But you know nothing of how MOSFET transistors work (not used in the LED itself, but in the power and control circuits driving the LED). They have two things that are affected by temperature: on-state resistance, and threshold voltage.
On-state resistance is proportional to increases in temperature. This means as temperature increases, the power dissipated by the power/control circuits will be higher.
In addition, as temperature rises, the current drawn by the MOSFET increases because the threshold voltage decreases. "On average this [threshold voltage] variation is between 4 mV/C and 2 mV/C depending on doping level," which means you can see a significant current increase with a device designed for 50C running at 100C, since current is proportional to the inverse of the threshold voltage.
When you combine these two properties, you get the danger of,a href="http://www.smpstech.com/tmos0000.htm">MOSFET thermal runaway, especially in a cheap device stuck in a light socket with no active cooling Normally you can implement special control circuits to remove this dependency, or add aggressive cooling, but not at the price points LED bulbs have to meet and the environments the have to work in.
You can spend just $60 on a Radeon HD 4670. This affordable gem is more powerful than the GPU of the PS3, and should give you similar gaming performance.
Want a little better performance? You can spend just $100 on a nice 9800GT. That's twice as powerful as the GPU in the PS3, and will support rendering resolutions the PS3 can only dream of.
Want a little better performance? Then step-up to a GTS 250 or 4850 at $110-120! That's the beauty of PC gaming: you can buy as little or as much performance as you desire.
Yes, it pisses me off that DX9 is the development target precisely because that's the feature level of the two most powerful consoles. It's funny you mention Borderlands, because the game is very fun to play, but the graphics are annoying.
1. They took the same graphics on the 360 version, and just added a whole bunch of pitch-black dynamic shadows for the PC (I think they don't realize that you can assign DYNAMIC RANGE for your shadow intensity). This makes the game way too dark, and since their gamma control tops-out at "dark alley" level, you can't see a thing unless you have a really bright monitor (and don't mind the game being washed-out). And since the braindead devs didn't have the foresight to include a FLASHLIGHT, I had to dig through the ini file and turn these shadows off just to play through some of the indoor sequences. No way this kinda shit would have shipped if this were a PC title first.
2. Unreal Engine 3 uses deferred rendering for some of the effects, and under DX9 this means that you can't have MSAA enabled. Epic Games acknowledged their stupidity, but also provided a carrot by promising that DX10 would allow AA to work in UE3 games. The lazy developers of Borderlands realized that actually supporting DX10 would require re-coding, so instead they released the game as pure DX9, with a hacked-on, completely unsupported DX10 mode buried in the ini file. Needless to say, neither mode allowed for anti-aliasing of every object on the screen.
These developers don't seem to realize that they are STALLING graphics development. Building a high-quality next-generation 3D accelerator is no easy task (just ask Intel, who has spent the last four years developing Larrabee), and for the past ten years the PC has traditionally been the source/seed for every new console 3D chipset. That's virtually FREE development work done for the console makers. But it's not happening much anymore, now that the PC is gaming's second-class citizen.
Hell, with LAME, VBR 256 is overkill. Preset Standard (target 180-220 ABR) has been good enough for years - it will automatically adjust bitrate upwards for tough tracks.
About the only thing it will fail at is tracks with continuous complexity, since it cannot steal bits from simpler parts of the song to fully-describe the more complex portions. But I have yet to encounter a track with the aforementioned description (in my mind, it would have to have lots of random noise and little variations in dynamics, and who makes a song like that?)
This may sound cold, but we'd probably be better off if this happened to more people. If everyone went around getting CP-downloading malware by mistake, I think they'd eventually give up on prosecuting people for it.
Many kids (not all) don't know how to properly weigh the consequences of their actions, and so they can be easily manipulated into doing just about anything. All it takes is one kid's testimony (zero hard evidence) to ruin your life forever (who cares if you're innocent or guilty, you'll still spend years jumping through all the legal hoops and dealing with the mob hysterics who hide in the shadows and sick authorities on the accused via "anonymous tips.")
If you think that more incidences of CP-downloading malware will somehow clarify the silly situation and stop frivolous prosecutions, you're a fool. The reason we have such knee-jerk laws in this country is because the mob loves a good burning at the stake. And since they can't do that anymore, they do the next best thing and rally around a law like a bunch of bloodthirsty jackals. If you are simply accused (not convicted), these people will sit in the shadows while dragging your name through the mud and trying you in the court of opinion. And if you have the balls to question the laws, they will fight you tooth and nail, and claim you love child molesters and child porn.
