Slashdot Mirror

Bullshit. VLSI code is almost always verified by finite models, and many processors are verified down to the level of mathematical axiom.

Bullshit on your bullshit, no it's not, and no they're not, not even close. Hardware companies have a fetish for formally verifying floating point stacks because, 20 years ago, the fickle and vacuous mainstream press people decided one particular piece of errata in one particular processor -- the Pentium FDIV bug from 1994 -- was important for some reason, even though every processor ever made and used has errata. AMD took advantage of Intel's bad publicity to formally verify their own FDIV instruction -- JUST the FDIV instruction, mind you -- and then doing formal verification with floating point stacks became something of a thing. There's nothing more going on than that.

Take a look here, in the section "Errata": http://download.intel.com/desi...

Doesn't look like the "proved" that VLSI very well to me, although they doubtless subjected it to a fuckton of simulation hours. Which is what they should be doing; theorem proving software or silicon is, usually, a ton of effort for little gain. Simulation hours cost much less than developer time. Our processors would likely be 486 level today if the designers had to prove everything correct. If that.

Provably correct software code exists in small amounts, and it's emergence is inevitable.

Said the formal verification researchers, for 30 years or so now.

Second this. Whether using Eclipse or Android Studio, if your development machine has an Intel with VT-x, EM64T and Execute Disable Bit functionality enabled in the BIOS, then you will see considerable speedups using the Intel x86 Atom system image with HAXM* over the ARM image.

*Intel's Hardware Accelerated Execution Manager (HAXM) dedicates RAM to speed up virtualization. Reboots are required to change the amount allocated. The ARM emulator was borderline useless and horribly frustrating to slog through but using the Intel image + HAXM is really snappy.

http://newsroom.intel.com/comm...

The customizable platform will be available to research and technology communities by January of next year.

However that still doesn't mean it will be open source as the newspapers claim.

I just measured (it's on an UPS): It's using a little over 400W at top effect (Valley benchmark). It peaks at only 55 degrees water temperature.

An alternative solution is of course steam stream. I'm surprised noone mentioned this. Then a NUC or so in the livingroom would be enough.

Saved cost, by going with a chip that wasn't heavily subsidized by Intel? I don't think so.
The HW specs are pretty much centered around the Intel® Atom Processor Z3735F, that was offered to them at (supposedly) dumping-prices.

Spec: http://ark.intel.com/products/...

Basically, a different chip would have resulted in higher costs.

I'm not trying to be a troll or flame here, so please don't take it like that. But, you may be comparing AMD processors and Intel in the same price range. Because the highest passmark bench results for laptops certainly belong to the 4th generation Intel mobile processors.

You must also work in an area of the country that is either really ahead, or really behind. I'm Sr sysadmin for a medium sized company and I haven't encountered a single person - outside of a geologist or engineer that needs real power - who prefers a desktop to a laptop in many years.

If Intel follows what they did with Ivy Bridge and Haswell, they will release a NUC based on these designs.

For ECC, many modern Pentium and i3 CPUs also support ECC. See http://ark.intel.com/products/... for example - 2 core @ 3.8 GHz for $150, perfect for a single process task. Most i5 and i7s have ECC disabled. At that point, just pay the relatively small premium for the Xeon versions.

The screen on the Lenovo has an odd 48Hz refresh rate. The MBPr does 60Hz.

The chip in the Lenovo is a year older and not as battery efficient.
http://ark.intel.com/compare/8...

Using a similar Y50...
http://www.notebookcheck.net/L...
The Sequential read/write on the Apple SSD are roughly 200MB/s faster (using LAST years SSD on the Mac as a comparison.)
http://www.notebookcheck.net/A...
Getting a SATA SSD to halve that gap costs $375 (850 pro). To actually close the gap requires PCIe and about $500.

Lenovo's abysmal 54Wh vs Apple 95Wh battery. You can't even get a replacement or extended battery from Lenovo.

And then there's build quality.

