"it is possible to run the Athlon X2 4800+ at 3.14 ghz , which is a 30% overclock, albeit with a very serious cooling solution."
Note the "very serious cooling solution". In the case of the OCed Core Duo, they used the stock cooler and it didn't even get hot when clocked WAY beyond the speed they tested it at. If you read the article they were able to run it as high as 3.1 GHz or so and the stock heatsink was still cold to the touch at that speed.
I've been a big AMD fan for a long time, but now that I own a Core Duo laptop (Intel has managed to maintain superiority in the mobile market) I am definately considering going to Conroe for my next desktop upgrade depending on price and what AMD comes out with. I already have an X2 3800+ based file server, and in terms of raw CPU, the Core Duo beats it. I'll admit that other than transcoding of video I can't do many comparisons between the two machines though.
That would've been a good comparison to make - Core Duo T2500 (approx $350-360) vs. Athlon 64 X2 3800+ (Approx $300). Like the FX-60 vs. OCed Core Duo, they run at the same clock speed. Note that the Core Duo has a TDP of 31W and the X2 3800+ has a TDP of 89W. Price/performance is close if not equal in my experience, and performance per watt of the Core Duo is WAY ahead.
One key here - Intel has a roadmap that includes a clear performance boost beyond that of the Core Duo within months, while AMD only has incremental upgrades (Socket AM2) planned. I've also seen references to a new A64 core with an extra FPU, but nothing about availability of that.
Just because AMD has been the king of the desktop for a while doesn't mean you should count Intel out now that they've finally ditched the Netburst architecture.
"Especially since all new Thinkpads have a fucking TCPA chip. Can you trust a chinese fabbed uber security module for critical national security purposes?"
Wherever that TCPA chip was fabbed, it is almost guaranteed to come from the same source as those found in laptops from any other manufacturer.
If there were ANY chips in the Lenovo that were built in China without a clear paper trail leading back to a non-Chinese supplier (or a Chinese supplier that isn't also selling chips to manufacturers that the government ISN'T freaking out about) then this would be justified. But it's an Intel CPU with an Intel northbridge/southbridge, Intel/ATI/NVidia graphics, the LCD is probably Japanese or South Korean just like nearly every other LCD on the market. The only place where Lenovo could backdoor the machines is in the BIOS (easy to audit) and in the OS (hire a sysadmin to nuke and repave).
I second the issues with GPUs and power management. NVidia's drivers disable clock adjustments on mobile chipsets under Linux currently, and also do not officially support PowerMizer. That said, battery life on my old Dell I8200 and on my new E1705 were both quite good despite these issues. Suspend to disk and suspend to RAM are both very iffy with NVidia or ATI chipsets. I haven't gotten them to work yet on my machine, others say suspend2 works great on their NV-based systems.
Speaking of ATI, avoid them at all costs under Linux. Their drivers are even worse under Linux than under Windows.
The Intel GMA950 is extremely difficult to get running under Linux right now, although historically Intel chipsets have been pretty well documented so this is likely to change in the future.
As to WiFi - most Core Duo-based systems have an Intel PRO/Wireless chipset. In the case of the Dell E1705, it's the IPW3945 a/b/g chipset, and if you go to the sourceforge project for the 3945 and get the driver it works great.
There are kernel patches that allow the SD/MMC readers on the E1705 (and other laptops) to function under Linux, although the last version of the patches I tried didn't like my new 2gig Corsair 133x card. I've never had any problems with any other card though. I haven't tried checking for a new patch recently.
Overall, with the exception of a few minor issues (covered above), all of the hardware on the Dell E1705 with an NVidia chipset works amazingly well under Linux. Reports on the Gentoo forums indicate that ATI-based E1705s are extremely problematic and Intel GMA950s are a royal PITA too at the moment.
Similarly, my Netgear WPN511 is also rock solid with madwifi on my old laptop.
Newer Intel-based laptops almost always come with an Intel PRO/Wireless chipset, all of which are rather well supported. The 3945 isn't in the kernel yet and can be a bit of a pain to install due to dependency on a newer version of the 802.11 stack than what is in the kernel, but once installed it works GREAT in my Dell Inspiron E1705.