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech
Congress shall make no law. But there is no law in-question here.
On CNN's website, they ARE the law (when it comes to censoring posted speech). You agreed to the terms in order to use their service. The censoring happens only after you read the EULA, click the box and press OK, and it happens entirely on their servers (not public property).
If you hate the policies so much, just make your own news discussion site (blog). Or if you're lazy like me, you could just submit it to Slashdot and wait two days:)
Also, since the bastards didn't include the sequential read speed graphs for the USB2/3 drives, we have no idea what the CPU utilization is.
We all knew USB3 would be faster. In fact, at the top-end it's probably more than the 5-6x performance increase they're reporting, since they're obviously drive-limited. What we don't know is, is it more efficient than the processor-heavy USB2?
Right, and all the claims from people about how we already use flash in millions of iPods - stop. SSD is a totally different application from an iPod, despite using the same flash memory:
1. The controllers used are cutting-edge, high-performance clocked, and thus more prone to defects as time goes by.
2. The workload presented to an SSD is completely different from that of your average mp3 player: the write cycles SSDs are subjected to are astronomical. This is bound to wear them out faster.
I think we're going to see over the next few years that the rates of failure are high, and the industry will take that data and re-evaluate reliabilutyt metrics to something more realistic.
The board features Gigabit Ethernet, and if you need 2 NICs there is a PCI slot and a MiniPCIe. Performance should be plenty for anything server-related (2 SATA), and if you need more drives you can add them via USB.
Absolutely. Intel looked like they were going to start producing "decent" integrated graphics when, in 2004, they announced the GMA 900. It looked to Microsoft like the world's largest GPU maker would finally have something capable of desktop compositing, so they figured they could finally add this capability to Windows without a huge performance hit.
Then, in 2006 Intel announced the GMA X3000, but couldn't produce drivers to enable the advanced features like Vertex Shaders (this took eighteen months). In the end, the perfrormance sucked, and most OEMs passed-over the capable G965 for the craptacular 915G and 945G. So, in early 2007 Vista launched, and Microsoft was screwed because Intel hadn't delivered acceptable 3D performance in time, and had to put "Vista-Capable" logos on 915G and 945G machines that were still shipping.
The whole Netbook debacle hit Microsoft like a ton of bricks becaused Intel tried to segment the market, and used predatory pricing bundles to prevent OEMs from making netbooks with dual-core Atom or 3rd-party chipsets (e.g. pricing the Atom N270 + 945G chipset less than the Atom CPU itself). This meant that ALL Netbooks were incapable of running Vista, not just the low-end ones, because the only way to make a profitable Netbook was to follow the herd and take the Intel deal.
Now that Intel is finally upping the spec on their new Atom dual-core netbooks (end of this year), and now that Windows 7 has been optimized to the point that it's "usable" even on low-end Netbooks, I think Windows "performance" is poised for a comeback.
After reading this story on Ars (Slashdot is always days late), I tried out the free account system. Three things that really made me happy:
1. They never asked for a credit card. I just created an account like I was signing up for a forum. 2. The download was quick and painless (maxed-out my 25Mbit connection, was playing in 30 minutes). 3. I felt so free that I didn't even feel bad about not getting the chance to play it last night.
I've been itching to try one of these MMOs, but couldn't stomach the monthly fee (I don't often have lots of time to play, and I also tend to put a game down after a couple months, then pick it back up later). For me, a monthly fee would be wasted. I like this pricing structure because I won't be forced to pay for anything, but if I really like the game I could see myself making small purchases here and there. If I find I really like this, and get worried about spending too much, I still have the option to upgrade to the VIP account for the normal $15/month.
I could certainly see some improvement with a larger, higher-resolution screen. The screen on the DS is so small that you can't make-out small details. This really killed my enjoyment of Professor Layton - style games, because the details needed to solve puzzles are often hidden in a single pixel.
To be honest, I really haven't been that satisfied with any of these "fun" games. My most recent disappointment: Scribblenauts. The game is captivating for about 2 hours, but then you get tired of it: the physics engine is clunky, and I'm constantly trying to anticipate what direction my objects or player character will go shooting off to when I touch the screen. The word engine is nice, initially - but it gets old fast because a lot of the objects are extremely limited in what you can do with them (e.g. you can only dig holes DOWN with a steam shovel. Apparently, digging through a wall is impossible. I was also disappointed when I couldn't entice a mole to dig through the ground, nor could I figure out how to get the drill to DRILL through anything. Also, when I tried to use the physics engine to my advantage, and try launching my player across the screen, I was constantly disappointed because he never got far off the ground).