I assume this is the processor the new Mac Minis will use compared to the 2010 processor The clock speed isn't a much benchmark as it once was considering that the new CPU can ramp up to 2.7GHz. It's more of a powersave feature as the new processor has a 15W TDP as opposed to 25W. The newer chip uses a 2 x 256KB L2 cache and a 3MB L3 cache whereas the older chip only uses a 3MB L2 cache. The bus speed on the new chip is 5GT/s and the old one was 1.066 GT/s. The most important difference would be integrated graphics vs discrete graphics required on the older one.

I assume this is the processor the new Mac Minis will use compared to the 2010 processor The clock speed isn't a much benchmark as it once was considering that the new CPU can ramp up to 2.7GHz. It's more of a powersave feature as the new processor has a 15W TDP as opposed to 25W. The newer chip uses a 2 x 256KB L2 cache and a 3MB L3 cache whereas the older chip only uses a 3MB L2 cache. The bus speed on the new chip is 5GT/s and the old one was 1.066 GT/s. The most important difference would be integrated graphics vs discrete graphics required on the older one.

Not this one?
Pentium 4 at 4.0 GHz
I see some old news articles about it being canceled, but I remember the hype for it.

You're probably not terribly bothered, but Xeons are just about up to 18 cores now..
http://ark.intel.com/products/...

I'll make this simple for you: The overclocker-friendly -T and -K versions of the CPU don't support virtualization. That's why they're 1-2% faster.

Wrong, 4790K is overclocking friendly and supports virtualization.

Thanks. That slate article is good!

Game companies and developers are now reaching out directly to quasi-amateur enthusiasts as a better way to build their brands, both because the gamers are more influential than the gaming journalists, and because these enthusiasts have far better relationships with their audiences than gaming journalists do. (Admittedly, most anyone does.)

I went through the 2-1/2 months of press releases at Intel. They weren't all about chips. For example, there was one about the Michael J Fox Foundation. It's too bad that Intel, in making this decision, didn't post a press release on their site explaining why they are dropping the ads. It would have been a lot better than

"Intel has pulled its advertising from website Gamasutra,” Intel spokesperson Bill Calder said. "We take feedback from our customers very seriously especially as it relates to contextually relevant content and placements."

... which sounds so last-minute.

They could have posted something along the lines of:

Gamers are a diverse and growing segment of the population, and we are enthusiastic about making products that help them get the most out of their gaming experience. We have decided to pull our advertising from [list of sites] because they continue to promote an outdated view of gamers that stigmatizes people. At Intel, we listen to the men and women who buy our products, and we appreciate both their patronage and their concerns. Therefore, we have decided not to subsidize web sites that have become platforms for disparaging our customers. We will, of course, continue to advertise in other, more appropriate venues. If you feel that a site showing our advertising is violating this policy, please let us know. Thank you.

That would have sent a powerful message.

The laptops are based on the Celeron N2840, with 2GB of RAM. I can't seem to find much in the way of benchmarks; but I suspect that they are surprisingly adequate. What is a bit surprising is that the the N2840 has a quoted tray price of $107, so either Intel is cutting HP one hell of a deal, or I don't even want to know what HP cobbled the rest of the system together from...

I don't think that tray price has much basis in reality. The "$107" N2840 looks, at least on the face, to be not vastly different from the "$86" 1037U. If Biostar can sell a motherboard + 1037U + heatsink + fan for $79.99, it doesn't take much of a stretch to think maybe these prices are just "list" prices with no basis in reality. Biostar is just selling a bare motherboard so there can't be any Microsoft kickbacks or ad revenue.

The laptops are based on the Celeron N2840, with 2GB of RAM. I can't seem to find much in the way of benchmarks; but I suspect that they are surprisingly adequate. What is a bit surprising is that the the N2840 has a quoted tray price of $107, so either Intel is cutting HP one hell of a deal, or I don't even want to know what HP cobbled the rest of the system together from...

Sure... You can tell it's the future cause the vaporware now comes as vaporwear.

From TFA:

Absolutely no information about availabilty seems to be listed anywhere, but if you head on over to the official website (linked to below), you can add your email to the company's mailing list to keep up-to-date.