Not initially, but every drug eventually has to undergo clinical trials in humans, usually multiple phases with increasing numbers of patients each phase. Clinical trials are a critical final step in FDA approval for any drug. (Of course, animal trials are the usual predecessor to this before the FDA will even allow a clinical trial in humans.)
Some drugs do get shortcutted through the trials phases faster than others. These are usually the class of drugs where the drug may possibly kill you or injure you, but what the drug treats will definately kill you. This is one such case - Even if the new drug has some nasty side effects, they're better than the alternative (MRSA or VRSA), and it sounds like even initial animal trials have indicated fewer side effects than vancomycin, which is a nasty drug that was approved anyway because the alternative to using it is nastier.
Depends on exactly how you define "public airwaves". While "public airwaves" may not be the proper term for what the parent poster meant, by his definition XM isn't public. To receive any XM content other than their preview channel, you must pay them.
There is plenty of precedent for XM being considered different from public OTA broadcasts because of the fact that it is a subscription service. Note that FCC censorship rules do not apply to XM or Sirius because they're subscription services, just as they don't apply to cable TV. Yes, XM and Sirius do some self-sensoring (All XM channels with "questionable" language/content are marked with XL in their channel names, non-XL channels have censored content.)
If you want to charge extra, provide additional service beyond what you already advertise.
Hint: Multicast will solve many of the problems you complain about. There's been a standard for it for HOW long without you implementing it?
(I'm not sure, but I have a feeling a multicast-enabled BitTorrent would use less than a tenth of the bandwidth it currently does for most swarms, if not less.)
There are multiple types of radiation. There is ionizing radiation (usually from nuclear sources) which can definately cause cancer, and IS cumulative in its effects. It has the effect of essentially flipping bits in your DNA. Flip enough bits and bad things happen.
There is also non-ionizing radiation. All electromagnetic radiation with wavelengths longer than UV falls into this category. This includes visible light and microwaves. There has never been any proof that such radiation can cause cumulative damage unless it exceeds a certain power threshold. In the case of microwaves, tissue damage only occurs when the power level becomes high enough to begin causing thermal heating.
Faulty and malfunctioning equipment could, in theory have significantly higher output power. That said, these faults would be obvious and would result in the tower owner's maintenance people being notified immediately and most likely would involve shutdown of the equipment within minutes.
Power amplifiers for cellular systems need to handle a certain average power, but the signal they transmit has peaks 7-9 dB higher in power than average (somewhere around 6-8 times the average power). Thus the transmitter must be sized to handle the peaks at least in the short term. Thus, an amp capable of 45 watts average can transmit 400 watts, BUT most amplifiers are not designed to handle the accompanying thermal load and power drain and would quickly overheat and shut down if run continuously at peak power for very long.
Even at 400W output power, you'd be completely safe more than 3-4 meters away in free space, and even less than that with a roof in between you and the transmitter.
At my last job, the company I worked for built transmitters for cell towers. We were routinely running these things without their covers on, and microwave emissions leaked all over the building. There was a story that once the leakage alone from a dummy load they were running a demo unit into was jamming Sprint service for the entire conference hall the demo booth was in. Many of my coworkers were working with tower equipment since the very first Motorola demo. There was not any evidence whatsoever of an increased incidence of cancer. If everyone's fears were true, one of my coworkers would've been a hideously deformed mutant instead of an active, healthy (but completely grey-haired) old guy.
The article said the tumors were "consistent with radiation" as the cause. People will continue to be stupid and not realize the difference between electromagnetic radiation (better turn off your light bulbs... There's RADIATION coming from them!) and nuclear radiation. Yes, I know gamma rays are a form of electromagnetic radiation, but they have extremely high energy and can cause ionization, while microwaves cannot cause ionization until their field strength exceeds the breakdown threshold of the dielectric they're in - by the time that happened to someone they'd be thoroughly cooked by thermal heating.
It's an old building, there's probably traces of asbestos or some other nasty construction material that's been banned for decades.
I've used wireless hardware from the following manufacturers:
D-Link - DWL-650, DWL-G650, DI-624 The original 650 wasn't a stellar performer but it wasn't horrible. The G650+624 combo was pretty decent. I only returned it in favor of waiting to see where things went as far as MIMO gear.
Belkin - Can't remember, it was a b-only router Utter crap. Couldn't last more than 2-3 days without crashing. Died permanently in just over a year.
