I have the download ready and waiting to install but I was looking for threads and articles like this one before deciding on upgrading (I have an iPhone 5 which runs perfectly fine as it is). I am planning on upgrading soon but want to hear about any problems that users are running into. It seems like some are very happy, some not so happy and those in-between. I figure that any show-stoppers will turn up this week.
My son had to buy a phone two months ago (his feature phone died). I suggested the Nexus 4, HTC-1 (Google edition), and S4 (Google Edition). I recommended the Nexus 4 as he's picking up the tab (it's nice when your kids graduate with a CS degree so that they can get a job and pay for their own toys) and that's what he chose and he loves it. He used to carry around a feature phone and an iPod Touch for music and pictures. Now he just has one device with a very good camera, screen, battery life and he doesn't have to keep multiple cables around to charge multiple devices.
My wife's feature phone is going too and I'm planning on getting her a Nexus 5 when it comes out.
I like the file-level stuff that you can do on Android but I prefer the Apps on iOS. So I have an iPhone and a Nexus 7 - I can do what I want between the two of them.
Plan is to run 1,000 miles this year. I do weights a few times a week and extended stretching a few times a week. Also 1.5 to 2.5 hours of tennis one to two times a week. It helps to have a fitness center at the office.
I also use a fairly low-carb diet. The LiveStrong website has a good calorie tracker - diet is more important than exercise in losing fat.
Avoid sugar. Watch the YouTube videos from Lustig on this.
Join a support group.
The hierarchy of fat loss:
Intervals
Strength-Training
High-intensity cardio
Low-intensity cardio
The earlier ones are more efficient for fat loss.
BTW, you can't out-train a crappy diet.
Many homeschoolers in my area are sending their homeschooled kids at 14, 15, 16, and 17 to college. They get their group education in an environment that isn't as caustic as middle- and high-school. They can also earn college credit to reduce college expenses in the future and there's a nice Federal tax credit to pay for the college courses.
Something that I've just read about is parents with students in private school considering homeschooling in order to get them into dual-enrollment programs at community colleges. Some community colleges provide free courses to homeschoolers and tax credits could take care of remaining expenses.
Our kids started college at 15. They get the group benefits and learn things that we're happy to farm out with students that are generally more mature than those in secondary schools.
That's what I thought too. Just download and install the PSDK. Perhaps it also works with VS2005 Express Edition (free too) for the Visual Studio environment. I have VS2005 Pro but used to build
x64 applications with the PSDK.
The random number code in Firefox is a fair amount of 64-bit integer code and some double precision floating point code which is why this test performs quite a bit better on 64-bit browsers than on 32-bit browsers. Of course I'm assuming that Internet Explorer's code is similar. I don't know what Opera uses. But perhaps Firefox would perform better with a 32-bit integer portion of the randome number generator. I had a look at the generated assembler code on an x64 bit build and it found what I'd call two performance bugs in the random number code. There are two constants that are generated. One of them is used for a double-precision floating point divide and the other is used as a mask. The constant is stored in a structure so the compiler can't optimize the division into a multiply by recipricol. I changed the code to do the division by the constant and the resulting performance improvement was five percent. I didn't try it with the mask as the savings would only be the cost of a memory access.
Obviously Opera still has a big edge over Firefox.
I think that the JS garbage collection approach is a major drag on some of the JS benchmarks that I've seen and perhaps that will have to get overhauled for Firefox to do better on benchmarks. I've tried to get performance improvements into Firefox for a while and it's a rather frustrating exercise.
BTW, my 32-bit Firefox results were with my own build.
I always get $200 out of the ATM and never count it - I just stuff it in the back of my wallet and get some more when I run out. If they determined that it gave me too many 20s, then I'd guess that they could just deduct the additional from my account. I wouldn't be able to dispute them.
The ATMs that I use don't give you the option of fives. I guess that's one way to avoid spitting out the wrong bills - just only support 20s.
