While it’s presently believed that 22nm will be the maximum achievable process shrink using silicon, recent discoveries might allow chip makers to cut the 2020 goal set by Moore’s Law loose.
The discoveries come in the form of a manganese-doped germanium substrate, which will allow the creation of nanowires that can be easily magnetized. The magnetizing effect is reportedly showing “the potential of using these nanowires as building blocks for electronic devices,” such as “ferromagnetism above 300 K and a superior performance with respect to the hole mobility of around 340 cm2/Vs and other industrially relevant parameters.”
So what does all this mumbo-jumbo mean? Well, in short there’s a chemical element (number 32 on the periodic table, if you’ve got one handy) that, when mixed with a magnetic field, is showing some promise for chipmakers looking to break the 22nm barrier. With any luck, in the 10 years between today and the marked date for the 22nm barricade, the research will come full circle.
This past year we've seen a major push by several manufacturers to move solid state drives (SSDs) into the mainstream market, but the lower pricing has often come at the expense of performance. Enter Intel, who did away with any notion of bang/buck and instead focused on lightning-fast read speeds with its X-25M SSD.
Now OCZ is getting into the high performance SSD game with the introduction of its new Vertex series. Unlike the company's existing Core series SSDs,which target average users, the Vertex is aimed squarely at enthusiasts.
"The new Vertex Series of SSD drives are a premium MLC based SSD solution that are designed for consumers that require fast, rugged, and reliable solid state storage,” commented Eugene Chang, Director of Product Management for the OCZ Technology Group. “The Vertex makes use of our newest architecture and controller design complete with 64MB of cache to offer faster transfers and superior overall system response times in a broad range of applications and games."
Write speeds have traditionally been a weak spot for MLC-based SSDs, but that doesn't appear to be the case with the Vertex drives, at least on paper. OCZ claims read and write speeds of up to 200MB/s and 160MB/s respectively. By comparison, Intel claims up to 250MB/s and 70MB/s read and write speeds for its X-25M, making the Vertex appear to be a more balanced higher performance solution.
No word yet on availability, but OCZ did say the Vertex series will come in 30GB, 60GB, 120GB, and 250GB capacities. MSRPs for the 30GB-250GB will be set at $130, $250, $470, and $870 respectively.
Yesterday, Logitech announced that they’ll be releasing the G13, a gameboard keypad peripheral designed to streamline PC gaming by allowing one-handed access to dozens of programmable keys. It’s akin to niche controller products like the Belkin Nostromo Speedpad or Zboard Fang. Well, it just so happens that today we got a shiny new G13 delivered to our office, which we were more than happy to playtest. Read on to find out what we thought of the device.
VIA has announced the ARTiGo A2000 barebones storage mini-server, a tiny box with a small price tag. The compact mini-server offers a high capacity, low power power storage system while also claiming to keep noise levels below 26.8 dB.
1.6GHz VIA C7-D processor
VIA VX8000 Unified Digital Media IGP chipset
1 x DDR2 SO-DIMM Socket (up to 2GB)
2 x 3.5" SATA II
1 x CF socket
3 x USB 2.0 ports (1 on front panel)
Other specs include a LAN port, audio ports, wireless LAN support, built-in HD audio, and support for Windows XP/Vista, and Linux. But perhaps the ARTiGo A2000's biggest appeal is it's small stature. The mini-server is designed using a custom Nano-ITX form factor and up to 3TB of data can be crammed into a chassis no higher than a CD and only 10.2 inches long.
Included software gives uses the ability to create up to 10 encrypted virtual drives, with the encryption being "performed with virtually no CPU load."
Several e-tailers have begun offering the device on pre-order for $299, and depending on where you order it from, could ship as early as this month.
Don't fret if you just plunked down a wad of cash for a 45nm processor, you're still ahead of the curve. And while 32nm chips aren't 'just around the corner,' they have officially moved into the production phase and it looks as though Intel will make the transition on schedule. The first commercial processors on the shrunken die process are expected to debut by the end of 2009.
Moving to 32nm isn't without its technological challenges, and Intel will use a second-generation high-K gate material to contain leakage current, TGDaily reports. The chip maker will also transition to a 193nm immersion lithography production technology to print the circuits on the chips, something AMD already does on its 45nm parts.
Should Intel not run into any product-delaying roadblocks, the switch to 32nm will put the chip maker at least a year ahead of AMD, whose roadmap shows a 32nm server processor scheduled for 2010 (consumer processors built on a 32nm manufacturing process aren't expected until 2011). And looking past 32nm, 22nm technology has moved out of research and into development, putting it on pace for a 2013 release.