Microsoft - MN500 In true Microsoft tradition, their software may be crap but their hardware products are actually decent. The MN500 was the most solid and consistent performing 11b kit I have used so far.
Netgear - WPN824 router + WPN511 NIC Stellar. Utterly stellar. I love the 824. WPN511 is only retired due to the fact that my new laptop has a built-in Intel PRO/Wireless 8945 a/b/g NIC.
"Ok, so it might have marginally better spectral efficiency per Mbps but really, what we want to see is true beamforming dynamic-arrays that will properly 'point' the RF where it's supposed to go in real-time."
Actually, to Airgo's credit, this is exactly what they do. Effectively, they form seperate beams for each possible path between units, and send different data on each. When connected to a "vanilla" B/G unit, they fall back to steering all of the power along the path with the least loss.
OK, yeah, they also implement "turbo" mode (channel bonding of multiple channels) unfortunately, but that isn't where the huge improvements over 108Mbps gear come from.
Although if you want realtime beamforming that's designed to work well with "vanilla" 11g gear, I suggest the Netgear RangeMax (Not RangeMax 240, that's Airgo-based) line (WPN824 router). They combine a (mostly) standard Atheros chipset (It does support Super G including "Turbo" mode, but still works wonders even with all that stuff disabled.) with an adaptive phased array antenna from Ruckus Wireless. The client cards are actually just vanilla Atheros XR (eXtended Range, which simply means 100mW transmit instead of 25 and higher receive sensitivity, both Good Things which are fully compatible with all other 11g gear) chips, NOT the Atheros MIMO chips (which is what D-Link's MIMO gear uses). I'm currently using the WPN824 with a "vanilla" Intel PRO/Wireless chipset most of the time and it still works far better than any other AP I've used.
Um, all of Dell's current laptop products use Intel WLAN chipsets.
Personally in that regard I'd prefer the Dell. The Intel PRO/Wireless silicon is designed for rock-solid performance within the 802.11g standard, not a bunch of flashy compatibility-breaking extensions that will rarely work as advertised. Their chipsets also (unlike Airgo's) work extremely well under Linux.
Linksys does not make their own silicon. Neither, I believe, does Cisco. (As evidenced by one of the major reasons for Lucent spinning off their microelectronics business as Agere - Despite being a Cisco competitor, Lucent was selling a lot of silicon to Cisco and Agere stood to sell even more as a non-competitor of Cisco.)
Airgo (along with Broadcom and many other companies) are SUPPLIERS of Linksys. Note that this does not make Linksys "rebranded Airgo", as Linksys just buys the ICs and builds a box around them.
By your definition, Intel is "rebranded Dell".
Commenting on the announcement itself - Stock 802.11g is more than sufficient for online gaming. The bottleneck will be one's connection to the outside world for a LONG time. To the geeks of Slashdot, this means that if you want to run Linux on your laptop and have everything work, stay far away from this laptop. Now that reverse engineered Broadcom WLAN drivers exist for Linux, Airgo is basically the only chipset vendor without any Linux support. (At least that was the case as of December when I was shopping for new WLAN hardware.)
I find it interesting that Linksys was perfectly happy to profit from Linux by using it in the WRT54G, but probably sells the highest percentage of non-Linux-compatible WLAN hardware. (Almost all recent Linksys products, especially any that contain extensions beyond vanilla 802.11g such as the SRX and SpeedBooster lines, use Broadcom or Airgo silicon.)
The exec who uttered that quote seems to be rather out of touch with current events.
I'm out of touch and even *I* know that numerous politicians have been calling for an investigation of the oil companies for price gouging/price fixing within the past few weeks.
"It's not clear to me that they'll be any more successful in getting across that this is "new and improved" than they were with the Pentium 2,3,4 series."
Probably because there wasn't much as far as "new and improved" going from 2->3 (same basic architecture, I believe only minor changes such as adding SSE and cache changes, and many of the latter cache changes changed frequently within a revision.)
There was also very little as far as "new and improved" going from 3 to 4. In fact, the initial P4 requirement for RDRAM plus the P4's bad IPC meant that at the time the P4 was initially released, the P3 was a better choice for processor. It wasn't until the P4 scaled up more in frequency that it became the obvious choice, and the return to a P3 descendant makes one wonder if the P4 would EVER have been the clear choice if the P3 architecture had been continually maintained and also continued ramping up in clock speed.
