I'm assuming your ancient system uses an AGP interface for graphics, which has a very fast download rate, but very poor upload. The maximum performance of AGP uploading data from the card memory to the rest of the machine is pretty slow (less than 100MB/sec, IIRC), and it will vary depending on the implementation. This is probably the reason you got such slow benches.
PCIe will likely give you performance more in-line with main memory (most implementations now are hitting 1-2 GB/s).
Exactly. The price of electronics is constantly dropping, so people are unwilling to spend more on newer devices; they expect that newer devices will cost the same, or possibly less. This is why you cannot bring inflation into a value consideration.
Let me clarify why you really can't use inflation as a direct comparison.
Today, people expect to pay the same or less for electronics they bought a decade or more ago. The difference today is, (1) they get more features for their dollar (convergence), and (2) thanks to the convergence and lack of price increase coupled with increased spending power, they can buy many times more electronic devices than they used to in the 80s and 90s.
Today, what 2-child family can live without multiple TVs, DVD players, home and portable gaming systems, computers, stereo systems, mp3 players, cell phones, etc. People have been trained by the industry to buy "more" rather than "better," because it is a far easier sell. Your "average" family today can afford many times the number of these devices they could 20 years ago, and so they play the game they have grown used to: buy more devices at similar prices. Unfortunately, this marketing method means it is harder than ever to upsell on quality and featureset, because people know they only have to wait a little bit for the same device with the same features to get cheaper.
Yes, in the past both compiler developers and chip designers have been known to optimize for certain aspects of benchmarks. Thus, to make numbers at all useful, you would have to have multiple benchmarks stressing different aspects of computing. This would prevent optimizing for benchmarks while hurting performance in other areas.
In addition to that, you have the problem that computer usage needs are constantly changing. This is the main reason why benchmarks change from year-to-year. How are you supposed to compare processor models when they advertise benchmark numbers from different versions of a given test suite?
In the end, marketing people from both AMD and Intel have discovered one ultimate truth: in a world where anyone and their dog can sell you an x86 computer, price is one of the easiest ways to gauge performance. People who care about performance will simply pay more and expect it. The rest of the world simply doesn't care, because they assume any processor is fast enough for their needs (and are probably right).
The same buffer configuration with full 720p resolution uses up 14MB of ram. Sure, you can move your framebuffer outside of the EDRAM, but you'll see a significant performance hit.
The dual-bus interface would also solve one of the problems of USB2 versus 1.1 backward compatibility, where a slower device on a shared bus will slow down high-speed devices. This is because bus transactions at the lower speed inherently take longer to complete, reducing the time slices available to the high-speed devices.
With USB3, you've got fibre for the high speed, and copper for the low speed.
Right, but there is one advantage to triple cores that Intel will be missing out on: A clearer and more complete lineup to fit various processing needs.
Right now, Intel sells the quad cores at a premium, and sells dual cores dirt-cheap. To the consumer who can't translate model numbers into actual clock speeds (most of the population), there is no clarity in the product lineup besides this division between two and four cores.
Really, would you be as willing to shell out extra money for a higher model number, if all the processors are already "Dual Core?"
AMD has a similar problem with their model numbering scheme, but now they can offer a clearer product lineup for stupid consumers: they can fill that wide gap between two and four cores with a triple-core option.
I mean, unless you can magically take a straight-line course between the two planets, how do you account for the off-trajectory component of the force exerted by the laser? The ship will be taking it's own course independent of the two planets, which means that the angle between the laser source and the ship will be constantly changing, and the laser will always be off-angle to the optimal.
Is there something I'm missing here? Is there an engine onboard to counteract the excess thrust? Are they planning on having multiple laser installations at different points in space to make this work? Am I missing a simple physics trick?
I used my G400 MAX to play Counterstrike, Quake III, MDK2, etc. The 32-bit Direct3D performance was second to none. The OpenGL performance using the TurboGL was outstanding, and supported all OpenGL games in existence. I had a lot of faith in Matrox, until they casually ignored the powerful GeForce 256 and announced the G450. What a horrible memory.
Sooooo,it is analagous to a tiny mercury transmission line?
That's exactly what I was thinking. This sounds very much like delay line memory, except at the MEMS scale. Everything that is old is new again.