So, you find what works, and continue to use the same objects, because the puzzles are all basically the same. And meanwhile, you get no bonus points because you've reached the edge of your creativity.
I've been disappointed with the DS game library overall. The only fun I've had is with remakes like Dragon Quest IV and Chrono Trigger - most of the "new" games are crapware or half-assed playtested releases with serious issues. If the games can at least look good, that would give me a little more immersive experience, despite the issues these games come with.
For once I can play a 3D game on a Nintendo portable and not be limited to 2000 polygons in a scene, and not be limited to shitty point-sample for texture filtering?
Hurray, we can finally have REAL 3D games on the DS, instead of crappy-looking 2D/3D hybrids!
The power consumption is also quite reasonable (Zune HD can do 8.5 hours video playback, and that stresses the GPU core and OLED screen), so it sounds plausible for Nintendo to sign-on.
Right, Google News is the equivalent of the front page on a traditional newspaper: tons of excerpts "continued on page XX," and almost no advertising.
Google beats the other newspapers by offering more varied content on their front page than any single newspaper can. If Rupert wants to compete, he needs to create his own news portal, instead of pushing for content segregation and paid access.
When I first read this, I was thinking of HARP, the (rather obvious) precursor to the SHARP program. His goal of making HARP a space launch platform was a failure, but the lead engineer (Gerald Bull) was so disgusted with the politics, he went on to created Project Babylon for Iraq. I suppose the moral of the story is: keep the big gun makers employed, or they will go work for someone else:)
Back to the original topic: from the press release, they've doubled the velocity achieved by HARP. If that is true, then it's only a small hop with a booster rocket to LEO. This could really work!
Thanks Charlie. You don't have to say any more about ATI, because the cat's already out of the bag (some site broke the Tuesday NDA). They'll be moving exclusively to GDDR5 on 128-bit bus for their midrange parts. This means that right now, they could sell a cheap 512MB 5850 with 4 memory chips for next to nothing. And once the 2Gbit GDDR5 parts ship next year, those 1GB 5770 parts can be paired with just 4 memory devices, and could probably be sold for the same cheapo $100.
The power of a 4890 (almost) for around $100 six months from now? It's certainly possible, and it's just amazing what GDDR5 brings to the table!
Sounds to me like the venerable GDDR3 is finally headed for that big tech dump in the sky. It only took five years!
It's quite simple. Nvidia has graphics chips that cost $x to create. If they can't sell them for more than that due to competition with another company, then what is the point of creating said chips? Building something and selling it for a loss is a losing strategy. It's better to regroup and try again with something that makes you money.
What Nvidia needs are cheap powerful chips. What they have are expensive not-so-powerful chips that soon no one will want to buy. It's quite simple, their engineering dept. needs completely new leadership or they soon will be an ex-company.
I guess you're right. GT200b is a lot less expensive to make than GT200, but the build cost of the board has never come down. The 512-bit bus means the PCBs still require more layers, and board makers must include 16 memory chips in every build.
Since the high-end PC gaming market has partially stalled, there's no demand for a 2GB single GPU card at the top-end, so Nvidia can't even take advantage of increasing GDDR3 memory densities. So those GTX285 boards are still shipping with paltry 512Mbit chips.
That's great. Nvidia is outselling ATI chipsets by dumping stock of their Nforce4 (that is what the MCP61 is, you'd know these things if you read the PCPer article linked in the summary), a chipset from 2006 that doesn't even support PCIe 2.0. If that's not a sign of things to come, I don't know what is.
And Nvidia is developing ONE new chipset - ION2, for Apple. Since the rest of the world is moving-on to mobile i7/i5/i3, and even Atom is getting on-die graphics, I can't forsee Nvidia really investing anything in future chipset tech.
Charlie - you claim that Nvidia will be dropping their midrange graphics chipsets, but offer no explanation why. While I tend to agree with your insight, I can't see why Nvidia would be willing to give-up marketshare just to staunch the bleeding a little. I mean, what the hell else does Nvidia make money off of, aside from midrange graphics (Tegra? too early to tell. Chipsets? They're gone. HPC? Small market.)? It would be foolish to allow their one remaining profitable enterprise to languish.