I don't have time to read the detail, but your post and the original article's comment about "recycling power" sounds to me like they are using some sort of adiabatic logic approach. Adiabatic logic is well known for significant power reduction, but at least historically it has required significantly more transistors per gate and cannot run as fast as traditional CMOS.
The ring thing sounds like it's just a new clock generation scheme to go with the existing adiabatic logic techniques (which do have rather unusual clocking requirements that are a bit harder to generate than typical single-square-wave clocks.)
Given the parent's comment about "cheaping down the ingredients", it's less likely that they used higher-end ingredients (Ace-K to replace some of the aspartame - BTW Aspartame=NutraSweet), but used a lower-end ingredient instead.
Did you know that the Diet Coke and (I believe) Pepsi served in restaurant soda fountains is usually not the same as that sold in stores? In many cases it's actually sweetened with saccharin, and that stuff DOES taste nasty, even to someone who has been drinking diet for over a decade. The Coke Light in the Phillipines may likely have been saccharin-sweetened.
Also in the "traditional flavors" arena, try Splenda-based drinks. They're still not the same as regular, although they do taste different. Disclaimer: As an insulin-dependent diabetic who has been drinking nothing but diet sodas for over a decade except in the case of hypoglycemia emergencies, regular sodas taste extremely strange to me. Note that below, assume that I am ALWAYS talking about the diet version.
My personal favorites: Lime Coke/Pepsi. I can't tell much difference here between the two. Vanilla Pepsi. Vanilla Coke seems to be not quite as good. Lemon Coke/Pepsi are OK. Well, one of them is. The other has a distinct medicine-like taste, I can't remember which. Coca-Cola Blak - This isn't zero-carb, although at 45 calories per bottle it's acceptable even to a diabetic. Of course, that is due partly due to the fact that bottles of Blak are only 8 ounces. You can emulate the taste of Blak somewhat by adding coffee flavoring to Coke. Same technique also works well with Vanilla Coke. This of course depends on the nature of the flavoring. I've been using a bottle of Kahlua-flavored sugar free syrup from DaVinci Gourmet, but I'll be looking into a less expensive solution than the DaVinci syrup soon. Diet MD Code Red - Oh my this is delicious. As the parent said, it's damn near impossible to find. I wish they'd release MD LiveWire or Pitch Black in diet form too. Slice ONE - New to the market. Traditional Slice is just orange, but they have grape (the first major brand to offer diet grape soda in over a decade, until now Diet Polar Grape has been the only option, and that's hard to find and expensive) and carbonated fruit punch flavors. REALLY good stuff, not caffeinated though. Vault Zero is also good. Despite Coke's marketing, it is NOT an energy drink, it's just highly caffeinated soda similar to MD. IMO Vault is a little smoother tasting, MD has more bite. Vault Zero is difficult to find in any packaging other than single bottles.
In terms of energy drinks, I really like Monster Lo-Carb and MDX. Red Bull and Rockstar are NASTY. Full Throttle sugar free (both FT and FT Fury) are OK.
Intel has recently come up with a series of totally unoriginal and ultimately confusing names for their CPUs.
For example, the "Core Duo" is a pretty unoriginal name for a dual core processor, and I've seen a lot of people start referring to dual core CPUs as "DuoCore" or other such nonsense.
Core 2 Duo? Talk about redundant and confusing naming...
Perhaps there is a band with an album called Weapons of Mass Destruction?
As to body parts - keep in mind that those body parts don't have to be human or animal. Vehicles have body parts too, which are almost surely sold on eBay. Same for dolls. (do a search on eBay itself for body parts...)
Same for warheads - lots of REAL eBay results, none of which are actually explosive, including "warheads" candy.
Given that few fossil fuel powered electric plants in the United States actually use petroleum-based products.
Instead they use coal, which is far more readily available in the United States than petroleum is. In fact I believe we are entirely self sufficient as far as coal goes and may even be exporting significant amounts of it.
Slashdot Mirror
I suggest you click on the "Public Companies" link to the left on the page you linked to.
Your opinion might change.
