, allocated a single core, MySQL backend. Been running for years (although moved from a dedicated machine to a VM a bit over a year ago), solid as a rock. Confluence is running on the same VM.

One of the bits of infrastructure I never have to worry about.

edit: never tried any of their other products however. We use Pulse by Zutubi for continuous builds.

Static analysis is a great tool - Clang's isn't too bad - but it isn't the right one to drop into a normal workflow.

We still develop using GCC - for our use case Clang performance isn't there yet - but have our continuous integration system perform Clang builds (as well as other platform/compiler variants, with unit+regression tests, etc.) so we miss out on the immediate build breakage that you mention, but do find out within ~30 mins if someone slipped up.

Both Google and Perforce have talked about the install from time to time - for example this 2007 paper: http://research.google.com/pubs/archive/34459.pdf

More interestingly (to me) is the quadratic performance they are reporting for the SSE optimized variants - ISTR when they were introduced there was some performance figures quoted. Looks like they didn't measure very well.

What is Microsoft's track record for spinning off companies?

The Photoshop/CS part of Adobe is somewhat profitable, correct? Would MS let it go (even for a bucket of cash)?

Maybe various comments indicating how they are unhappy they've missed the mobile bus, including seemingly oddball suggestions like a tool to "port" iOS apps to x86: http://www.ibtimes.com/articles/69637/20101007/intel-porting...

I did something similar to learn a bit of Haskell (to knock the rust off my C geared brain) by picking through a bunch of the problems in http://www.haskell.org/haskellwiki/H-99:_Ninety-Nine_Haskell...

I can say that it has indeed improved my Python scripting as well.

That chunk of "third_party" stuff is a pretty tempting target.

Instead of using shared libs - like those that you've probably got on your system already - Google appear to like slurping everything into the one blob to solve dependency/versioning issues.