But I have to believe you're right, because Nvidia shows none of the normal signs of competing. Normally when ATI releases something better, and Nvidia wants to compete, they introduce some impressive price drops, and we're not seeing that this time around. Is Nvidia really going to do something as stupid as sacrifice marketshare just to save a few dollars?
Awesome, now we can have the miniaturized nucleics of The Foundation! And we can sell them to backwater countries and manipulate them with our fake techno-religion!
No, seriously, reliable nuclear batteries could be less expensive than electrical infrastructure in many countries. But they need to be cheap, or else people will have a whole new excuse to fight wars.
I bought an HL-1240 in 2001 for $300 ($100 less than HP's cheapest laser at the time). The printer (and the stock toner cartridge) have lasted me through 2009 without any issues (I print occasionally, and it's nice to never have toner dry-out). Over the years, Brother supplied XP and Vista drivers (despite the fact that none of these OSes were out when I bought the thing), and good CUPS support meant it worked well on OS X and Linux.
I only had to buy a replacement recently because the toner cart got damaged, and I had to choose between a new cart ($50 or more) or a new printer ($120). I decided to see how much the technology had improved in 8 years, so I bought the Brother HL-2170W. On XP and OS X, the wireless configuration was a breeze, and it has worked without a hitch. The Linux support for the wireless is more involved - there is a CUPS driver, but you'll have to configure the wireless manually.
The new printer is even faster than my old one, and because it's wireless, I can stick it in whatever damn room I please. And the networking already supports IPv6, so I can depend on this network printer being future-proof.
It does. A minute number of high school athletes make professional careers out of it and are successful, the rest end up working in construction or a similar career and spending the rest of their life thinking about how great high school was.
Why mod this up "informative," when it's bullshit?
Sports as a career is just like math as a career: only the truly exceptional will ever see that big paycheck. Decent minor-league players (AA of AAA) and most decent math majors will make a respectable paycheck, and people who are not very good at baseball are just as screwed as people who are not very good at math (they will both work construction).
The fact is, just like people have physical limitations, most people have mental limitations. If you convince people to only concentrate on math, you'll be amazed how many hit their limit, having trouble visualizing ever-more-complex concepts. In a field which is just as-competitive as Baseball, if you have limits in math, you're not going to stay employed.
Why the hell do you think most people stay away from math? It's because we're asking the math heads to to more and more incredible things, and every year the bar runs higher.
Kidding aside, thanks for the rundown. It now makes sense.:) One thing I don't understand, though, is the write cycle life. Does the phase change substance gradually settle into a third state? Or does the heating mechanism 'wear out'?
According to this detailed paper, it's being conservatively limited to a million writes for now because they have no real experience from which to determine true write lifetime. Right now, they don't know which will fail first: the phase-change material, or the BJT driver transistor.
ExpressCard v2 is already in the pipeline, and will feature a single lane of PCIe v2.0 (5.0 Gbit) and USB 3.0. Really, it's no surprise USB 3.0 has a similar speed to PCIe 2.0, it's based on the exact same signaling tech (including full-duplex support).
NES v. SMS - doing just a quick comparison via youtube, I'd say the NES had better sound but the SMS has better graphics ability - almost like a Genesis in quality (look at Sonic 1 on the SMS). Plus the SMS' CPU runs about twice as fast. The NES was a 1982 technology and SMS used more-advanced 1985 tech, which you can see on the screen.
And if you compared one of the "best" games on the SMS to one of the "best" games on the NES, you'd end-up tied. I'd suggest you compare SMB3 to Sonic before you decide. SMB3 uses the PCM channel on the NES to produce exceptional drum tracks and effects. The large color palette and advanced scrolling effects (split screen, scroll in both directions at-once) are unmatched by any other game on the NES.
I will fully admit that the SMS had a slight hardware lead, but very few games made use of the hardware edge (Phantasy Star and Sonic come to mind). In real terms, there was nothing exceptional about either system, except that one beat the other to market, and locked-in the game publishers.
Slashdot Mirror
But you know nothing of how MOSFET transistors work (not used in the LED itself, but in the power and control circuits driving the LED). They have two things that are affected by temperature: on-state resistance, and threshold voltage.