"it is possible to run the Athlon X2 4800+ at 3.14 ghz , which is a 30% overclock, albeit with a very serious cooling solution."
Note the "very serious cooling solution". In the case of the OCed Core Duo, they used the stock cooler and it didn't even get hot when clocked WAY beyond the speed they tested it at. If you read the article they were able to run it as high as 3.1 GHz or so and the stock heatsink was still cold to the touch at that speed.
I've been a big AMD fan for a long time, but now that I own a Core Duo laptop (Intel has managed to maintain superiority in the mobile market) I am definately considering going to Conroe for my next desktop upgrade depending on price and what AMD comes out with. I already have an X2 3800+ based file server, and in terms of raw CPU, the Core Duo beats it. I'll admit that other than transcoding of video I can't do many comparisons between the two machines though.
That would've been a good comparison to make - Core Duo T2500 (approx $350-360) vs. Athlon 64 X2 3800+ (Approx $300). Like the FX-60 vs. OCed Core Duo, they run at the same clock speed. Note that the Core Duo has a TDP of 31W and the X2 3800+ has a TDP of 89W. Price/performance is close if not equal in my experience, and performance per watt of the Core Duo is WAY ahead.
One key here - Intel has a roadmap that includes a clear performance boost beyond that of the Core Duo within months, while AMD only has incremental upgrades (Socket AM2) planned. I've also seen references to a new A64 core with an extra FPU, but nothing about availability of that.
Just because AMD has been the king of the desktop for a while doesn't mean you should count Intel out now that they've finally ditched the Netburst architecture.
"Especially since all new Thinkpads have a fucking TCPA chip. Can you trust a chinese fabbed uber security module for critical national security purposes?"
Wherever that TCPA chip was fabbed, it is almost guaranteed to come from the same source as those found in laptops from any other manufacturer.
If there were ANY chips in the Lenovo that were built in China without a clear paper trail leading back to a non-Chinese supplier (or a Chinese supplier that isn't also selling chips to manufacturers that the government ISN'T freaking out about) then this would be justified. But it's an Intel CPU with an Intel northbridge/southbridge, Intel/ATI/NVidia graphics, the LCD is probably Japanese or South Korean just like nearly every other LCD on the market. The only place where Lenovo could backdoor the machines is in the BIOS (easy to audit) and in the OS (hire a sysadmin to nuke and repave).
I second the issues with GPUs and power management. NVidia's drivers disable clock adjustments on mobile chipsets under Linux currently, and also do not officially support PowerMizer. That said, battery life on my old Dell I8200 and on my new E1705 were both quite good despite these issues. Suspend to disk and suspend to RAM are both very iffy with NVidia or ATI chipsets. I haven't gotten them to work yet on my machine, others say suspend2 works great on their NV-based systems.
Speaking of ATI, avoid them at all costs under Linux. Their drivers are even worse under Linux than under Windows.
The Intel GMA950 is extremely difficult to get running under Linux right now, although historically Intel chipsets have been pretty well documented so this is likely to change in the future.
As to WiFi - most Core Duo-based systems have an Intel PRO/Wireless chipset. In the case of the Dell E1705, it's the IPW3945 a/b/g chipset, and if you go to the sourceforge project for the 3945 and get the driver it works great.
There are kernel patches that allow the SD/MMC readers on the E1705 (and other laptops) to function under Linux, although the last version of the patches I tried didn't like my new 2gig Corsair 133x card. I've never had any problems with any other card though. I haven't tried checking for a new patch recently.
Overall, with the exception of a few minor issues (covered above), all of the hardware on the Dell E1705 with an NVidia chipset works amazingly well under Linux. Reports on the Gentoo forums indicate that ATI-based E1705s are extremely problematic and Intel GMA950s are a royal PITA too at the moment.
Similarly, my Netgear WPN511 is also rock solid with madwifi on my old laptop.
Newer Intel-based laptops almost always come with an Intel PRO/Wireless chipset, all of which are rather well supported. The 3945 isn't in the kernel yet and can be a bit of a pain to install due to dependency on a newer version of the 802.11 stack than what is in the kernel, but once installed it works GREAT in my Dell Inspiron E1705.
Not initially, but every drug eventually has to undergo clinical trials in humans, usually multiple phases with increasing numbers of patients each phase. Clinical trials are a critical final step in FDA approval for any drug. (Of course, animal trials are the usual predecessor to this before the FDA will even allow a clinical trial in humans.)