Delay line memory actually makes sense because you can pack so many more bits into it than a normal SRAM or Flash cell. Normal cells require interface logic for every bit (or every 2-4 bits on the most advanced Flash processes, thanks to multi-level cells). With this loop-based technology, you can place several times more bits at a time on the loop, and build only one controller to manipulate all those bits as they pass through.
What can I say? I'm disappointed that they stuck with a 3-issue architecture - while it is true that Intel's 4-issue setup is often data-starved, even with exceptional I/O performance AMD can only hope to match the Core platform in most situations. The lack of progress in their cache technology means AMD gets as much burden as benefit out of the L3 cache over 20ns access time!).
In the I/O arena, AMD potentially has the edge, and for HPC there's no question Barcelona will do well: this architecture is built for scaling beyond 8 cores (potentially as high as 8 sockets, 32 cores). The question is: how big is that market, and will it really pay-off for them?
In the meantime, small-scale servers and workstations are the majority of the market, and Intel still has a commanding lead in that space. Intel's introduction of the San Clemente Xeon chipset with DDR2 instead of FBDIMMs will not help things for AMD, because these systems can most certainly compete with AMD's low power consumption.
And much liks Vista, all three of the above operating systems had service pack releases within the first year. I'd say this puts Vista well on-track for three service packs.
That's the way it is: most of these Steam gamers are still CS junkies...I mean CS 1.6 plus CS:Source combined. If you want to know why people don't upgrade, you just have to look at the numbers Valve's survey doesn't tell you:
You can play CS 1.6 on Intel GMA series graphics...hell, you can even run software mode if that's your bag. CS:Source runs decently even on a Radeon 9600 Pro or GeForce 6600, so there's no real reason to upgrade there. Why upgrade when the game you play every day is smooth?
Exactly. If I made $1 last year, and my friend made $1 million last year, we both "averaged" $500,000. Which is a figure not even close to representing what either of us actually made.
Your example fails because your sample size is too small. A sample size of "all men" would not have this problem when you apply an average.
Your example fails again because you assume that there is somehow a gap within the distribution of the size of all men. The distributed size of all men would be a fairly continuous dataset (translation: there are men of all shapes and sizes). This means the average is a very good tool with which to draw conclusions.
You don't notice it, initially, but eventually it grates on you.
Example: I own the Queens of the Stone Age: Songs for the Deaf album mentioned on the Wikipedia loudness war page. I bought the album because the songs sounded pretty nice. I think the songs are hard-hitting, and fun.
I have noticed one issue: I have NEVER been able to sit down and listen to the entire album in one sitting. This is something I can do with almost any album I own, even heavy metal albums (even previous albums from QotSA); but with Songs for the Deaf, even though the music is great, just the effort of listening really wears on me.
Now that I'm aware of the problem, I try to pay more attention.
Back to your point, my ears can't really make out the difference between a CD and an MP3 as long as the MP3 is encoded at a decent bitrate. I can clearly make out the difference between a 128kbps and 256kbps, but not between 256kbps and 320kbps or VBR.
I have to agree, it is hard to tell the difference these days. I've been a long-time proponent of VBR: r3mix was the first encoder setting besides 256 or 320k CBR where I couldn't tell the difference between the CD and the compressed file, even on my Allesandros. Today's VBR settings are far more impressive, with alt preset standard pushing the limits of audio quality with mp3.
True story: recently, for about 6 months I was accidentally encoding my alt preset standard mp3s with a peak bitrate locked at 224k. I encoded DOZENS of CDs without noticing. When I finally noticed, I re-encoded for consistency's sake, but I couldn't tell the difference. That's how good LAME is today: ABR of 160-192k is transparent, even without 320k peaks.
Yes, the quad-core chips will have the fourth link. In addition, the chips will be able to split their 16-bit HT links into dual 8-bit HT links, allowing for 8-way CPU configurations without hops (8 x 8-bit HT links per socket). In reality, this is the reason why AMD is pushing the new HyperTransport 3.0: so they can cut the bus lines to 8 without sacrificing too much bandwidth.
Sure, it's nice to have single data sets which can occupy more than ~2^30-2^32 bytes (depending on alignment and bit robbing for tags) without having to window them, but such applications are rare -- most processes that use pointers will never come anywhere close to the limits of a 32-bit pointer.