On-state resistance is proportional to increases in temperature. This means as temperature increases, the power dissipated by the power/control circuits will be higher.
In addition, as temperature rises, the current drawn by the MOSFET increases because the threshold voltage decreases. "On average this [threshold voltage] variation is between 4 mV/C and 2 mV/C depending on doping level," which means you can see a significant current increase with a device designed for 50C running at 100C, since current is proportional to the inverse of the threshold voltage.
When you combine these two properties, you get the danger of ,a href="http://www.smpstech.com/tmos0000.htm">MOSFET thermal runaway, especially in a cheap device stuck in a light socket with no active cooling Normally you can implement special control circuits to remove this dependency, or add aggressive cooling, but not at the price points LED bulbs have to meet and the environments the have to work in.
And you don't have to.
You can spend just $60 on a Radeon HD 4670. This affordable gem is more powerful than the GPU of the PS3, and should give you similar gaming performance.
Want a little better performance? You can spend just $100 on a nice 9800GT. That's twice as powerful as the GPU in the PS3, and will support rendering resolutions the PS3 can only dream of.
Want a little better performance? Then step-up to a GTS 250 or 4850 at $110-120! That's the beauty of PC gaming: you can buy as little or as much performance as you desire.
Yes, it pisses me off that DX9 is the development target precisely because that's the feature level of the two most powerful consoles. It's funny you mention Borderlands, because the game is very fun to play, but the graphics are annoying.
1. They took the same graphics on the 360 version, and just added a whole bunch of pitch-black dynamic shadows for the PC (I think they don't realize that you can assign DYNAMIC RANGE for your shadow intensity). This makes the game way too dark, and since their gamma control tops-out at "dark alley" level, you can't see a thing unless you have a really bright monitor (and don't mind the game being washed-out). And since the braindead devs didn't have the foresight to include a FLASHLIGHT, I had to dig through the ini file and turn these shadows off just to play through some of the indoor sequences. No way this kinda shit would have shipped if this were a PC title first.
2. Unreal Engine 3 uses deferred rendering for some of the effects, and under DX9 this means that you can't have MSAA enabled. Epic Games acknowledged their stupidity, but also provided a carrot by promising that DX10 would allow AA to work in UE3 games. The lazy developers of Borderlands realized that actually supporting DX10 would require re-coding, so instead they released the game as pure DX9, with a hacked-on, completely unsupported DX10 mode buried in the ini file. Needless to say, neither mode allowed for anti-aliasing of every object on the screen.
These developers don't seem to realize that they are STALLING graphics development. Building a high-quality next-generation 3D accelerator is no easy task (just ask Intel, who has spent the last four years developing Larrabee), and for the past ten years the PC has traditionally been the source/seed for every new console 3D chipset. That's virtually FREE development work done for the console makers. But it's not happening much anymore, now that the PC is gaming's second-class citizen.
Hell, with LAME, VBR 256 is overkill. Preset Standard (target 180-220 ABR) has been good enough for years - it will automatically adjust bitrate upwards for tough tracks.
About the only thing it will fail at is tracks with continuous complexity, since it cannot steal bits from simpler parts of the song to fully-describe the more complex portions. But I have yet to encounter a track with the aforementioned description (in my mind, it would have to have lots of random noise and little variations in dynamics, and who makes a song like that?)
This may sound cold, but we'd probably be better off if this happened to more people. If everyone went around getting CP-downloading malware by mistake, I think they'd eventually give up on prosecuting people for it.
Yeah, because we totally stopped prosecuting people for molestation without evidence because kids lie about it all the time. I mean, it's not like children have ever been coerced into trumping-up child molestation accusations, and they certainly don't lie about molestation to get attention and revenge.
Many kids (not all) don't know how to properly weigh the consequences of their actions, and so they can be easily manipulated into doing just about anything. All it takes is one kid's testimony (zero hard evidence) to ruin your life forever (who cares if you're innocent or guilty, you'll still spend years jumping through all the legal hoops and dealing with the mob hysterics who hide in the shadows and sick authorities on the accused via "anonymous tips.")
If you think that more incidences of CP-downloading malware will somehow clarify the silly situation and stop frivolous prosecutions, you're a fool. The reason we have such knee-jerk laws in this country is because the mob loves a good burning at the stake. And since they can't do that anymore, they do the next best thing and rally around a law like a bunch of bloodthirsty jackals. If you are simply accused (not convicted), these people will sit in the shadows while dragging your name through the mud and trying you in the court of opinion. And if you have the balls to question the laws, they will fight you tooth and nail, and claim you love child molesters and child porn.