Some drugs do get shortcutted through the trials phases faster than others. These are usually the class of drugs where the drug may possibly kill you or injure you, but what the drug treats will definately kill you. This is one such case - Even if the new drug has some nasty side effects, they're better than the alternative (MRSA or VRSA), and it sounds like even initial animal trials have indicated fewer side effects than vancomycin, which is a nasty drug that was approved anyway because the alternative to using it is nastier.
Depends on exactly how you define "public airwaves". While "public airwaves" may not be the proper term for what the parent poster meant, by his definition XM isn't public. To receive any XM content other than their preview channel, you must pay them.
There is plenty of precedent for XM being considered different from public OTA broadcasts because of the fact that it is a subscription service. Note that FCC censorship rules do not apply to XM or Sirius because they're subscription services, just as they don't apply to cable TV. Yes, XM and Sirius do some self-sensoring (All XM channels with "questionable" language/content are marked with XL in their channel names, non-XL channels have censored content.)
If you want to charge extra, provide additional service beyond what you already advertise.
Hint: Multicast will solve many of the problems you complain about. There's been a standard for it for HOW long without you implementing it?
(I'm not sure, but I have a feeling a multicast-enabled BitTorrent would use less than a tenth of the bandwidth it currently does for most swarms, if not less.)
This is partially false.
There are multiple types of radiation. There is ionizing radiation (usually from nuclear sources) which can definately cause cancer, and IS cumulative in its effects. It has the effect of essentially flipping bits in your DNA. Flip enough bits and bad things happen.
There is also non-ionizing radiation. All electromagnetic radiation with wavelengths longer than UV falls into this category. This includes visible light and microwaves. There has never been any proof that such radiation can cause cumulative damage unless it exceeds a certain power threshold. In the case of microwaves, tissue damage only occurs when the power level becomes high enough to begin causing thermal heating.
Faulty and malfunctioning equipment could, in theory have significantly higher output power. That said, these faults would be obvious and would result in the tower owner's maintenance people being notified immediately and most likely would involve shutdown of the equipment within minutes.
Power amplifiers for cellular systems need to handle a certain average power, but the signal they transmit has peaks 7-9 dB higher in power than average (somewhere around 6-8 times the average power). Thus the transmitter must be sized to handle the peaks at least in the short term. Thus, an amp capable of 45 watts average can transmit 400 watts, BUT most amplifiers are not designed to handle the accompanying thermal load and power drain and would quickly overheat and shut down if run continuously at peak power for very long.
Even at 400W output power, you'd be completely safe more than 3-4 meters away in free space, and even less than that with a roof in between you and the transmitter.
At my last job, the company I worked for built transmitters for cell towers. We were routinely running these things without their covers on, and microwave emissions leaked all over the building. There was a story that once the leakage alone from a dummy load they were running a demo unit into was jamming Sprint service for the entire conference hall the demo booth was in. Many of my coworkers were working with tower equipment since the very first Motorola demo. There was not any evidence whatsoever of an increased incidence of cancer. If everyone's fears were true, one of my coworkers would've been a hideously deformed mutant instead of an active, healthy (but completely grey-haired) old guy.
The article said the tumors were "consistent with radiation" as the cause. People will continue to be stupid and not realize the difference between electromagnetic radiation (better turn off your light bulbs... There's RADIATION coming from them!) and nuclear radiation. Yes, I know gamma rays are a form of electromagnetic radiation, but they have extremely high energy and can cause ionization, while microwaves cannot cause ionization until their field strength exceeds the breakdown threshold of the dielectric they're in - by the time that happened to someone they'd be thoroughly cooked by thermal heating.
It's an old building, there's probably traces of asbestos or some other nasty construction material that's been banned for decades.
I've used wireless hardware from the following manufacturers:
D-Link - DWL-650, DWL-G650, DI-624
The original 650 wasn't a stellar performer but it wasn't horrible. The G650+624 combo was pretty decent. I only returned it in favor of waiting to see where things went as far as MIMO gear.
Belkin - Can't remember, it was a b-only router
Utter crap. Couldn't last more than 2-3 days without crashing. Died permanently in just over a year.
