A competitive product is the Linear Technology LTC3441
which also dynamically moves from buck to boost to compensate for draining batteries.
I used it to provide a stable output voltage regardless of whether my input was 2.7-3.3V or 4.7 to 5.0V
No, the bandwidth (2.5 Gbps in each direction) of the 1x PCI Express Lane in an ExpressCard can't compete with an AGP8X built in solution (2.1 GB/sec). The ExpressCard power limitations would keep it from working as well, but primarily its bandwidth.
However, the beauty of PCI Express (and ExpressCard) is that your entire machine will be using the same buses for all communication (from the Northbridge on...). None of this AGP for the GPU, PCI for the other devices and an adapter chip to provide CardBus/16-Bit PC Card support that adds complexity and cost to every motherboard. Each device will speak PCI Express and/or USB. All you need is a good Northbridge and the appropriate connectors.
Two ExpressCard/32 cards will fit in the space currently used by a PC Card slot. And the fact that the PCIExpress/USB connection simply taps into the existing buses means that your ExpressCard connectors don't have to be co-located. (Right now your CardBus/PC Card slot MUST be as close as possible to the (typically TI) Cardbus Host Adapter.)
Similarly it should be trivial to add an ExpressCard connector to a desktop (as all new desktops have PCIExpress and USB) and your ExpressCard modules can be reused.
Software isn't an issue because PCI Express is simply a Physical Layer change. All existing PCI code will work with PCI Express out of the box.
There's a fly in the ointment though. Along with ExpressCard, there's also PCI Express Mini Card. PCI Express Mini Card (51mm x 30mm) is very similar in physical size to ExpressCard (75mm x 34mm), but though they both have USB and PCIExpress connections, the connectors are completely different. ExpressCard is a mere 26 pins, but PCI Express Mini Card is 52 (and a PCI Express 1x is 36).
ExpressCard and PCI Express Mini Card aren't competitors, though. PCI Express Mini Cards are for internal connections while your ExpressCard is physically encased and protected. As a hardware designer, though, I wish the two standards shared a bit more commonality. If, at least, the PCB sizes could be exactly the same, we would probably see most ExpressCard products come out as PCI Express Mini Cards (and visa versa). I don't know if that will happen as the circuit board would have to be completely redesigned (for space-constrained designs).
Oh, and one more thing...
PCI Express, even in its 32 Lane (8 GB/sec unencoded) form, has too much latency for direct memory access. Hypertransport does not suffer from this same deficiency.
More like a two step plan and a setup for attack
on
A Six-Step Plan for Apple
·
· Score: 2, Interesting
Unfortunately Alex seems to like making mileage out of a couple of old ideas.
Cheap
Points one and two are redundant. People want cheap and they find it at Target. But Alex overlooks Apple's method of compensating for lower volume sales (than say Dell) - gross margins. Apple's healthy margins are what have helped it accumulate is near $5 billion cash. If you try to play the "cheap" high volume game with Dell, you'll end up like Gateway - bleeding to death.
Point three is synonymous with points one and two. To sell something cheap, its typically no-frills and as basic as possible. Selling a headless iMac just pushes the display revenue into someone else's pocket and kills your consumer-oriented style. But I concede that a product reminiscent of the LC may spread appleseeds into non-BMW families. But all in one is much simpler than headless for newbies - the tradeoff may be worth it, but its hard to say.
Dell's move with the iPod bounty almost screams "product failure". If the DJ isn't selling of its own merits, then why would I want to swap my well-loved iPod for one? Apple's position is more healthy with its desktops than Dell is with its DJ. An interesting competitive upgrade idea but more of a last-ditch effort.
As for try before you buy, what the hell do you think the Apple Retail Stores are for? Salkever must not be hanging around his local Apple store enough. The one here has a steady stream of people just coming in to play with the machines. But the stores don't have to worry about sending out 10+ iBooks everyday to people who may never return them, and if they did return them, they'd certainly have to be in non-mint condition and have to be sold at a discount to someone else. I know I never want to buy a non-mint Apple if I'm paying the Apple-premium, and I don't expect anyone else to.