As far as criminal charges go, seems like a way to improve police metrics. Why call someone on the phone to ask them about it when you can take them away in handcuffs?
Two weeks ago, I walked into an AT&T Store and bought a flip phone for $59.99. Doesn't have much in the way of features but it has address book, makes and receives phone calls and you don't need a plan. I paid $25 for 100 minutes, good for 3 months. So I pay about $8.33 a month for the service. The phone was for my daughter when she's taking classes.
They have a cheaper candy bar model for $39.99. And a more expensive model with camera for $100.
We have two more of these phones. One for me and one for our son. Just what we need for short communications and calling for pizza.
Intel's quad-core is one socket. AMD's FX quad-core is two-sockets (and boatloads of electricity).
AMD has been touting native quad-core for quite some time but Intel's approach has big advantages in manufacturing the chip. If you run into defects on the wafer, you have to toss one big die instead of a die about half the size. Using pairs, you have the flexibility to pair core-pairs that use little power, or pair higher-power cores with lower-power cores. And you have the flexibility to respond easily to market demands for dual-core vs quad-core.
So when you're talking about actually bringing something to market that sells, makes a profit and makes customers happy, Intel's approach is one that's actually been working. They claim to have sold 1 million quad-core chips so far. What's the score for AMD?
Intel delivers Nehalem in the second-half of 2008. What would be incredibly funny is if they demo'd Nehalem this fall. Can you imagine the Architectural advantages of Core 2 Duo with the SSE4 and 45 nm enhancements of Penryn combined with cache coherency and an integrated memory controller? Intel has said that Nehalem is doing very well.
As far as 64-bit Operating Systems go, I would hope that you could use your iPhone with the Mac Pro which runs a 64-bit operating system. Of course if it isn't, then I guess that 16 GB of ram that you can order it with is mostly going to waste.
64-bit computers and operating systems have been around since the 1980s. They make a lot of sense for servers today. They will make their way into desktops soon. For Apple, it should be by the end of 2007.
AMD and Microsoft didn't do a very good job in marketing and in the driver detail work for x64. Microsoft, of course, has been pretty busy with Vista and dealing with security holes. And AMD spent too much time celebrating their success with K8 over Netburst.
I remember the old Task Builder days under RSTS and RSX in the early 1980s where you did program segment overlays to get effectively get larger programs to fit into memory. I'm pretty happy with the flexibility of virtual models and would rather not go back into the age of hacks to address more memory.
I started working on 64-bit processors in the 1980s. Probably MIPs. Was working on DEC Alphas by the 1990s as I worked for DEC back then and we had early experimental prototypes. 64-bits has been around for quite some time.
The current approach is to use the 32-bit version of flash to support the 64-bit version of Firefox. The work to do this hasn't been done for Firefox Windows x64 yet.
They fixed the Win32 installer problem in iTunes 7.
iTunes 6 had a native x64 build which apparently was dropped in 7. I was using the x64 native version but can't find a link to it right now as PlanetAMD64 apparently was hacked and lost the older posts.
If your application uses SIMD programming, then you get 8 extra registers which means that you don't have to toss stuff on the stack for temporary storage. AMD and Intel are adding considerable performance improvements in SIMD instructions (witness SSE4 and Super Shuffle in Penryn and the move to 128-bit instructions from 64-bit SIMD instructions).
Perhaps they can access more than 2 GB of Ram. Or they're running on Itaniums which are 64-bit processors. I've read that there are some hacks that can allow a process to access more than 2 GB and I suppose that one could do some kind of bank switching to technically address more than 2 GB of memory or use a RamDisk.
But the big problem is address space. Addresses are 32 bits in Win32 programs and unless you're doing some kind of special memory mapping, you're limited to 4 GB of address space.
I've accessed 5 GB on my laptop in the past and this thing is three years old. I was running a beta of Windows x64.