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech
Congress shall make no law. But there is no law in-question here.
On CNN's website, they ARE the law (when it comes to censoring posted speech). You agreed to the terms in order to use their service. The censoring happens only after you read the EULA, click the box and press OK, and it happens entirely on their servers (not public property).
If you hate the policies so much, just make your own news discussion site (blog). Or if you're lazy like me, you could just submit it to Slashdot and wait two days :)
I agree.
Also, since the bastards didn't include the sequential read speed graphs for the USB2/3 drives, we have no idea what the CPU utilization is.
We all knew USB3 would be faster. In fact, at the top-end it's probably more than the 5-6x performance increase they're reporting, since they're obviously drive-limited. What we don't know is, is it more efficient than the processor-heavy USB2?
Right, and all the claims from people about how we already use flash in millions of iPods - stop. SSD is a totally different application from an iPod, despite using the same flash memory:
1. The controllers used are cutting-edge, high-performance clocked, and thus more prone to defects as time goes by.
2. The workload presented to an SSD is completely different from that of your average mp3 player: the write cycles SSDs are subjected to are astronomical. This is bound to wear them out faster.
I think we're going to see over the next few years that the rates of failure are high, and the industry will take that data and re-evaluate reliabilutyt metrics to something more realistic.
I just saw this reviewed on SPCR: Intel finally released a MiniITX motherboard with an Atom processor and and a mobile 945 chipset! Even with a 2.5" hard drive, idle power is less than 10w, and peak power is around 16w.
The board features Gigabit Ethernet, and if you need 2 NICs there is a PCI slot and a MiniPCIe. Performance should be plenty for anything server-related (2 SATA), and if you need more drives you can add them via USB.
Absolutely. Intel looked like they were going to start producing "decent" integrated graphics when, in 2004, they announced the GMA 900. It looked to Microsoft like the world's largest GPU maker would finally have something capable of desktop compositing, so they figured they could finally add this capability to Windows without a huge performance hit.
Then, in 2006 Intel announced the GMA X3000, but couldn't produce drivers to enable the advanced features like Vertex Shaders (this took eighteen months). In the end, the perfrormance sucked, and most OEMs passed-over the capable G965 for the craptacular 915G and 945G. So, in early 2007 Vista launched, and Microsoft was screwed because Intel hadn't delivered acceptable 3D performance in time, and had to put "Vista-Capable" logos on 915G and 945G machines that were still shipping.
The whole Netbook debacle hit Microsoft like a ton of bricks becaused Intel tried to segment the market, and used predatory pricing bundles to prevent OEMs from making netbooks with dual-core Atom or 3rd-party chipsets (e.g. pricing the Atom N270 + 945G chipset less than the Atom CPU itself). This meant that ALL Netbooks were incapable of running Vista, not just the low-end ones, because the only way to make a profitable Netbook was to follow the herd and take the Intel deal.
Now that Intel is finally upping the spec on their new Atom dual-core netbooks (end of this year), and now that Windows 7 has been optimized to the point that it's "usable" even on low-end Netbooks, I think Windows "performance" is poised for a comeback.
After reading this story on Ars (Slashdot is always days late), I tried out the free account system. Three things that really made me happy:
1. They never asked for a credit card. I just created an account like I was signing up for a forum.
2. The download was quick and painless (maxed-out my 25Mbit connection, was playing in 30 minutes).
3. I felt so free that I didn't even feel bad about not getting the chance to play it last night.
I've been itching to try one of these MMOs, but couldn't stomach the monthly fee (I don't often have lots of time to play, and I also tend to put a game down after a couple months, then pick it back up later). For me, a monthly fee would be wasted. I like this pricing structure because I won't be forced to pay for anything, but if I really like the game I could see myself making small purchases here and there. If I find I really like this, and get worried about spending too much, I still have the option to upgrade to the VIP account for the normal $15/month.
I could certainly see some improvement with a larger, higher-resolution screen. The screen on the DS is so small that you can't make-out small details. This really killed my enjoyment of Professor Layton - style games, because the details needed to solve puzzles are often hidden in a single pixel.