Businesses aren't going to just disappear with xServes like consumers will, and any unpurchased trial machines being sold at a discount will negatively impact gross margins.
Any prideful statement about a lack of viruses and exploits is nothing more than a HUGE invitation to be attacked and exploited. Security through obscurity is wrong, but so is inviting mayhem if you're not absolutely convinced that you will be able to withstand the attack. As the user-base grows, the level of security confidence should also increase, but don't set yourself up for a potential black-eye.
Actually if you SINGLE click on the removable devices icon in the system tray, you can disable/remove a device with a single click.
Just as easy as clicking the eject icon in the Mac OS X Finder. Just avoid double clicking.
Actually, European operators ALL have UMTS plans and they have either started to roll it out or will be sometime this year.
Vodaphone, T-Mobile, Hutchison, Telephonica, Orange, O2, etc.. They all bought spectrum licences and are now legally bound to roll out UMTS service THIS year.
UMTS in Europe runs in its own 2100 band and therefore does NOT interfere or disturb existing GSM/GPRS traffic in the EGSM and DCS bands.
That is, they DON'T take any spectrum away from existing users, they're simply using NEW spectrum to support the new service.
Your points are very valid. However, for Europe, UMTS is running in the 2100 band which will keep it from affecting existing traffic in the EGSM and DCS bands.
UMTS in North America is running on the already congested 1900 band, but I do believe that it uses either a smaller band width or smaller channel separation (I don't recall).
1) your usage of the word "shared" isn't clear in its meaning, so I'll just ignore that statement.
2) Security. Each CDMA phone in the world is has a unique ESN that is hardcoded into the phone. Even if it were possible to reprogram a phone with a duplicate ESN, no two mobiles would be allowed on the network with the same ESN. Both mobiles would be denied service and your account most certainly would be flagged. The ESN is used to create a unique offset in the main long orthogonal code (41 days long) that enables your "calls" to be uniquely encrypted/encoded with your own version of this orthogonal code (in combination with Walsh Codes and Turbo Codes). Not to mention the fact that all packes are "chipped" up and reorganized and duplicated into a random order to improve error correction.
48 users was (about) the maximum number that an AMPS system could handle. The technology has significantly improved in the last 10 years. Try 100+ users per cell.
3) This statement, as a blanket statement about VOIP is FUD. But over EVDO I'd graciously admit that you're probably true. Its designed for Data, not Voice.
4) EVDO eats no more "spectrum" than 1xRTT or IS-98 or IS-95. The other standards don't "coexist" with voice, they CARRY voice. Data over the older standards (even 1xRTT) was a side benefit. EVDO can co-exist in the same sell as 1xRTT handling Voice traffic. The system designers aren't the morons you seem to be implying they are. Just think about it.
"Sorry folks, no cell phones work within this 5 mile area. Data modems only!!!"
802.11 will do nicely within your office building or at your coffee shop. It won't do as you ride in a cab from the airport to your hotel to the conference center to the local park bench (all without having to scan for a new, open, AP).
And EVDO can lay over a 1xRTT network - so indeed they are intermingling.
Oh, and UMTS runs on the 1900 band in North America and the 2100 band in Europe. So much for penetrating walls! Better hope for a good multipath through the windows!
EDGE is better than GPRS but is limited by the spectral inefficiency of its GSM heritage. CDMA (Code Division Multiple Access) based systems are by their nature far more efficient. That's why ALL wireless (cellular) standards are migrating toward and incorporating advanced CDMA techniques and technology.
UMTS (Universal Mobile Telecommunications System) is the CDMA-based successor to GPRS (& EDGE). Its theoretical transfer ceiling (or the ceiling of the Qualcomm UMTS chipset) is 384 kbps down and 64 kbps up. UMTS will be replaced in a few years with HSDPA (High Speed Data Packet Access) which has 7 a Mbps ceiling (or there abouts). UMTS is primarily a European (and Asian) standard but North American GSM/GPRS providers will migrate forward to it.
UMTS is NOT faster than Verizon's EVDO service nor will it ever be. UMTS IS faster than Verizon's (and Sprint's) older CDMA 1xRTT service.