The better performance comes for desktop users where memory isn't a concern is that you can operate on 64-bits of data at a time, the additional 8 general purpose registers and the additional 8 vector registers.
I have Windows XP x64 running and use an XP Home Win32 VM running under Virtual PC for applications that won't run or install on x64. One of the applications is for work (VPN) and one of them is for an online class that our son takes.
iTunes did have a native x64 image in the 6.0 line but I think that they pulled it. The Win32 version installs and works fine. I think that there's a bit of a drop in performance on the Win32 version.
I generally prefer working in x64 as it performs better for what I do and I like the VM so that I can be connected to work in one window and connect directly to the internet in all of my other applications.
Slashdot Mirror
I have the download ready and waiting to install but I was looking for threads and articles like this one before deciding on upgrading (I have an iPhone 5 which runs perfectly fine as it is). I am planning on upgrading soon but want to hear about any problems that users are running into. It seems like some are very happy, some not so happy and those in-between. I figure that any show-stoppers will turn up this week. My son had to buy a phone two months ago (his feature phone died). I suggested the Nexus 4, HTC-1 (Google edition), and S4 (Google Edition). I recommended the Nexus 4 as he's picking up the tab (it's nice when your kids graduate with a CS degree so that they can get a job and pay for their own toys) and that's what he chose and he loves it. He used to carry around a feature phone and an iPod Touch for music and pictures. Now he just has one device with a very good camera, screen, battery life and he doesn't have to keep multiple cables around to charge multiple devices. My wife's feature phone is going too and I'm planning on getting her a Nexus 5 when it comes out. I like the file-level stuff that you can do on Android but I prefer the Apps on iOS. So I have an iPhone and a Nexus 7 - I can do what I want between the two of them.
Plan is to run 1,000 miles this year. I do weights a few times a week and extended stretching a few times a week. Also 1.5 to 2.5 hours of tennis one to two times a week. It helps to have a fitness center at the office. I also use a fairly low-carb diet. The LiveStrong website has a good calorie tracker - diet is more important than exercise in losing fat. Avoid sugar. Watch the YouTube videos from Lustig on this. Join a support group. The hierarchy of fat loss: Intervals Strength-Training High-intensity cardio Low-intensity cardio The earlier ones are more efficient for fat loss. BTW, you can't out-train a crappy diet.
Many homeschoolers in my area are sending their homeschooled kids at 14, 15, 16, and 17 to college. They get their group education in an environment that isn't as caustic as middle- and high-school. They can also earn college credit to reduce college expenses in the future and there's a nice Federal tax credit to pay for the college courses. Something that I've just read about is parents with students in private school considering homeschooling in order to get them into dual-enrollment programs at community colleges. Some community colleges provide free courses to homeschoolers and tax credits could take care of remaining expenses. Our kids started college at 15. They get the group benefits and learn things that we're happy to farm out with students that are generally more mature than those in secondary schools.
This person was unschooled by her parents. She seems to be doing fairly well. http://en.wikipedia.org/wiki/Alison_Miller
That's what I thought too. Just download and install the PSDK. Perhaps it also works with VS2005 Express Edition (free too) for the Visual Studio environment. I have VS2005 Pro but used to build x64 applications with the PSDK.
I ran the benchmark on my Dell E521 X2 5600+ and found:
IE: 1515
IE x64: 891
FF: 594
FF x64: 453
Opera 9.23 266
The random number code in Firefox is a fair amount of 64-bit integer code and some double precision floating point code which is why this test performs quite a bit better on 64-bit browsers than on 32-bit browsers. Of course I'm assuming that Internet Explorer's code is similar. I don't know what Opera uses. But perhaps Firefox would perform better with a 32-bit integer portion of the randome number generator. I had a look at the generated assembler code on an x64 bit build and it found what I'd call two performance bugs in the random number code. There are two constants that are generated. One of them is used for a double-precision floating point divide and the other is used as a mask. The constant is stored in a structure so the compiler can't optimize the division into a multiply by recipricol. I changed the code to do the division by the constant and the resulting performance improvement was five percent. I didn't try it with the mask as the savings would only be the cost of a memory access.