And videos are just your run of the mill advertisement for imaginary products.
Showing diddly-squat of actual operation or even wearing of the product, while showing instead obviously fake videos of them throwing the "prototype" off screen (which does not even clip on to the hand at this point) and "drone footage" which is too well focused and stabilized to be from a wrist-mountable drone camera, obviously NOT wrist-mountable drones flying around, 3D renderings, and not even a single 360-degree shot to prove it was done with at least a camera hanging off of a drone (or a movable crane).

Oh... It's a part of a contest sponsored by Intel?
With prizes of $50,000 to finalists (10) and a $500,000 grand prize (1)?

Well why didn't you say so? I've got a design for a floating cloud sofa I could have entered.
It's like this only with an "intel inside" logo taped to it.

Sure... You can tell it's the future cause the vaporware now comes as vaporwear.

From TFA:

Absolutely no information about availabilty seems to be listed anywhere, but if you head on over to the official website (linked to below), you can add your email to the company's mailing list to keep up-to-date.

And videos are just your run of the mill advertisement for imaginary products.
Showing diddly-squat of actual operation or even wearing of the product, while showing instead obviously fake videos of them throwing the "prototype" off screen (which does not even clip on to the hand at this point) and "drone footage" which is too well focused and stabilized to be from a wrist-mountable drone camera, obviously NOT wrist-mountable drones flying around, 3D renderings, and not even a single 360-degree shot to prove it was done with at least a camera hanging off of a drone (or a movable crane).

Oh... It's a part of a contest sponsored by Intel?
With prizes of $50,000 to finalists (10) and a $500,000 grand prize (1)?

Well why didn't you say so? I've got a design for a floating cloud sofa I could have entered.
It's like this only with an "intel inside" logo taped to it.

All i care about is performance. If i want performance, i will learn how to use C++,

Actually, if all you care about is performance, generally Fortran is the language for you. Its when you start caring about something like toolchain support and your own established codebase that C++ starts looking like a much better choice to most folks. But if you "don't care" about those things, here's your compiler.

I stuck in the following mobo - http://www.intel.com/content/w... - into an old casing, put in an old psu, and an ethernet card on its slot, with an SSD card, and that's all to it

And it has been running for the past 3 years, 24/7 without giving me any problem

Intel's best kept secret is that many of Intel's cheapest processors support ECC (including most of the i3 series), and as such enable you to build some surprisingly low-cost low-power file servers.

Here's the list of Intel desktop CPUs that support ECC:

http://ark.intel.com/search/ad...

Looks like the MSRP starts at $64 or so. The downside is that you need a chipset that supports ECC too, and those are only server chipsets. Luckily, a motherboard with one of those (like the Intel C222 chipset) start at ~$140 or so.

Slapping together a low-end server motherboard with an i3, some 8-drive HBAs, and a bunch of ECC RAM, it's a popular way to make a low-end file server.

It isn't 100% clear: All 'Silvermont' SKUs appear to be 64 bit capable, and this board has a Silvermont-based Atom on it(two of them actually); but Intel's Edison Native Applications Guide definitely appears to be walking you through setting up a build environment for 32 bit x86.

Whether that means that Intel actually lasered support for 64 bit execution off when they were designing this chip(which isn't like most of the other Silvermont devices, which have a GPU and more PC-style I/O, so clearly some cutting was done), or whether it means that they decided to default to 32 bits for everything because the device only has 1GB of RAM soldered on and no support for more, is not clear from anything I've been able to find.

TFS is simply incorrect; but may have been confused by Intel's "Galileo" board, which is based on Quark (at 400MHz). Curiously, 'Galileo' is much more Arduino-esque (a bit more raw punch, weaker hard-real-time bitbang); but also has a full PCIe lane(brought out on a miniPCIe connector) and 100Mb ethernet(optional PoE in Gen2); but no RF.

The much more PC-like, or at least BB Black/rPi-like Edison has a substantially punchier CPU; but no PCIe, wired ethernet, and includes wifi and BT.

I'm not sure if we are misjudging intel through the lens of existing products, or if Intel can't quite decide what niche to hit; but I find this mix a trifle confusing personally...

The teeny little one has the high powered CPU (relatively speaking); but not the high speed expansion bus or wired networking and PoE options. The relatively big one has high speed expansion (but severely limited GPIO, making compatibility with MCU projects that depend on bitbanging rather tepid); but the weak CPU and limited RAM.