CDMA 1X (RTT) has a theoretical ceiling of 152 kbps. It and EDGE are probably quite comparable in terms of throughput. CDMA 1xEVDO is a significant step above any of these other services.
EVDO stands for EVolution Data Only. This means that you will NEVER be able to purchase an EVDO phone (unless of course it was a VOIP phone...) EVDO's big brother, EVDV will allow for both high speed Data and the traditional Voice coverage. EVDO and 1xRTT can be serviced by the same base station. Its as simple as adding a Nortel (or other) card to the base station and Voila! Verizon is providing EVDO. As it is a Data Only standard, it has been designed from the ground up to provide high speed data access and to do it well. Its built to service a large number of data customers off the same cell. Its completely different from GPRS where voice customers slow your data connection.
EVDO is in face Always On, even as your cellular phone is Always On. You don't dial in and you will probably want to disconnect if you're not actively using the connection (and want to save batteries).
You will NOT be able to flash upgrade a phone to EVDO. EVDO can provide high data rates because of its computational intensiveness. Remember Turbo Codes? They help provide such significant data rates, but demand incredible computing power. Your older phone's baseband processor, and RF Chipset, just don't cut it. Another EVDO technique is its diversity receiver (two antennas and two rake receivers) that improve the signal to noise ratio and help the modem decode weaker and more corrupt signals than any other system could allow.
And for security, this has WiFi annhilated. Do you worry about people stealing your CDMA (voice) phone's ESN, learning your Walsh Codes and calculating the time offset that your phone is using in the main (long) orthogonal code (streaming by at 1.2288 Mega chips per second) of the CDMA/GPS system in real time? Its as secure as the CDMA voice system that we all trust. Certainly higher authorities, with access to the base stations and ESN information are capable of "listening in" on your traffic, but unless you have direct access to the Base Station Controllers, there's no way for anyone to "listen in".
Slashdot Mirror
A competitive product is the Linear Technology LTC3441 which also dynamically moves from buck to boost to compensate for draining batteries. I used it to provide a stable output voltage regardless of whether my input was 2.7-3.3V or 4.7 to 5.0V
This has nothing to do with the Blackberry Product itself though. Here's a good synopsis of the system in question and why the US patent shouldn't necessarily have effect. http://patentlaw.typepad.com/patent/2005/01/canada _challeng.html
No, the bandwidth (2.5 Gbps in each direction) of the 1x PCI Express Lane in an ExpressCard can't compete with an AGP8X built in solution (2.1 GB/sec). The ExpressCard power limitations would keep it from working as well, but primarily its bandwidth.
However, the beauty of PCI Express (and ExpressCard) is that your entire machine will be using the same buses for all communication (from the Northbridge on...). None of this AGP for the GPU, PCI for the other devices and an adapter chip to provide CardBus/16-Bit PC Card support that adds complexity and cost to every motherboard. Each device will speak PCI Express and/or USB. All you need is a good Northbridge and the appropriate connectors.
Two ExpressCard/32 cards will fit in the space currently used by a PC Card slot. And the fact that the PCIExpress/USB connection simply taps into the existing buses means that your ExpressCard connectors don't have to be co-located. (Right now your CardBus/PC Card slot MUST be as close as possible to the (typically TI) Cardbus Host Adapter.)
Similarly it should be trivial to add an ExpressCard connector to a desktop (as all new desktops have PCIExpress and USB) and your ExpressCard modules can be reused.
Software isn't an issue because PCI Express is simply a Physical Layer change. All existing PCI code will work with PCI Express out of the box.
There's a fly in the ointment though. Along with ExpressCard, there's also PCI Express Mini Card. PCI Express Mini Card (51mm x 30mm) is very similar in physical size to ExpressCard (75mm x 34mm), but though they both have USB and PCIExpress connections, the connectors are completely different. ExpressCard is a mere 26 pins, but PCI Express Mini Card is 52 (and a PCI Express 1x is 36).