Obviously Opera still has a big edge over Firefox.
I think that the JS garbage collection approach is a major drag on some of the JS benchmarks that I've seen and perhaps that will have to get overhauled for Firefox to do
better on benchmarks. I've tried to get performance improvements into Firefox for a while and it's a rather frustrating exercise.
BTW, my 32-bit Firefox results were with my own build.
I always get $200 out of the ATM and never count it - I just stuff it in the back of my wallet and get some more when I run out. If they determined that it gave me too many 20s, then I'd guess that they could just deduct the additional from my account. I wouldn't be able to dispute them.
The ATMs that I use don't give you the option of fives. I guess that's one way to avoid spitting out the wrong bills - just only support 20s.
As far as criminal charges go, seems like a way to improve police metrics. Why call someone on the phone to ask them about it when you can take them away in handcuffs?
Two weeks ago, I walked into an AT&T Store and bought a flip phone for $59.99. Doesn't have much in the way of features but it has address book, makes and receives phone calls and you don't need a plan. I paid $25 for 100 minutes, good for 3 months. So I pay about $8.33 a month for the service. The phone was for my daughter when she's taking classes. They have a cheaper candy bar model for $39.99. And a more expensive model with camera for $100. We have two more of these phones. One for me and one for our son. Just what we need for short communications and calling for pizza.
Intel's quad-core is one socket. AMD's FX quad-core is two-sockets (and boatloads of electricity). AMD has been touting native quad-core for quite some time but Intel's approach has big advantages in manufacturing the chip. If you run into defects on the wafer, you have to toss one big die instead of a die about half the size. Using pairs, you have the flexibility to pair core-pairs that use little power, or pair higher-power cores with lower-power cores. And you have the flexibility to respond easily to market demands for dual-core vs quad-core. So when you're talking about actually bringing something to market that sells, makes a profit and makes customers happy, Intel's approach is one that's actually been working. They claim to have sold 1 million quad-core chips so far. What's the score for AMD? Intel delivers Nehalem in the second-half of 2008. What would be incredibly funny is if they demo'd Nehalem this fall. Can you imagine the Architectural advantages of Core 2 Duo with the SSE4 and 45 nm enhancements of Penryn combined with cache coherency and an integrated memory controller? Intel has said that Nehalem is doing very well.
As far as 64-bit Operating Systems go, I would hope that you could use your iPhone with the Mac Pro which runs a 64-bit operating system. Of course if it isn't, then I guess that 16 GB of ram that you can order it with is mostly going to waste. 64-bit computers and operating systems have been around since the 1980s. They make a lot of sense for servers today. They will make their way into desktops soon. For Apple, it should be by the end of 2007. AMD and Microsoft didn't do a very good job in marketing and in the driver detail work for x64. Microsoft, of course, has been pretty busy with Vista and dealing with security holes. And AMD spent too much time celebrating their success with K8 over Netburst.