No, not for the near future. But multiple applications + OS memory + hardware address mapping can quite easily use up the space of one 32-bit pointer, and that is the reason for my post. If Intel had made more of an effort to standardize the industry on 64-bit architecture, or alternatively made a stronger push to support PAE at the motherboard level (full 32-bit pointer space for every application), we wouldn't be having this conversation.
This makes requiring x86-64 more important than you imagine it is, because there's nothing optional in the standard (unlike PAE hardware address remapping and 36-bit address extension in IA32). If the OS requires it, then OEMs are forced to offer complete support, which means no more address space games.
The major weakness of IA32 is that the ISA has no native 64-bit datatype, so any manipulations of 64 bit datatypes require lots of extra instructions which are superfluous in an architecture where chip internals handle 64 bit words as quickly as 32 bit ones.
No, it's not really. You just stated that there's no need (for the moment) of 64-bit ints if you can break the 4GB barrier with clever paging, and virtual address spaces beyond 2^32...and because of this, once again I have to point back to the poor implementation of PAE as one of the biggest downfalls in modern systems.
The major advantage of x86-64 is therefore not in pointers, nor in integers, being 32-bits-only, because on other architectures LP64 tends to outperform ILP64 for most integer-heavy tasks, and 32 bit pointer heavy tasks tend to outperform the same tasks done with 64 bit pointers, for similar reasons.
The major advantage of x86-64 is that support for large memory spaces is not optional. In the industry, everyone is trying to shave a buck, and it's tempting to not include support for an "enterprise-level" feature in mainstream motherboards and chipsets. It's unfortunate, but because PAE started in the server world, much like PCI-X it will live and die in the server world.
This problem has many causes. I will now lay blame on the appropriate parties.
1. Intel. Yes, Intel, our savior. Intel did a very bad thing for the future of computing when they refused to add 64-bit support to the Pentium M line. They then continued the disservice by shipping the CoreDuo, continuing their legacy of 32-bit only for the notebook platform. Only now in 2007 are we seeing the last 32-bit only mobile chips come off the mainstream market (they're used for embedded devces only, where they belong).
2. Microsoft, for not having the balls to push the 64-bit Vista as the "standard" platform. The OEM versions of Vista are actually 32-bit only. You can understand this decision when you consider that Intel was still shipping 32-bit mobile chips at the time, but it would have been refreshing to see Microsoft push the market rather than follow.
3. PC vendors. What the hell are they doing? Why the hell are they shipping motherboards that can't handle PAE device remapping? Why the hell aren't they configuring PCs for PAE out-of-the-box? They've got tons of cheap memory, and can't do shit with it. They've known about the messy 32 to 64-bit transistion for a long time now, and have done nothing to help ease it.
The PC vendors can make most of the troubles of 32-bit Vista go away, but they don't seem all that driven to bridge the gap.
(since back in the days when the Reality Distortion Field convinced people, myself included, that they only needed 2 colors)
Some things never change. For instance, in Tiger you can have any desktop color scheme you like, so long as it's graphite or blue. About the only thing you can customize is the background image.
I've learned that they both have their flaws and any reasonable person is perfectly capable of hating them both equally
Exactly. For this specific issue, Windows does allow much more customization of the look and feel, but even that has its limits. If you want more, both Windows and OS X require $20 third-party programs to load custom themes. In other words, they BOTH suck.
But that's exactly what they did with DVI. The connector uses LVDS serial transmission lines much like other modern interfaces. The link design was made scalable (two links) for two reasons:
1. In 1999 literally NOBODY needed dual-link DVI. 2. In 1999, the cost (in terms of silicon) for the second controller was too high to justify.
They have slowly introduced the second link as the need grew, and now most add-in boards have at least one dual-link port.
Unfortunately, the current connector has no room for growth, but there's an extra facet people ignore about dual-link DVI: chipsets that support the dual link also support faster clock speeds than the original single-link spec of 165MHz, which means that dual-link DVI can potentially deliver MORE than twice the bandwidth of single-link DVI. This makes the improvements brought by DisplayPort even less impressive.
Yes, that's the long and short of what we all expected.