Intel will give more details on its next generation chips at the International Electron Devices meeting (IEDM) in San Francisco on December 15th.
ATI videocard owners take note - AMD has released new Catalyst drivers, v8.12, for both Windows XP and Vista. The new Catalyst release brings with it performance improvements in several DX9 and DX10 games, including up to a 25 percent boost in Crysis (DX10) for both Single and Crossfire mode. More recent releases, such as Left 4 Dead and Fallout 3, also receive a claimed performance boost by up to 10 percent and 15 percent respectively.
Many bug fixes accompany 8.12, such as improved HD video playback no longer causing Vista to stop responding, better support for connecting an All-in-Wonder card to a Yamaha receiver via an HDMI cable, and several more.
The new drivers also unlock the ATI Stream compute acceleration capabilities built into Radeon HD 3870, HD 3870 X2, and all HD 4000 series graphics cards. Similar to Nvidia's CUDA technology, ATI Stream is a set of advanced hardware and software technologies that enable AMD GPUs to work with the CPU and accelerate applications other than just graphics. To show off the technology, AMD has made available its free ATI Video Converter utility.
OCZ's making a pitch for its new Slate Series ExpressCard, a storage expansion drive the company claims is better suited than USB flash devices and external hard drives.
Compatible with USB 2.0
18 MB/sec read
12.5 MB/sec write
Voltage: 2.7V - 3.6V
The new ExpressCard storage drives aren't going to win any speed crowns, so OCZ is touting convenience and low power consumption over alternative backup solutions. Users who don't like to lug around external hard drives or who are prone to bumping into USB keys sticking out of a notebook may find appeal in an ExpressCard that stays put and out of the way.
Specific pricing and availability has not yet been announced, though OCZ did say its new Slate Series will come in 8GB, 16GB, and 32GB capacities.
If you've been thinking about upgrading to Nvidia's GeForce GTX 260 videocard, you may want to hold off for a few weeks. According to Chinese site Expreview, Nvidia will release a new 55nm-based GTX 260 along with a 55nm GTX 295 (GTX 260 GX2) in January 2009. And if history tells us anything, Nvidia tends to do well with core revisions (G92-based 8800GT, for example). Expreview posted several pics of the revised GTX 260, which it claims were sent in from Zotac.
In addition to a die shrink, the new GTX 260, or at least Zotac's version, looks to be built with a 10-layer PCB design rather than 14 layers as found on current GTX 260/280 videocards, Expreview says. The new revision also upgrades its 3+2 phase power modules to 4+2 phase.
Other specs look to remain the same, such as the number of stream processors (216) and core and memory frequencies. This means you might not see a leap in stock performance, but in theory, the power consumption, heat output, and overclocking potential should all be improved.
No word yet on projected pricing, which could either sweeten or spoil the whole deal.
In case you haven't noticed, multi-core processing has taken hold and the race is on to cram more cores onto a single die. But assuming developers can keep up, at some point, chip manufacturers are going to have address a potential major problem that could make adding more cores a useless endeavor. More specifically, a "memory wall" looms large in the not too distant future that, as Jon Stokes from ArsTechnica puts it, could make more than 16 cores pointless.
The problem stems from memory bandwidth not being able to keep pace with faster processors, whether those speed bumps come from a faster frequency or more cores. Put simply, memory is creating a bottleneck and can't feed the processor fast enough, a problem that has existed for some time. Intel and AMD have been able to mask the problem by adding more cache, but doing so doesn't overcome the memory wall, which looks poised to really rear its ugly head as more cores are piled on to new chip packages.
"Engineers at Sandia National Laboratories, in New Mexico, have simulated future high-performance computers containing the 8-core, 16‑core, and 32-core microprocessors that chip makers say are the future of the industry," writes Samuel K. Moore at IEEE Spectrum Online. "The results are distressing. Because of limited memory bandwidth and memory-management schemes that are poorly suited to supercomputers, the performance of these machines would level off or even decline with more cores."
Hit the jump to find out what solutions are being proposed.
Back in our September 2008 issue, we published a list of 9 Skills Every Nerd Needs – a lighthearted examination of the essential abilities Maximum PC readers should have in their geek arsenal. We still stand by that list, but we were somewhat one-upped last month when we saw that Gizmodo had since run its own list of 50 key geek skills. Their list was very respectable, but we thought that we could do better by not only expanding and refining our original story, but actually teaching you these skills. The highest echelon of geeks will be able to do everything in this list, and this is by no means a full categorization of the complete geek skillset – only what we consider to be the most indispensable abilities. Have anything to add to our list? Post it in the comments!