I'm interested, and it's always worth keeping an eye on Intel; but I'm a bit confused about what they are aiming at here.

Intel has the Product Brief up. Unless they specifically decided to hide the fact (which would seem unlikely), video appears to be absent.

Wireless connectivity looks pretty nice (wifi and bluetooth out of the box, though BTLE is mysteriously 'in Q4-14', which makes one wonder if perhaps the driver situation has a few grim little stories that we should know about...) and the inclusion of both 2x Atom cores at 500MHz and 1x Quark core at 100MHz is potentially interesting, depending on how easily and elegantly the Atom and Quark 'sides' of the device can talk to one another and either share control of, or at least transfer control of, the various I/O lines.

Not going to shove the rPi out of the way for video-pushing applications, and I suspect that PWM and bitbang-heavy applications may still be underimpressed by Intel compared to the classic microcontroller options; but it could certainly be a contender for a lot of the 'arduino-connected-to-the-network' applications which don't lean too hard on squeezing every last drop out of bare-metal-MCU work; but which could really use a bit more punch on the networking and storage/logging side.

> These shiny new processor having working TSX instruction sets? The ones that are supposed to help with virtualization?

TSX is not for virtualization, but for transactional synchronization, it provides efficient transaction locking for multi-threaded applications. Not necessarily virtualization, although it can benefit from efficient locking as well

No, as far as I know, these have TSX disabled, or will be with a microcode update, as TSX isn't expected to be fixed until 2015 in Broadwell or Haswell-EX Xeons (not Haswell-EP which these are).

Why just tablets? This sounds like a full powered, full featured processor for smart phones and a serious attack against AMD's mobile market share.

Their marketing promises are largely useless (productivity up to 19% better? 3-D graphics up to 47% better? What does that even mean?) but with graphics, wireless, and fast processing in a low power chip they're already there.

When I saw a mention of a "small L3 cache" I looked at Intel's site (warning: PDF) which also didn't five actual L3 cache sizes. By the way you'll find more technical information about the chip there than in the article. L2 is 256k per core. Cache aside in general it's a solid midrange processor at mobile level power usage. Here's hoping to see them in a phone soon.

because the school systems are grinding the future brilliant, passionate creative scientists into drones.

I have two kids in public school, and I have seen no evidence at all that the schools discourage creativity. In elementary school, my kids did an independent science fair project every year. They learned to do graphical programming in Scratch. The school had several teams that competed in robotic competitions. In high school, they have the full range of science classes, and students are encouraged to do original research or development as an independent study project with a mentor recruited from a research center or tech corporation. Last year, several students from my daughter's school competed in the Intel Science Talent Search. The public schools seem to be doing a much better job than they did when I was a kid.

They run the same benchmarks. It's a lot of copypasta from Intel's marketing material. Boring. How many enthusiasts are helped by a photo of the chip with the cores labelled?

From the reviews I could not figure out whether vPro or the virtualisation bits were turned on.

vPro: no
VT-x: yes
VT-d: yes

Source: http://ark.intel.com/products/...

You asked: Can anyone tell us a simple way to check? [if my laptop's CPU supports TSX-NI]

Here is a list (as of November 2013), scroll down for an Intel reply:
Where are the Haswell laptops with TSX-NI ?
https://communities.intel.com/message/211616

The list starts with i5-4200H, i5-4350U, i5-4300U, i5-4300M, ... and continues up to the i7 chips

I know this was a troll, but I feel compelled to reply in case someone doesn't know.

ALL CPUs have errata. Some of it more significant than others.

A quick Google for "AMD errata" revealed Revision Guide for AMD Family 16h Models 00h-0Fh, published June 2013, and applying to AMD's Mobile A,E, and G series, and Opteron X1100/X2100 (These are modern CPUs)

There are 21 entries, with descriptions, system impact, and suggested workaround (if any)

Haswell's errata has 131 entries

Write a letter with proof of purchase to Intel.
http://www.intel.com/content/w...