Home users don't know enough to care, and even most geeks don't care enough to hook them up with an OEM still selling XP. Most home users are either too scared to take a dive into OS X, or are freaked out when they see the price tag.
Businesses will stick with MS because there's really no other alternative (well, nothing that delivers the "whole package" like MS can). They will wait until service pack 1 or 2, just like they did with Windows 2000 and XP. This is "business as ususal" for them.
About the only place I've seen people pushing Macs over PCs this year have been college campus IT departments, because many of them are not yet ready to handle Vista. This is also a standard practice, because most of the staff is usually students, and they tend to lag current technology.
I am waiting patently to see the web trends for August and September to see if these college-level sales make a dent in the OS X versus Vista battle, but realistically I'm not expecting much.
Simple phone with bluetooth, good reception and decent battery life. Really cheap too. Offered by Sprint.
I have one and I love it. I was really wary about replacing my 5 year-old Touchpoint 1100, but after the hinge broke, I was forced to. I tried out the LX 150, and was really impressed with the responsiveness of the interface and the reception.
My biggest complaint with Fable was, being evil didn't really do anything. Sure, people hateed you and scowled in your general direction, but the quests themselves were virtually the same, so there were really no advantages/disadvantages to being a pure evil/good character.
I had fun for about two hours being evil to people, but after I realized it has no effect on the game at-large, I really didn't care anymore. If I'd wanted to just play a sandbox game, I'd have played The Sims.
Slashdot Mirror
I'm assuming your ancient system uses an AGP interface for graphics, which has a very fast download rate, but very poor upload. The maximum performance of AGP uploading data from the card memory to the rest of the machine is pretty slow (less than 100MB/sec, IIRC), and it will vary depending on the implementation. This is probably the reason you got such slow benches.
PCIe will likely give you performance more in-line with main memory (most implementations now are hitting 1-2 GB/s).
Exactly. The price of electronics is constantly dropping, so people are unwilling to spend more on newer devices; they expect that newer devices will cost the same, or possibly less. This is why you cannot bring inflation into a value consideration.
Let me clarify why you really can't use inflation as a direct comparison.
Today, people expect to pay the same or less for electronics they bought a decade or more ago. The difference today is, (1) they get more features for their dollar (convergence), and (2) thanks to the convergence and lack of price increase coupled with increased spending power, they can buy many times more electronic devices than they used to in the 80s and 90s.
Today, what 2-child family can live without multiple TVs, DVD players, home and portable gaming systems, computers, stereo systems, mp3 players, cell phones, etc. People have been trained by the industry to buy "more" rather than "better," because it is a far easier sell. Your "average" family today can afford many times the number of these devices they could 20 years ago, and so they play the game they have grown used to: buy more devices at similar prices. Unfortunately, this marketing method means it is harder than ever to upsell on quality and featureset, because people know they only have to wait a little bit for the same device with the same features to get cheaper.
Yes, in the past both compiler developers and chip designers have been known to optimize for certain aspects of benchmarks. Thus, to make numbers at all useful, you would have to have multiple benchmarks stressing different aspects of computing. This would prevent optimizing for benchmarks while hurting performance in other areas.
In addition to that, you have the problem that computer usage needs are constantly changing. This is the main reason why benchmarks change from year-to-year. How are you supposed to compare processor models when they advertise benchmark numbers from different versions of a given test suite?
In the end, marketing people from both AMD and Intel have discovered one ultimate truth: in a world where anyone and their dog can sell you an x86 computer, price is one of the easiest ways to gauge performance. People who care about performance will simply pay more and expect it. The rest of the world simply doesn't care, because they assume any processor is fast enough for their needs (and are probably right).
I don't think this is a pixel-pushing performance problem. I believe they're hitting the limits of 12MB EDRAM:
2 working framebuffers plus 1 complete output buffer plus one 32-bit Z-buffer make for fully-used EDRAM:
1152x640 pixels * 4 Bytes/pixel * 3 buffers + 1152x640 Z * 4 Bytes/Z = 11.25MB.
The same buffer configuration with full 720p resolution uses up 14MB of ram. Sure, you can move your framebuffer outside of the EDRAM, but you'll see a significant performance hit.