Microsoft - MN500
In true Microsoft tradition, their software may be crap but their hardware products are actually decent. The MN500 was the most solid and consistent performing 11b kit I have used so far.
Netgear - WPN824 router + WPN511 NIC
Stellar. Utterly stellar. I love the 824. WPN511 is only retired due to the fact that my new laptop has a built-in Intel PRO/Wireless 8945 a/b/g NIC.
"Ok, so it might have marginally better spectral efficiency per Mbps but really, what we want to see is true beamforming dynamic-arrays that will properly 'point' the RF where it's supposed to go in real-time."
Actually, to Airgo's credit, this is exactly what they do. Effectively, they form seperate beams for each possible path between units, and send different data on each. When connected to a "vanilla" B/G unit, they fall back to steering all of the power along the path with the least loss.
OK, yeah, they also implement "turbo" mode (channel bonding of multiple channels) unfortunately, but that isn't where the huge improvements over 108Mbps gear come from.
Although if you want realtime beamforming that's designed to work well with "vanilla" 11g gear, I suggest the Netgear RangeMax (Not RangeMax 240, that's Airgo-based) line (WPN824 router). They combine a (mostly) standard Atheros chipset (It does support Super G including "Turbo" mode, but still works wonders even with all that stuff disabled.) with an adaptive phased array antenna from Ruckus Wireless. The client cards are actually just vanilla Atheros XR (eXtended Range, which simply means 100mW transmit instead of 25 and higher receive sensitivity, both Good Things which are fully compatible with all other 11g gear) chips, NOT the Atheros MIMO chips (which is what D-Link's MIMO gear uses). I'm currently using the WPN824 with a "vanilla" Intel PRO/Wireless chipset most of the time and it still works far better than any other AP I've used.
Um, all of Dell's current laptop products use Intel WLAN chipsets.
Personally in that regard I'd prefer the Dell. The Intel PRO/Wireless silicon is designed for rock-solid performance within the 802.11g standard, not a bunch of flashy compatibility-breaking extensions that will rarely work as advertised. Their chipsets also (unlike Airgo's) work extremely well under Linux.
NOT rebranded Linksys.
Linksys does not make their own silicon. Neither, I believe, does Cisco. (As evidenced by one of the major reasons for Lucent spinning off their microelectronics business as Agere - Despite being a Cisco competitor, Lucent was selling a lot of silicon to Cisco and Agere stood to sell even more as a non-competitor of Cisco.)
Airgo (along with Broadcom and many other companies) are SUPPLIERS of Linksys. Note that this does not make Linksys "rebranded Airgo", as Linksys just buys the ICs and builds a box around them.
By your definition, Intel is "rebranded Dell".
Commenting on the announcement itself - Stock 802.11g is more than sufficient for online gaming. The bottleneck will be one's connection to the outside world for a LONG time. To the geeks of Slashdot, this means that if you want to run Linux on your laptop and have everything work, stay far away from this laptop. Now that reverse engineered Broadcom WLAN drivers exist for Linux, Airgo is basically the only chipset vendor without any Linux support. (At least that was the case as of December when I was shopping for new WLAN hardware.)
I find it interesting that Linksys was perfectly happy to profit from Linux by using it in the WRT54G, but probably sells the highest percentage of non-Linux-compatible WLAN hardware. (Almost all recent Linksys products, especially any that contain extensions beyond vanilla 802.11g such as the SRX and SpeedBooster lines, use Broadcom or Airgo silicon.)
The exec who uttered that quote seems to be rather out of touch with current events.
I'm out of touch and even *I* know that numerous politicians have been calling for an investigation of the oil companies for price gouging/price fixing within the past few weeks.
"It's not clear to me that they'll be any more successful in getting across that this is "new and improved" than they were with the Pentium 2,3,4 series."
Probably because there wasn't much as far as "new and improved" going from 2->3 (same basic architecture, I believe only minor changes such as adding SSE and cache changes, and many of the latter cache changes changed frequently within a revision.)
There was also very little as far as "new and improved" going from 3 to 4. In fact, the initial P4 requirement for RDRAM plus the P4's bad IPC meant that at the time the P4 was initially released, the P3 was a better choice for processor. It wasn't until the P4 scaled up more in frequency that it became the obvious choice, and the return to a P3 descendant makes one wonder if the P4 would EVER have been the clear choice if the P3 architecture had been continually maintained and also continued ramping up in clock speed.