Can anyone tell us a simple way to check?
Intel has on their website info on the processors.
For example, for yours (i7-4700mq) you would look at:

http://ark.intel.com/products/75117/Intel-Core-i7-4700MQ-Processor-6M-Cache-up-to-3_40-GHz

Or you can look for all products that were "formerly haswell":
http://ark.intel.com/products/codename/42174/Haswell#@All

how to apply the "disable the broken feature" fix - without installing windows

I would do some searches for updating BIOS from linux - ex:
https://wiki.archlinux.org/index.php/Flashing_BIOS_from_Linux

Or doing a microcode update:
https://wiki.archlinux.org/index.php/Microcode

Until there is a chip for sale that really supports TSX I wouldn't expect anyone to be distributing software that uses it. So I wouldn't be too worried about it yet.

Can anyone tell us a simple way to check?
Intel has on their website info on the processors.
For example, for yours (i7-4700mq) you would look at:

http://ark.intel.com/products/75117/Intel-Core-i7-4700MQ-Processor-6M-Cache-up-to-3_40-GHz

Or you can look for all products that were "formerly haswell":
http://ark.intel.com/products/codename/42174/Haswell#@All

how to apply the "disable the broken feature" fix - without installing windows

I would do some searches for updating BIOS from linux - ex:
https://wiki.archlinux.org/index.php/Flashing_BIOS_from_Linux

Or doing a microcode update:
https://wiki.archlinux.org/index.php/Microcode

Until there is a chip for sale that really supports TSX I wouldn't expect anyone to be distributing software that uses it. So I wouldn't be too worried about it yet.

Check the Intel ARK page for your model number Ex: http://ark.intel.com/products/...

Fast RAM is mainly important for graphics. AMD has a more powerful IGP, the Intel equivalent performs worse and so requires less. That is why Intel went with embedded DRAM on their best IGPs (brand name "Intel Iris Pro Graphics 5200"), though none of these are retail chips but only for laptops and AIOs. Personally I'm of the opinion that either you don't care about the GPU at all and it doesn't matter, or you should care enough to get a decent graphics card. Putting a CPU+GPU on a 65W power budget won't ever be great unless you want to play Dota 2.

In-memory tables allow the indexes of database tables to reside in memory to speed up transactional updates. These in-memory indexes are typically hashed for unique versioning so queries can spread throughout all of the processors in a computer, which presents the problem of the table de-syncing as each processor/core makes a change.

So, this Xeon model has special instruction set that helps keep the in-memory index synced across all cores in the server. Here is an Intel brief describing the technology and it's use:

https://software.intel.com/sit...

I bet Google DOES use some moderate amount of assembly. I once worked for an audio-recognition company and we did indeed use about 100 lines of x64 assembly to perform the inner loop, which was some complex audio signal processing routine. Similar to an FFT.

This was easily 10x faster than the C version, which we had for reference purposes, even when using the Intel compiker with all optimizations turned on.

So, just because you never saw a Tapir in your life, does not mean they can't exist because their dick is longer than you can imagine.

Maybe you shouldn't have been using an AMD processor:
(Intel has been slammed for their compiler creating code that directs non-Intel CPUs to completely unoptimized code, not taking advantage of SSE, etc, even when present in the non-Intel processor)

http://www.agner.org/optimize/...

Section 2.3 of this:
http://download.intel.com/pres...

It's a shame the software didn't have a handy dandy instruction that it could execute without reference to the OS or libraries or permissions.

Intel has 107,600.

How many people do you think work in the graphics division of AMD? How many at NVidia?

http://ark.intel.com/m/product... The Intel Silvermont Atom boards are very electrically efficient and offer surprisingly good performance. You can buy a board for under US$100 and all you need to add is case, PSU, RAM and mass storage. Some boards have VGA, some DVI, with or without legacy serial and parallel, lots of choices. Manufacturers include gigabyte, msi, Asus, supermicro.

Parent poster here, I use the 840 Pros I mentioned above on my laptop, I already have extensive caching going on with about 12GB of my 32GB of RAM, but its still saturates the SATA bus due to the mostly random nature of the I/Os. Its basically a giant 300GB b+tree with 2MB leaf nodes and about a 40% insertion 40% lookup and 20% deletion ratio.