The dual-bus interface would also solve one of the problems of USB2 versus 1.1 backward compatibility, where a slower device on a shared bus will slow down high-speed devices. This is because bus transactions at the lower speed inherently take longer to complete, reducing the time slices available to the high-speed devices.
With USB3, you've got fibre for the high speed, and copper for the low speed.
Right, but there is one advantage to triple cores that Intel will be missing out on: A clearer and more complete lineup to fit various processing needs.
Right now, Intel sells the quad cores at a premium, and sells dual cores dirt-cheap. To the consumer who can't translate model numbers into actual clock speeds (most of the population), there is no clarity in the product lineup besides this division between two and four cores.
Really, would you be as willing to shell out extra money for a higher model number, if all the processors are already "Dual Core?"
AMD has a similar problem with their model numbering scheme, but now they can offer a clearer product lineup for stupid consumers: they can fill that wide gap between two and four cores with a triple-core option.
My video card is a Geforce 7900 GTX (512mb) and my CPU is an AMD Athlon X2 4600+. With that setup, would my RAM be holding me back?
No, not really. The difference in performance is typically %5 or less. Games are not usually memory-limited.
I mean, unless you can magically take a straight-line course between the two planets, how do you account for the off-trajectory component of the force exerted by the laser? The ship will be taking it's own course independent of the two planets, which means that the angle between the laser source and the ship will be constantly changing, and the laser will always be off-angle to the optimal.
Is there something I'm missing here? Is there an engine onboard to counteract the excess thrust? Are they planning on having multiple laser installations at different points in space to make this work? Am I missing a simple physics trick?
I guess you have a poor memory. How about a refresher?
I used my G400 MAX to play Counterstrike, Quake III, MDK2, etc. The 32-bit Direct3D performance was second to none. The OpenGL performance using the TurboGL was outstanding, and supported all OpenGL games in existence. I had a lot of faith in Matrox, until they casually ignored the powerful GeForce 256 and announced the G450. What a horrible memory.
Sooooo,it is analagous to a tiny mercury transmission line?
That's exactly what I was thinking. This sounds very much like delay line memory, except at the MEMS scale. Everything that is old is new again.
Delay line memory actually makes sense because you can pack so many more bits into it than a normal SRAM or Flash cell. Normal cells require interface logic for every bit (or every 2-4 bits on the most advanced Flash processes, thanks to multi-level cells). With this loop-based technology, you can place several times more bits at a time on the loop, and build only one controller to manipulate all those bits as they pass through.
What can I say? I'm disappointed that they stuck with a 3-issue architecture - while it is true that Intel's 4-issue setup is often data-starved, even with exceptional I/O performance AMD can only hope to match the Core platform in most situations. The lack of progress in their cache technology means AMD gets as much burden as benefit out of the L3 cache over 20ns access time!).
In the I/O arena, AMD potentially has the edge, and for HPC there's no question Barcelona will do well: this architecture is built for scaling beyond 8 cores (potentially as high as 8 sockets, 32 cores). The question is: how big is that market, and will it really pay-off for them?
In the meantime, small-scale servers and workstations are the majority of the market, and Intel still has a commanding lead in that space. Intel's introduction of the San Clemente Xeon chipset with DDR2 instead of FBDIMMs will not help things for AMD, because these systems can most certainly compete with AMD's low power consumption.
Go with "not til SP3 comes out."
You'll never have to use MS software again!
No.
Windows NT: final release SP6.
Windows 2000: final release SP4.
Windows XP: final release SP3 (to be released in 2008).
And much liks Vista, all three of the above operating systems had service pack releases within the first year. I'd say this puts Vista well on-track for three service packs.
That's the way it is: most of these Steam gamers are still CS junkies...I mean CS 1.6 plus CS:Source combined. If you want to know why people don't upgrade, you just have to look at the numbers Valve's survey doesn't tell you:
Have you ever seen PC online FPS game stats? CS 1.6 and CS:Source vie for the top title, and combined they hoard over %75 of all online FPS gamers.
You can play CS 1.6 on Intel GMA series graphics...hell, you can even run software mode if that's your bag. CS:Source runs decently even on a Radeon 9600 Pro or GeForce 6600, so there's no real reason to upgrade there. Why upgrade when the game you play every day is smooth?