To be honest, I really haven't been that satisfied with any of these "fun" games. My most recent disappointment: Scribblenauts. The game is captivating for about 2 hours, but then you get tired of it: the physics engine is clunky, and I'm constantly trying to anticipate what direction my objects or player character will go shooting off to when I touch the screen. The word engine is nice, initially - but it gets old fast because a lot of the objects are extremely limited in what you can do with them (e.g. you can only dig holes DOWN with a steam shovel. Apparently, digging through a wall is impossible. I was also disappointed when I couldn't entice a mole to dig through the ground, nor could I figure out how to get the drill to DRILL through anything. Also, when I tried to use the physics engine to my advantage, and try launching my player across the screen, I was constantly disappointed because he never got far off the ground).
So, you find what works, and continue to use the same objects, because the puzzles are all basically the same. And meanwhile, you get no bonus points because you've reached the edge of your creativity.
I've been disappointed with the DS game library overall. The only fun I've had is with remakes like Dragon Quest IV and Chrono Trigger - most of the "new" games are crapware or half-assed playtested releases with serious issues. If the games can at least look good, that would give me a little more immersive experience, despite the issues these games come with.
For once I can play a 3D game on a Nintendo portable and not be limited to 2000 polygons in a scene, and not be limited to shitty point-sample for texture filtering?
Hurray, we can finally have REAL 3D games on the DS, instead of crappy-looking 2D/3D hybrids!
The power consumption is also quite reasonable (Zune HD can do 8.5 hours video playback, and that stresses the GPU core and OLED screen), so it sounds plausible for Nintendo to sign-on.
Right, Google News is the equivalent of the front page on a traditional newspaper: tons of excerpts "continued on page XX," and almost no advertising.
Google beats the other newspapers by offering more varied content on their front page than any single newspaper can. If Rupert wants to compete, he needs to create his own news portal, instead of pushing for content segregation and paid access.
When I first read this, I was thinking of HARP, the (rather obvious) precursor to the SHARP program. His goal of making HARP a space launch platform was a failure, but the lead engineer (Gerald Bull) was so disgusted with the politics, he went on to created Project Babylon for Iraq. I suppose the moral of the story is: keep the big gun makers employed, or they will go work for someone else :)
Back to the original topic: from the press release, they've doubled the velocity achieved by HARP. If that is true, then it's only a small hop with a booster rocket to LEO. This could really work!
Thanks Charlie. You don't have to say any more about ATI, because the cat's already out of the bag (some site broke the Tuesday NDA). They'll be moving exclusively to GDDR5 on 128-bit bus for their midrange parts. This means that right now, they could sell a cheap 512MB 5850 with 4 memory chips for next to nothing. And once the 2Gbit GDDR5 parts ship next year, those 1GB 5770 parts can be paired with just 4 memory devices, and could probably be sold for the same cheapo $100.
The power of a 4890 (almost) for around $100 six months from now? It's certainly possible, and it's just amazing what GDDR5 brings to the table!
Sounds to me like the venerable GDDR3 is finally headed for that big tech dump in the sky. It only took five years!
It's quite simple. Nvidia has graphics chips that cost $x to create. If they can't sell them for more than that due to competition with another company, then what is the point of creating said chips? Building something and selling it for a loss is a losing strategy. It's better to regroup and try again with something that makes you money.
What Nvidia needs are cheap powerful chips. What they have are expensive not-so-powerful chips that soon no one will want to buy. It's quite simple, their engineering dept. needs completely new leadership or they soon will be an ex-company.
I guess you're right. GT200b is a lot less expensive to make than GT200, but the build cost of the board has never come down. The 512-bit bus means the PCBs still require more layers, and board makers must include 16 memory chips in every build.
Since the high-end PC gaming market has partially stalled, there's no demand for a 2GB single GPU card at the top-end, so Nvidia can't even take advantage of increasing GDDR3 memory densities. So those GTX285 boards are still shipping with paltry 512Mbit chips.
That's great. Nvidia is outselling ATI chipsets by dumping stock of their Nforce4 (that is what the MCP61 is, you'd know these things if you read the PCPer article linked in the summary), a chipset from 2006 that doesn't even support PCIe 2.0. If that's not a sign of things to come, I don't know what is.
And Nvidia is developing ONE new chipset - ION2, for Apple. Since the rest of the world is moving-on to mobile i7/i5/i3, and even Atom is getting on-die graphics, I can't forsee Nvidia really investing anything in future chipset tech.