Wouldn't something like this or another enterprise drive be a better match for you?

That's a fun post! 36-core is immense! As an aside: It's been a while since we've seen any decent rise in processor Ghz. I remember IBM talking about functioning reasonably cool 10 Ghz processors (ref needed) in the early 2000s, but no one has them in the shops yet! I'm sure this was discussed in Moore's Law lectures prior to Y2K, but mention it these days and everyone scowls! So some people can (and they run cool) and some people can't, what normally happens in computing when the faster items are released?

It's a step down from the 48 core CPU Intel created in 2009. http://www.intel.com/pressroom...

Cache coherency has been one of the banes of multicore architecture for years. It's nice to see a different approach but chip manufacturers are already getting high performance results without introducing additional complexity. The Oracle (Sun) Sparc T5 architecture has 16 cores with 128 threads running at 3.6Ghz. It gives a few more years to Solaris at least but it's still a hell of a processor. For you Intel fans the E7-2790 v2 sports 15 cores with 30 threads with a 37.5MB cache so they're doing something right because it screams and is capable of 85GB/s memory throughput.

I'm sure the chip architects are looking at this research but somehow I think they're already ahead of the curve because these kinds of cores/threads are jumps ahead of where we were just a few years ago. Anybody remember the first Pentium Dual Core and The UltraSparc T1?

My guess would be that the real perk is bandwidth and latency. Unless Intel really phones it in on integration, the FPGA should have about the fastest, lowest-latency, link to the CPU, possibly even some of the cache, especially if they throw in a big chunk of eDRAM, as they have for 'Iris Pro' parts, that money can buy.

As usual, the slashdot post has the absolute worst story link. compare http://www.enterprisetech.com/... which gives you links to where it gets its info, namely https://communities.intel.com/... and http://gigaom.com/2014/06/18/i... ... the latter is the interesting link because it tells us that the FPGA will have access to main memory. I personally would presume that means it's tied into the memory controller somehow.

Less of a "Hey, let's do this instead of GPU compute!" and more of a "It sucks that our weirdo application-specific operation is probably never going to be one of Intel or AMD's extensions to x86; but this is the closest we can get to having it added" thing.

What I began fantasizing about immediately upon reading the article was some sort of optimizer that would semi-automatically build functional units to perform whatever function the CPU was grinding on at the moment, with some sort of recognition engine and periodic updates garnered from participating customers to help special-yet-common cases. As well, seeing how customers actually use FPGA with their products will help Intel decide what functionality to add to their next (or next+1, etc) processor.

There are already options to add an FPGA to your Xeon system, with its own blob of RAM. Since they talk about this being fundamentally different, I'm not sure what makes sense except the idea of it being connected at the memory controller. Hopefully there will be a talk with some nice block diagrams released soon.

Intel's x86 emulator.

I decided to check to see if it would support my programs. It didn't take long to hit a roadblock.

Requirements for Office 2013 - http://office.microsoft.com/en...

Hardware acceleration Graphics hardware acceleration with DirectX10 graphics card

According to http://www.intel.com/products/... , there's no Directx10 support from this board.

Google is your friend, but I'll save you a little time. Some info from Google and Intel.

The "ultra-low power" 2 core Haswell has a 35 w power budget.

There are many Haswell processors below 25W TDP, in fact that list is made up of most of the actual "ultra-low power" ones (U and Y branded). ARK will list every Haswell processor for you. Do they make lots of processors that draw more than 25W? Sure, but the trend has been flat or downward since the Core 2 release while providing more processing power (so regularly improving performance per watt). If they were binning to throw away anything over 25W they'd just end up with a lot of waste throwing away "bad" parts that work just fine in an environment that isn't that power constrained, like my local desktop. I know my processor is 84W because I wanted good performance when needed and for a desktop that level of power draw just isn't that relevant. When it's not working, it idles about the same as one of the better power bins.

Even according to their own developer website, only three phones on the entire market use x86.

This is a problem for tablets, then. But wait! This could all be solved by Google simply filtering their store offerings by what's available to x86 tablets vs ARM phones, which would be a real kick in the nuts for Intel.