Exactly. If I made $1 last year, and my friend made $1 million last year, we both "averaged" $500,000. Which is a figure not even close to representing what either of us actually made.
Your example fails because your sample size is too small. A sample size of "all men" would not have this problem when you apply an average.
Your example fails again because you assume that there is somehow a gap within the distribution of the size of all men. The distributed size of all men would be a fairly continuous dataset (translation: there are men of all shapes and sizes). This means the average is a very good tool with which to draw conclusions.
Why'd you buy them?
You don't notice it, initially, but eventually it grates on you.
Example: I own the Queens of the Stone Age: Songs for the Deaf album mentioned on the Wikipedia loudness war page. I bought the album because the songs sounded pretty nice. I think the songs are hard-hitting, and fun.
I have noticed one issue: I have NEVER been able to sit down and listen to the entire album in one sitting. This is something I can do with almost any album I own, even heavy metal albums (even previous albums from QotSA); but with Songs for the Deaf, even though the music is great, just the effort of listening really wears on me.
Now that I'm aware of the problem, I try to pay more attention.
Back to your point, my ears can't really make out the difference between a CD and an MP3 as long as the MP3 is encoded at a decent bitrate. I can clearly make out the difference between a 128kbps and 256kbps, but not between 256kbps and 320kbps or VBR.
I have to agree, it is hard to tell the difference these days. I've been a long-time proponent of VBR: r3mix was the first encoder setting besides 256 or 320k CBR where I couldn't tell the difference between the CD and the compressed file, even on my Allesandros. Today's VBR settings are far more impressive, with alt preset standard pushing the limits of audio quality with mp3.
True story: recently, for about 6 months I was accidentally encoding my alt preset standard mp3s with a peak bitrate locked at 224k. I encoded DOZENS of CDs without noticing. When I finally noticed, I re-encoded for consistency's sake, but I couldn't tell the difference. That's how good LAME is today: ABR of 160-192k is transparent, even without 320k peaks.
If it weren't for AMD forcing innovation down Intel's throat we would still be stuck with that crap they called the Pentium 4
And Rambus. And CPUs without x86-64.
Could you imagine the entire world switching to Itanium just to get 64-bit support? It would be a nightmare!
Yes, the quad-core chips will have the fourth link. In addition, the chips will be able to split their 16-bit HT links into dual 8-bit HT links, allowing for 8-way CPU configurations without hops (8 x 8-bit HT links per socket). In reality, this is the reason why AMD is pushing the new HyperTransport 3.0: so they can cut the bus lines to 8 without sacrificing too much bandwidth.
Check it out here.
Sure, it's nice to have single data sets which can occupy more than ~2^30-2^32 bytes (depending on alignment and bit robbing for tags) without having to window them, but such applications are rare -- most processes that use pointers will never come anywhere close to the limits of a 32-bit pointer.
No, not for the near future. But multiple applications + OS memory + hardware address mapping can quite easily use up the space of one 32-bit pointer, and that is the reason for my post. If Intel had made more of an effort to standardize the industry on 64-bit architecture, or alternatively made a stronger push to support PAE at the motherboard level (full 32-bit pointer space for every application), we wouldn't be having this conversation.
This makes requiring x86-64 more important than you imagine it is, because there's nothing optional in the standard (unlike PAE hardware address remapping and 36-bit address extension in IA32). If the OS requires it, then OEMs are forced to offer complete support, which means no more address space games.
The major weakness of IA32 is that the ISA has no native 64-bit datatype, so any manipulations of 64 bit datatypes require lots of extra instructions which are superfluous in an architecture where chip internals handle 64 bit words as quickly as 32 bit ones.
No, it's not really. You just stated that there's no need (for the moment) of 64-bit ints if you can break the 4GB barrier with clever paging, and virtual address spaces beyond 2^32...and because of this, once again I have to point back to the poor implementation of PAE as one of the biggest downfalls in modern systems.
The major advantage of x86-64 is therefore not in pointers, nor in integers, being 32-bits-only, because on other architectures LP64 tends to outperform ILP64 for most integer-heavy tasks, and 32 bit pointer heavy tasks tend to outperform the same tasks done with 64 bit pointers, for similar reasons.
