Saturday, October 14, 2017
Friday, October 13, 2017
Wednesday, September 13, 2017
Since my iPhone 6+ works just fine and I am happy with it, I can't really justify spending hundreds of dollars on a new phone. I could be happy with either the iPhone 8 or the iPhone 10, but I would rather wait until something major goes wrong with my phone before I spend the big bucks.
When I was working, I used to get a new phone with a new contract every 2 years. The upfront cost was relatively low, usually either $200 or $300, because the carrier would subsidize the cost of the phone. Many carriers won't do this anymore,but some companies like Sprint might. But right now I am saving $30 per month on Total Wireless, so I would have to pay an extra $720 over two years to get a subsidized phone.
One reason I bought a new phone every two years is that there would be some major changes every two years. However, now the phones have gotten so good that the changes are more incremental. The new phones always have a faster processor and a better camera, but this is hardly a reason to abandon your old phone.
Apple's 6 core processor is kind of exciting. Two high power cores with 4 low power cores, and unlike the previous phone, it can use all the cores at the same time. This rivals some desktop computers in terms of power. (Some competing brands of phones will allow you to hook up a monitor, keyboard and mouse and use it like an internet based computer. Microsoft Windows Phone does this, and I think so does the Google phone.)
Wednesday, September 6, 2017
Thursday, June 8, 2017
Wednesday, June 7, 2017
Wednesday, March 29, 2017
Wednesday, February 17, 2016
'Tim Cook Opposes Court Order That Apple Must Help FBI Unlock iPhone'
Sunday, December 27, 2015
Wednesday, September 2, 2015
the mobile workstation market is one of the few areas where Windows PC sales are growing -- it only makes sense for Intel to court the increasing numbers of creatives and engineers that want to do their jobs away from the office.'
'Skylake ships with an integrated graphics core inside of it, which Intel had previously said would be optimized for low-power video decoding. What was surprising, however, was apparently how good it is – Skaugen showed off the chip decoding 15 4K videos using the chip's H.265 decoding engine, and compared it against to a machine with a pair of Nvidia Titan chips performing the same task. They struggled.'
Wednesday, August 26, 2015
The letter states: "AI technology has reached a point where the deployment of [autonomous weapons] is – practically if not legally – feasible within years, not decades, and the stakes are high:'
Musk and Hawking have warned that AI is "our biggest existential threat" and that the development of full AI could "spell the end of the human race". But others, including Wozniak have recently changed their minds on AI, with the Apple co-founder saying that robots would be good for humans, making them like the "family pet and taken care of all the time".
Tuesday, August 25, 2015
In future combat, all machines could leverage this mega-smart skin, detecting heat, damage and stress. Combat aircraft, drones, tanks and other land vehicles, as well as naval vessels could covered with the smart skin. Drones operating in air, on land, at sea or underwater could also deploy the technology.
Using a skin loaded with a range of sensors, machines could "feel" and sense things like an animal does. The smart skin would cover a combat aircraft -- reading, recording and processing the machine's sensations.'
Friday, August 21, 2015
Working with Frank Chang at Caltech UCLA, the JPL boffin Adrian Tang is keen on ways to let devices with relatively low communications needs do without recharging.
As NASA explains here, their idea is to let a chip either reflect a signal back to a base station or access point (representing a binary 1), or absorb it (representing a binary 0). That way, the Wi-Fi device (be it a smart-watch or a a bio-sensor, for example), only needs enough energy for its own operations, instead of having to carry power for a full transceiver.
Not only that, but the device is fast. The NASA release says that at a short 2.5 metre distance, it can communicate at up to 330 Mbps, "using about a thousand times less power than a regular Wi-Fi link".'
Their technology—dubbed 3D Xpoint—doesn't quite match the speed of the chips known as DRAMs. But unlike those chips—and like NAND flash memory—the new chips will retain data even after they're powered off, the companies say.
"This is a whole new paradigm," said Mark Adams, Micron's president, predicting the technology will cause "a major disruption" in the $78.5 billion memory-chip market…
compare to nram…
The service previously offered translations between English and French, German, Italian, Portuguese, Russian, and Spanish. The app works both ways: Non-English speakers can also translate English signs into their native languages. For Hindi and Thai translations, however, Google Translate can only convert English to the two languages—not the other way around—due to the complexity of their characters.
The app also works in the absence of a data connection for a phone, which makes it optimal for travelers.
The instant translation feature is largely derived from the Word Lens app, which Google acquired last year when it purchased the company behind it, Quest Visual.'
Sunday, August 16, 2015
Vulkan actually comes by way of the Khronos Group, a non-profit consortium of companies creating open standard APIs for computer graphics rendering such as OpenGL and WebGL. Once Google integrates the new Vulkan APIs, developers can choose to use them or stick with the tried and true OpenGL ES.'
Sunday, August 2, 2015
A machine at China's National University of Defense Technology in Guangzhou, called Tianhe-2 (Milky Way-2) is thought to currently be the fastest supercomputer in existence — variously reported as doing either 34 or 55 petaflops (1 petaflop is equivalent to 1 quadrillion floating-point operations per second).
The executive order, issued by Obama on Wednesday, would set up a body known as the National Strategic Computing Initiative (NSCI) to research and build what is hoped to become the first machine to hit 1 exaflop, equivalent to 1,000 petaflops…
But Industry Tap notes that such a machine would require 200 megawatts of power (compared with 3 megawatts for the current-generation machines). That means "a power plant would be required to run it."