Here's a list from September 2006 from PlanetAMD64. BTW, you can add Mathematica to this list and, I'd guess, quite a few more applications. Antivirus / Spyware Software avast! Antivirus Symantec Antivirus Corp. 10 NOD32 AVG Antivirus Ghost Security Suite (Beta) System Utilities Stratup Monitor64 SpeedCommander Filemon OpenGL Extensions Viewer Prime95 Process Explorer Regmon ActivePython UltraMon DVDInfoPro 64 Stata TweakNow eXtended Task Manager Process Lasso dtSearch (Beta) MD5 Checksum Tool Benchmarking / Overclocking Tools Sisoftware Sandra 2007 Fraps CineBench SpeedFan PCI Latency Tool v2 SysTool Desktop / UI Customization True Launch Bar WindowBlinds StyleXP File Compression Squeez 7-Zip sfvsub 1.4 (Beta) Disk Management WinImage Filedisk O&O Defrag Diskeeper 10 Silicon RAID Management Utility Graphic Apps POV-Ray Softimage | XSI RealWorld Icon Editor Autodesk Maya Maxon Cinema 4D Internet / Email Apps Mozilla Firefox x64 Sun Java x64 PeerGuardian 2 Mozilla Thunderbird x64 Arctic Torrent cFosSpeed Firewall / Security Tools Tiny Firewall 64 UltraVNC With Firewall Control Networking CommView CommView Remote Agent NetResident CommTraffic FTP Software SmartFTP Audio / Video Editing / Production Cakewalk Sonar x64 VirtuaDub DVDx CD/DVD Authoring AnyDVD (Beta) CloneDVD (Beta) CloneCD (Beta) DAEMON Tools X64 Microsoft .NET Framework 2.0 x64
IntelliType Pro 5.5
IntelliPoint 5.5
Windows Media Encoder 9 x64 .NET Framework 2.0 SDK x64
Windows Defender x64 (Beta)
Internet Explorer 7 x64
Windows PowerShell x64
DirectX (8/8/06)
DirectX SDK (8/8/06)
PostScript Interpreters
AFPL Ghostscript
GSview
Games
Half Life 2
Far Cry
Unreal Tournament 2004
Unofficial
TweakUI64 (UI Tweaking)
PuTTy x64 (Networking)
Misc.
Gozoku x64 (Japanese to English dictionary)
I remember the old Task Builder days under RSTS and RSX in the early 1980s where you did program segment overlays to get effectively get larger programs to fit into memory. I'm pretty happy with the flexibility of virtual models and would rather not go back into the age of hacks to address more memory. I started working on 64-bit processors in the 1980s. Probably MIPs. Was working on DEC Alphas by the 1990s as I worked for DEC back then and we had early experimental prototypes. 64-bits has been around for quite some time.
The current approach is to use the 32-bit version of flash to support the 64-bit version of Firefox. The work to do this hasn't been done for Firefox Windows x64 yet.
They fixed the Win32 installer problem in iTunes 7. iTunes 6 had a native x64 build which apparently was dropped in 7. I was using the x64 native version but can't find a link to it right now as PlanetAMD64 apparently was hacked and lost the older posts.
If your application uses SIMD programming, then you get 8 extra registers which means that you don't have to toss stuff on the stack for temporary storage. AMD and Intel are adding considerable performance improvements in SIMD instructions (witness SSE4 and Super Shuffle in Penryn and the move to 128-bit instructions from 64-bit SIMD instructions).
Perhaps they can access more than 2 GB of Ram. Or they're running on Itaniums which are 64-bit processors. I've read that there are some hacks that can allow a process to access more than 2 GB and I suppose that one could do some kind of bank switching to technically address more than 2 GB of memory or use a RamDisk. But the big problem is address space. Addresses are 32 bits in Win32 programs and unless you're doing some kind of special memory mapping, you're limited to 4 GB of address space. I've accessed 5 GB on my laptop in the past and this thing is three years old. I was running a beta of Windows x64. The better performance comes for desktop users where memory isn't a concern is that you can operate on 64-bits of data at a time, the additional 8 general purpose registers and the additional 8 vector registers.
I have Windows XP x64 running and use an XP Home Win32 VM running under Virtual PC for applications that won't run or install on x64. One of the applications is for work (VPN) and one of them is for an online class that our son takes.
iTunes did have a native x64 image in the 6.0 line but I think that they pulled it. The Win32 version installs and works fine. I think that there's a bit of a drop in performance on the Win32 version.
I generally prefer working in x64 as it performs better for what I do and I like the VM so that I can be connected to work in one window and connect directly to the internet in all of my other applications.