ExpressCard and PCI Express Mini Card aren't competitors, though. PCI Express Mini Cards are for internal connections while your ExpressCard is physically encased and protected. As a hardware designer, though, I wish the two standards shared a bit more commonality. If, at least, the PCB sizes could be exactly the same, we would probably see most ExpressCard products come out as PCI Express Mini Cards (and visa versa). I don't know if that will happen as the circuit board would have to be completely redesigned (for space-constrained designs).
Oh, and one more thing...
PCI Express, even in its 32 Lane (8 GB/sec unencoded) form, has too much latency for direct memory access. Hypertransport does not suffer from this same deficiency.
Unfortunately Alex seems to like making mileage out of a couple of old ideas.
Cheap Points one and two are redundant. People want cheap and they find it at Target. But Alex overlooks Apple's method of compensating for lower volume sales (than say Dell) - gross margins. Apple's healthy margins are what have helped it accumulate is near $5 billion cash. If you try to play the "cheap" high volume game with Dell, you'll end up like Gateway - bleeding to death.
Point three is synonymous with points one and two. To sell something cheap, its typically no-frills and as basic as possible. Selling a headless iMac just pushes the display revenue into someone else's pocket and kills your consumer-oriented style. But I concede that a product reminiscent of the LC may spread appleseeds into non-BMW families. But all in one is much simpler than headless for newbies - the tradeoff may be worth it, but its hard to say.
Dell's move with the iPod bounty almost screams "product failure". If the DJ isn't selling of its own merits, then why would I want to swap my well-loved iPod for one? Apple's position is more healthy with its desktops than Dell is with its DJ. An interesting competitive upgrade idea but more of a last-ditch effort.
As for try before you buy, what the hell do you think the Apple Retail Stores are for? Salkever must not be hanging around his local Apple store enough. The one here has a steady stream of people just coming in to play with the machines. But the stores don't have to worry about sending out 10+ iBooks everyday to people who may never return them, and if they did return them, they'd certainly have to be in non-mint condition and have to be sold at a discount to someone else. I know I never want to buy a non-mint Apple if I'm paying the Apple-premium, and I don't expect anyone else to.
Businesses aren't going to just disappear with xServes like consumers will, and any unpurchased trial machines being sold at a discount will negatively impact gross margins.
Any prideful statement about a lack of viruses and exploits is nothing more than a HUGE invitation to be attacked and exploited. Security through obscurity is wrong, but so is inviting mayhem if you're not absolutely convinced that you will be able to withstand the attack. As the user-base grows, the level of security confidence should also increase, but don't set yourself up for a potential black-eye.
Actually if you SINGLE click on the removable devices icon in the system tray, you can disable/remove a device with a single click.
Just as easy as clicking the eject icon in the Mac OS X Finder. Just avoid double clicking.
Actually, European operators ALL have UMTS plans and they have either started to roll it out or will be sometime this year.
Vodaphone, T-Mobile, Hutchison, Telephonica, Orange, O2, etc.. They all bought spectrum licences and are now legally bound to roll out UMTS service THIS year.
UMTS in Europe runs in its own 2100 band and therefore does NOT interfere or disturb existing GSM/GPRS traffic in the EGSM and DCS bands.
That is, they DON'T take any spectrum away from existing users, they're simply using NEW spectrum to support the new service.
Your points are very valid. However, for Europe, UMTS is running in the 2100 band which will keep it from affecting existing traffic in the EGSM and DCS bands.
UMTS in North America is running on the already congested 1900 band, but I do believe that it uses either a smaller band width or smaller channel separation (I don't recall).
FUD, plain and simple
1) your usage of the word "shared" isn't clear in its meaning, so I'll just ignore that statement.
2) Security. Each CDMA phone in the world is has a unique ESN that is hardcoded into the phone. Even if it were possible to reprogram a phone with a duplicate ESN, no two mobiles would be allowed on the network with the same ESN. Both mobiles would be denied service and your account most certainly would be flagged. The ESN is used to create a unique offset in the main long orthogonal code (41 days long) that enables your "calls" to be uniquely encrypted/encoded with your own version of this orthogonal code (in combination with Walsh Codes and Turbo Codes). Not to mention the fact that all packes are "chipped" up and reorganized and duplicated into a random order to improve error correction.