The president's executive order is just the latest salvo in something of an international supercomputer arms race that has broken out in recent years...
A 100-petaflop machine is being developed in the U.S. and is expected to be ready by 2017...
Wednesday, July 29, 2015
For reference, a typical PC with a Haswell or Broadwell CPU and dual-channel DDR3 offers around 17GB+ of memory bandwidth, while lower-end machines survive on 9GBps or less. Higher-end systems reach into the 55GBps range, while graphics card memory far outstrips those.'
Sunday, June 21, 2015
Thursday, June 11, 2015
The new memory from Nantero, called NRAM (nonvolatile RAM), is based on carbon nanotubes. The memory is hundreds of times faster than flash storage used in mobile devices and SSDs, claimed Greg Schmergel, CEO of the company.
Carbon nanotubes are cylinders made out of carbon atoms, with a diameter of one to two nanometers. The nanotubes are known to be stronger than steel, and better conductors of electricity than other known materials used in chips, making the technology an excellent candidate for storage and memory.
Nantero's NRAM operates at the speed of DRAM and is nonvolatile, meaning it can store data. The small size of carbon nanotubes allows more data to be crammed into tighter spaces, and the storage chips will consume significantly less power than flash storage and DRAM. That could bring more storage and longer battery life to laptops and mobile devices…
Nantero, which was formed in 2001, has spent 14 years refining carbon nanotubes, which has been researched for decades by universities, the U.S. government and companies like IBM and Intel. Many top chip and device makers have shown interest in NRAM, which is now ready for manufacturing, Schmergel said…
Nantero won't make the NRAM, but license the technology to device makers and manufacturers. The first NRAM chips will appear as DRAM-compatible modules that can be plugged directly into memory slots on motherboards.
"We are designing chips that are DDR3 and DDR4 compatible, you just put in carbon-nanotube memory," Schmergel said.
Devices makers will be able to put carbon-nanotube storage on top of NAND flash circuitry so it fits in mobile devices and PCs. The technology will be compatible with storage, memory systems and protocols that exist today, Schmergel said.
The NRAM chips should arrive in the next few years, Schmergel said, adding that chip and device makers are designing the memory into new products. '
Wednesday, June 3, 2015
However, "soon" isn't "now". It's going to be a confusing and expensive journey before the promises are fulfilled.
The last piece in the jigsaw fell into place yesterday at Computex, and cemented the USB-C socket as the winner. Intel announced that the third generation of Thunderbolt will support USB-C plugs.
So only one kind of plug is needed to support power, video and audio, and high-throughput data peripherals such as disk drives.
But that doesn't mean one cable will support everything: there will be several different kinds of USB-C supporting different capabilities, ensuring confusion continues for some time to come.
The reason is obvious to the tech-savvy, but less so for the typical user who has wandered into PC World on a Saturday morning. The plugs may be the same, but the capabilities are defined by the gadgets at each end of it.
Since the expense is defined by the capabilities of the host controller, it all depends on how much the market-conscious manufacturer wanted to spend.
Most people who'll see a USB-C socket won't be getting Thunderbolt 3 performance, as the Thunderbolt hardware is a luxury-priced item that will continue to be in high-performance hardware, rather than the value mass-market.
So the industry is moving to "one plug", but retains lots of different standards. At least in the bad old days, you knew you couldn't plug your projector monitor into the modem port and expect it to work. It wouldn't fit.'
Thursday, April 30, 2015
Tuesday, April 21, 2015
Monday, April 13, 2015
Simda is a "pay-per-install" software nasty: fraudsters pay miscreants some sum of money for every 1,000 or so machines they compromise. The hackers effectively earn cash by selling access to the infected computers, renting out the botnet real-estate to other crooks.
The Simda malware, once installed and has set itself up to run after every system startup, kills off antivirus software, logs keystrokes made by the user so it can steal passwords and other sensitive information, downloads and executes banking Trojans and other malicious programs, upload copies of the user's files, and so on.
It opens a backdoor to a command-and-control server, so it can receive orders from the brains behind the malware, and send back any stolen data.
The botnet was seeded by compromising legitimate websites, and hijacking them to redirect visitors to sites hosting exploit kits – which are webpages booby-trapped with code that exploits software vulnerabilities to install the malware.
The most heavily infected countries were the US, UK, Russia, Canada and Turkey, although Simda spreads its tentacles worldwide. The vast majority of victims were located in the US, where there were more than 90,000 new infections since the start of 2015 alone.
In a series of raids last Thursday, 10 command-and-control servers were physically seized in the Netherlands, with additional servers taken down in the US, Russia, Luxembourg and Poland. The operation involved officers from the Dutch National High Tech Crime Unit (NHTCU), the FBI in the US, and the Russian Ministry of the Interior's Cybercrime Department "K" supported by the INTERPOL National Central Bureau in Moscow.
Security firms Trend Micro and Kaspersky Lab provided the cops the technical knowhow to locate the systems. The crackdown effectively decapitated the botnet by taking away the servers that sent infected PCs their instructions and received swiped passwords and other data.
Windows PCs keelhauled into the botnet remain compromised, hence the need for a cleanup operation. In order to help victims disinfect their PCs, Kaspersky Lab has created a website that will check your public IP address against a database of machines known to be infiltrated by Simda. This database was lifted from the command and control servers during the takedown raids.'