The major advantage of x86-64 is that support for large memory spaces is not optional. In the industry, everyone is trying to shave a buck, and it's tempting to not include support for an "enterprise-level" feature in mainstream motherboards and chipsets. It's unfortunate, but because PAE started in the server world, much like PCI-X it will live and die in the server world.
This problem has many causes. I will now lay blame on the appropriate parties.
1. Intel. Yes, Intel, our savior. Intel did a very bad thing for the future of computing when they refused to add 64-bit support to the Pentium M line. They then continued the disservice by shipping the CoreDuo, continuing their legacy of 32-bit only for the notebook platform. Only now in 2007 are we seeing the last 32-bit only mobile chips come off the mainstream market (they're used for embedded devces only, where they belong).
2. Microsoft, for not having the balls to push the 64-bit Vista as the "standard" platform. The OEM versions of Vista are actually 32-bit only. You can understand this decision when you consider that Intel was still shipping 32-bit mobile chips at the time, but it would have been refreshing to see Microsoft push the market rather than follow.
3. PC vendors. What the hell are they doing? Why the hell are they shipping motherboards that can't handle PAE device remapping? Why the hell aren't they configuring PCs for PAE out-of-the-box? They've got tons of cheap memory, and can't do shit with it. They've known about the messy 32 to 64-bit transistion for a long time now, and have done nothing to help ease it.
The PC vendors can make most of the troubles of 32-bit Vista go away, but they don't seem all that driven to bridge the gap.
(since back in the days when the Reality Distortion Field convinced people, myself included, that they only needed 2 colors)
Some things never change. For instance, in Tiger you can have any desktop color scheme you like, so long as it's graphite or blue. About the only thing you can customize is the background image.
I've learned that they both have their flaws and any reasonable person is perfectly capable of hating them both equally
Exactly. For this specific issue, Windows does allow much more customization of the look and feel, but even that has its limits. If you want more, both Windows and OS X require $20 third-party programs to load custom themes. In other words, they BOTH suck.
But that's exactly what they did with DVI. The connector uses LVDS serial transmission lines much like other modern interfaces. The link design was made scalable (two links) for two reasons:
1. In 1999 literally NOBODY needed dual-link DVI.
2. In 1999, the cost (in terms of silicon) for the second controller was too high to justify.
They have slowly introduced the second link as the need grew, and now most add-in boards have at least one dual-link port.
Unfortunately, the current connector has no room for growth, but there's an extra facet people ignore about dual-link DVI: chipsets that support the dual link also support faster clock speeds than the original single-link spec of 165MHz, which means that dual-link DVI can potentially deliver MORE than twice the bandwidth of single-link DVI. This makes the improvements brought by DisplayPort even less impressive.
Yes, that's the long and short of what we all expected.
Home users don't know enough to care, and even most geeks don't care enough to hook them up with an OEM still selling XP. Most home users are either too scared to take a dive into OS X, or are freaked out when they see the price tag.
Businesses will stick with MS because there's really no other alternative (well, nothing that delivers the "whole package" like MS can). They will wait until service pack 1 or 2, just like they did with Windows 2000 and XP. This is "business as ususal" for them.
About the only place I've seen people pushing Macs over PCs this year have been college campus IT departments, because many of them are not yet ready to handle Vista. This is also a standard practice, because most of the staff is usually students, and they tend to lag current technology.
I am waiting patently to see the web trends for August and September to see if these college-level sales make a dent in the OS X versus Vista battle, but realistically I'm not expecting much.
Simple phone with bluetooth, good reception and decent battery life. Really cheap too. Offered by Sprint.
I have one and I love it. I was really wary about replacing my 5 year-old Touchpoint 1100, but after the hinge broke, I was forced to. I tried out the LX 150, and was really impressed with the responsiveness of the interface and the reception.
Check out the reviews here.
My biggest complaint with Fable was, being evil didn't really do anything. Sure, people hateed you and scowled in your general direction, but the quests themselves were virtually the same, so there were really no advantages/disadvantages to being a pure evil/good character.
I had fun for about two hours being evil to people, but after I realized it has no effect on the game at-large, I really didn't care anymore. If I'd wanted to just play a sandbox game, I'd have played The Sims.