I don't have time to read the detail, but your post and the original article's comment about "recycling power" sounds to me like they are using some sort of adiabatic logic approach. Adiabatic logic is well known for significant power reduction, but at least historically it has required significantly more transistors per gate and cannot run as fast as traditional CMOS.
The ring thing sounds like it's just a new clock generation scheme to go with the existing adiabatic logic techniques (which do have rather unusual clocking requirements that are a bit harder to generate than typical single-square-wave clocks.)
Given the parent's comment about "cheaping down the ingredients", it's less likely that they used higher-end ingredients (Ace-K to replace some of the aspartame - BTW Aspartame=NutraSweet), but used a lower-end ingredient instead.
Did you know that the Diet Coke and (I believe) Pepsi served in restaurant soda fountains is usually not the same as that sold in stores? In many cases it's actually sweetened with saccharin, and that stuff DOES taste nasty, even to someone who has been drinking diet for over a decade. The Coke Light in the Phillipines may likely have been saccharin-sweetened.
Also in the "traditional flavors" arena, try Splenda-based drinks. They're still not the same as regular, although they do taste different. Disclaimer: As an insulin-dependent diabetic who has been drinking nothing but diet sodas for over a decade except in the case of hypoglycemia emergencies, regular sodas taste extremely strange to me. Note that below, assume that I am ALWAYS talking about the diet version.
My personal favorites:
Lime Coke/Pepsi. I can't tell much difference here between the two.
Vanilla Pepsi. Vanilla Coke seems to be not quite as good.
Lemon Coke/Pepsi are OK. Well, one of them is. The other has a distinct medicine-like taste, I can't remember which.
Coca-Cola Blak - This isn't zero-carb, although at 45 calories per bottle it's acceptable even to a diabetic. Of course, that is due partly due to the fact that bottles of Blak are only 8 ounces. You can emulate the taste of Blak somewhat by adding coffee flavoring to Coke. Same technique also works well with Vanilla Coke. This of course depends on the nature of the flavoring. I've been using a bottle of Kahlua-flavored sugar free syrup from DaVinci Gourmet, but I'll be looking into a less expensive solution than the DaVinci syrup soon.
Diet MD Code Red - Oh my this is delicious. As the parent said, it's damn near impossible to find.
I wish they'd release MD LiveWire or Pitch Black in diet form too.
Slice ONE - New to the market. Traditional Slice is just orange, but they have grape (the first major brand to offer diet grape soda in over a decade, until now Diet Polar Grape has been the only option, and that's hard to find and expensive) and carbonated fruit punch flavors. REALLY good stuff, not caffeinated though.
Vault Zero is also good. Despite Coke's marketing, it is NOT an energy drink, it's just highly caffeinated soda similar to MD. IMO Vault is a little smoother tasting, MD has more bite. Vault Zero is difficult to find in any packaging other than single bottles.
In terms of energy drinks, I really like Monster Lo-Carb and MDX. Red Bull and Rockstar are NASTY. Full Throttle sugar free (both FT and FT Fury) are OK.
Hmm, did I miss a buyout of Capcom by Intel?
Intel has recently come up with a series of totally unoriginal and ultimately confusing names for their CPUs.
For example, the "Core Duo" is a pretty unoriginal name for a dual core processor, and I've seen a lot of people start referring to dual core CPUs as "DuoCore" or other such nonsense.
Core 2 Duo? Talk about redundant and confusing naming...
At least some of those (possibly) make sense.
Perhaps there is a band with an album called Weapons of Mass Destruction?
As to body parts - keep in mind that those body parts don't have to be human or animal. Vehicles have body parts too, which are almost surely sold on eBay. Same for dolls. (do a search on eBay itself for body parts...)
Same for warheads - lots of REAL eBay results, none of which are actually explosive, including "warheads" candy.
Given that few fossil fuel powered electric plants in the United States actually use petroleum-based products.
Instead they use coal, which is far more readily available in the United States than petroleum is. In fact I believe we are entirely self sufficient as far as coal goes and may even be exporting significant amounts of it.
MRI
Metallic objects in the body and MRIs don't mix.