Charlie - you claim that Nvidia will be dropping their midrange graphics chipsets, but offer no explanation why. While I tend to agree with your insight, I can't see why Nvidia would be willing to give-up marketshare just to staunch the bleeding a little. I mean, what the hell else does Nvidia make money off of, aside from midrange graphics (Tegra? too early to tell. Chipsets? They're gone. HPC? Small market.)? It would be foolish to allow their one remaining profitable enterprise to languish.
But I have to believe you're right, because Nvidia shows none of the normal signs of competing. Normally when ATI releases something better, and Nvidia wants to compete, they introduce some impressive price drops, and we're not seeing that this time around. Is Nvidia really going to do something as stupid as sacrifice marketshare just to save a few dollars?
Awesome, now we can have the miniaturized nucleics of The Foundation! And we can sell them to backwater countries and manipulate them with our fake techno-religion!
No, seriously, reliable nuclear batteries could be less expensive than electrical infrastructure in many countries. But they need to be cheap, or else people will have a whole new excuse to fight wars.
I bought an HL-1240 in 2001 for $300 ($100 less than HP's cheapest laser at the time). The printer (and the stock toner cartridge) have lasted me through 2009 without any issues (I print occasionally, and it's nice to never have toner dry-out). Over the years, Brother supplied XP and Vista drivers (despite the fact that none of these OSes were out when I bought the thing), and good CUPS support meant it worked well on OS X and Linux.
I only had to buy a replacement recently because the toner cart got damaged, and I had to choose between a new cart ($50 or more) or a new printer ($120). I decided to see how much the technology had improved in 8 years, so I bought the Brother HL-2170W. On XP and OS X, the wireless configuration was a breeze, and it has worked without a hitch. The Linux support for the wireless is more involved - there is a CUPS driver, but you'll have to configure the wireless manually.
The new printer is even faster than my old one, and because it's wireless, I can stick it in whatever damn room I please. And the networking already supports IPv6, so I can depend on this network printer being future-proof.
It does. A minute number of high school athletes make professional careers out of it and are successful, the rest end up working in construction or a similar career and spending the rest of their life thinking about how great high school was.
Why mod this up "informative," when it's bullshit?
Sports as a career is just like math as a career: only the truly exceptional will ever see that big paycheck. Decent minor-league players (AA of AAA) and most decent math majors will make a respectable paycheck, and people who are not very good at baseball are just as screwed as people who are not very good at math (they will both work construction).
The fact is, just like people have physical limitations, most people have mental limitations. If you convince people to only concentrate on math, you'll be amazed how many hit their limit, having trouble visualizing ever-more-complex concepts. In a field which is just as-competitive as Baseball, if you have limits in math, you're not going to stay employed.
Why the hell do you think most people stay away from math? It's because we're asking the math heads to to more and more incredible things, and every year the bar runs higher.
Kidding aside, thanks for the rundown. It now makes sense. :) One thing I don't understand, though, is the write cycle life. Does the phase change substance gradually settle into a third state? Or does the heating mechanism 'wear out'?
According to this detailed paper, it's being conservatively limited to a million writes for now because they have no real experience from which to determine true write lifetime. Right now, they don't know which will fail first: the phase-change material, or the BJT driver transistor.
ExpressCard v2 is already in the pipeline, and will feature a single lane of PCIe v2.0 (5.0 Gbit) and USB 3.0. Really, it's no surprise USB 3.0 has a similar speed to PCIe 2.0, it's based on the exact same signaling tech (including full-duplex support).
NES v. SMS - doing just a quick comparison via youtube, I'd say the NES had better sound but the SMS has better graphics ability - almost like a Genesis in quality (look at Sonic 1 on the SMS). Plus the SMS' CPU runs about twice as fast. The NES was a 1982 technology and SMS used more-advanced 1985 tech, which you can see on the screen.
And if you compared one of the "best" games on the SMS to one of the "best" games on the NES, you'd end-up tied. I'd suggest you compare SMB3 to Sonic before you decide. SMB3 uses the PCM channel on the NES to produce exceptional drum tracks and effects. The large color palette and advanced scrolling effects (split screen, scroll in both directions at-once) are unmatched by any other game on the NES.
I will fully admit that the SMS had a slight hardware lead, but very few games made use of the hardware edge (Phantasy Star and Sonic come to mind). In real terms, there was nothing exceptional about either system, except that one beat the other to market, and locked-in the game publishers.