48 users was (about) the maximum number that an AMPS system could handle. The technology has significantly improved in the last 10 years. Try 100+ users per cell.
3) This statement, as a blanket statement about VOIP is FUD. But over EVDO I'd graciously admit that you're probably true. Its designed for Data, not Voice.
4) EVDO eats no more "spectrum" than 1xRTT or IS-98 or IS-95. The other standards don't "coexist" with voice, they CARRY voice. Data over the older standards (even 1xRTT) was a side benefit. EVDO can co-exist in the same sell as 1xRTT handling Voice traffic. The system designers aren't the morons you seem to be implying they are. Just think about it.
"Sorry folks, no cell phones work within this 5 mile area. Data modems only!!!"
802.11 will do nicely within your office building or at your coffee shop. It won't do as you ride in a cab from the airport to your hotel to the conference center to the local park bench (all without having to scan for a new, open, AP).
You CAN intermingle UMTS/EDGE/GSM (well except for EDGE right now) Otherwise how can you explain the Novatel Wireless U530 and the Option 3Globetrotter
And EVDO can lay over a 1xRTT network - so indeed they are intermingling.
Oh, and UMTS runs on the 1900 band in North America and the 2100 band in Europe. So much for penetrating walls! Better hope for a good multipath through the windows!
EDGE is better than GPRS but is limited by the spectral inefficiency of its GSM heritage. CDMA (Code Division Multiple Access) based systems are by their nature far more efficient. That's why ALL wireless (cellular) standards are migrating toward and incorporating advanced CDMA techniques and technology.
UMTS (Universal Mobile Telecommunications System) is the CDMA-based successor to GPRS (& EDGE). Its theoretical transfer ceiling (or the ceiling of the Qualcomm UMTS chipset) is 384 kbps down and 64 kbps up. UMTS will be replaced in a few years with HSDPA (High Speed Data Packet Access) which has 7 a Mbps ceiling (or there abouts). UMTS is primarily a European (and Asian) standard but North American GSM/GPRS providers will migrate forward to it.
UMTS is NOT faster than Verizon's EVDO service nor will it ever be. UMTS IS faster than Verizon's (and Sprint's) older CDMA 1xRTT service.
CDMA 1X (RTT) has a theoretical ceiling of 152 kbps. It and EDGE are probably quite comparable in terms of throughput. CDMA 1xEVDO is a significant step above any of these other services.
EVDO stands for EVolution Data Only. This means that you will NEVER be able to purchase an EVDO phone (unless of course it was a VOIP phone...) EVDO's big brother, EVDV will allow for both high speed Data and the traditional Voice coverage. EVDO and 1xRTT can be serviced by the same base station. Its as simple as adding a Nortel (or other) card to the base station and Voila! Verizon is providing EVDO. As it is a Data Only standard, it has been designed from the ground up to provide high speed data access and to do it well. Its built to service a large number of data customers off the same cell. Its completely different from GPRS where voice customers slow your data connection.
EVDO is in face Always On, even as your cellular phone is Always On. You don't dial in and you will probably want to disconnect if you're not actively using the connection (and want to save batteries).
You will NOT be able to flash upgrade a phone to EVDO. EVDO can provide high data rates because of its computational intensiveness. Remember Turbo Codes? They help provide such significant data rates, but demand incredible computing power. Your older phone's baseband processor, and RF Chipset, just don't cut it. Another EVDO technique is its diversity receiver (two antennas and two rake receivers) that improve the signal to noise ratio and help the modem decode weaker and more corrupt signals than any other system could allow.
And for security, this has WiFi annhilated. Do you worry about people stealing your CDMA (voice) phone's ESN, learning your Walsh Codes and calculating the time offset that your phone is using in the main (long) orthogonal code (streaming by at 1.2288 Mega chips per second) of the CDMA/GPS system in real time? Its as secure as the CDMA voice system that we all trust. Certainly higher authorities, with access to the base stations and ESN information are capable of "listening in" on your traffic, but unless you have direct access to the Base Station Controllers, there's no way for anyone to "listen in".
This is not your Mom's 900 MHz cordless phone!