The thing weighs 7 pounds (less than my 8200) and no, it probably won't last through a 2.5 hour DVD. My 8200 gets about 2 hours of battery life, tops. But, can you run C&C Generals and Solid Edge (3D CAD) on your Portege?
I own and Inspiron 8200 as well, and the thing has yet to blue-screen on me. Of course, this is due to the fact I run Linux, which has no blue screen. However, I used it for the first month of its life with Windows XP (it's a trauma it hasn't fully recovered from yet, no matter how much Gentoo I feed it!) and I never had a blue-screen. Ever. Maybe you should ditch the bundled drivers and D/L NVIDIA's from their website?
The 8200 really is a great machine. Fast, ugpradable, orgasmic screen, etc. However, it's also very flimsy, hot, heavy, and sucks battery life like anything. But most desktop replacements are like that:)
Actually, browsing porn sucks on a high res display. All the images are meant for 1024x768 displays, and appear too small. Scaling just makes things ugly.
I have one of those laptops. I use KDE, and I just set X to run at 133 DPI, which is the native DPI of the screen. Fonts are *extremely* crisp, and the same size as they would be on any other display. 12pt means 12/72 (1/6) of an inch, not 12 pixels. If you set things correctly (pass the -- -dpi 133 option to X, or change the DPI setting in Windows) pretty much everything should scale just fine, and you get a huge payoff in readibility.
While I certainly appreciate the flexibility here, I can't help but wonder if this will lead to some bad music listening habits. An album is a complete artistic whole. Even if you don't like particular songs, they can often be integral to the body of the work. Further, particular songs can easily grow on you as you listen to the complete album a few times. When I bought the latest Tori Amos CD (Scarlet's Walk) I didn't really like a couple of the songs. But I've listened to the CD more than a dozen times now, and I must say I now like many of the ones I didn't the first time through.
2 gigaBITS per channel, 32 *dual* channels. That's 128 gigabits/sec, which is 16 gigabytes per second. However, initial implementations will have up to 16 dual channels, in the case of the AGP replacement. And when I said that PCI-E will replace AGP, I meant that there would be no more seperate graphics bus like there is now. In order to do that, PCi-E will have to have at least as much bandwidth as AGP did.
PCI is (in theory) 133 MB/sec. PCI express is from 250MB/sec (2 channel) to 16GB/sec (32 channel). Also, PCI Express is designed to more than replace PCI, it's also designed to replace AGP. Lastly, there are several things bottlenecking on PCI these days. Gigabit ethernet (which Apple ships by default), HDDs that can burst > 80 MB/sec, RAID arrays, IEEE 1394b (400MB/sec!), etc.
Have you ever tried compiling something in Linux. Even on a 2GHz P4, KDE takes 8 hours to do a full build. Renders still take a couple of hours. Even some simple video encoding can take several minutes. Before making overly general comments like "there is no need for more speed from systems" try actually *doing* something with your system.
Also, there is no such thing as a $20 CPU. The only reason you can buy them on PriceWatch for $20 is because they're subsidized by sales of the $400 CPUs. And a PC with eight $20 CPUs would be dead slow. Most code is not that parallizable, and any interconnect that can handle 8 CPUs is going to be a whole lot more than a few hundred dollars.
If anything, I'd like to see an addon vector processor for high speed math. G4 motherboards have them under an Altivec core instruction set. >>>>>>>> It's not a component of the motherboard, but of the CPU. And Intel CPU's *do* have vector units, just like AltiVec. These units (SSE units) probably aren't as powerful as the G4's AltiVec units, but both are essentially limited by bus bandwidth, of which the P4 has 3x as much.
I would also want the ability to directly program (in chip asm) to do misc functions. >>>>>>> They do. It's called a "function call." If you're worried about the extra-overhead, don't be. Features like the Trace Cache in the P4 remove a whole lot of the decode overhead associated with x86 CPUs.
If anything, I'd want an architechure where EVERYTHING's on a very high speed, very high bandwidth quad-plane bus >>>>>> And I want to be the Queen of England.
with basic controllable logic. >>>>>>>>> No such thing as "basic" control logic anymore. Read the tome that is the Infiniband specs and weep.
To those who say this isnt possible, I believe the Altair 8800 used this similar architechure >>>>>> Life was easier back then. Today, we're hitting a whole bunch of limits, both engineering and physical. We can't engineer a bus that fast to those tolerences and have it work in existing environments. Not enough precision, too much environmental interference, etc.
Grognard. It's not 1992 anymore and PC's aren't lumbering beasts.
Thin clients: How are people going to use this at home? Over their 28.8 dialup connection? With the work I do, I can peg pretty much anything you throw at me. You think they want user's like me on shared systems? You think I want other users slowing me down?
PC architecture: A modern PC has more resources than most RISC workstations that are 5x the price. Ever since the P4 came out, PC memory bandwidth (one of it's traditional weak points) has skyrocketed. By the end of the year, it will be up at 6.4 GB/sec, which is an impressive number even for an SGI or Sun machine.
Bottlenecks: What do you do where the HDD is the bottleneck? After an hour of use or so, my Linux system pretty much runs out of RAM. On workstation tasks, the HDD is often not the benchmark. It's not the benchmark for the 3D rendering I do, the scientific sims, the gaming, the programming, pretty much everything. In fact, I thought it was going to really suck moving to a P4 laptop, because of the slow 4200 RPM hard drive. Ever since I put 640MB of RAM in there, I don't noticed any slowdown at all.
50% speedup means just that. If you've got a single CPU system, adding a second CPU can make things 50 percent faster. If a render takes 3 minutes, speeding things up by 50% will make the render take only 2 minutes. This is about in line with what is shown for 3D rendering. Games have less of a speed up, around 30% (for Quake, pretty much the only SMP game). In most cases, SMP has much less of a speedup. Specifically, it is widely acknolwedged that having 2 x GHz CPUs is nowhere near as good as 1 2x GHz CPU, because the base case speedup is the same, while the average case speedup is much lower (due to most code not having enough parallelism).
Unfortunately, the 2.5 + 2.5 > 4 is even more bullshit than the "the G4 can match a P4 even though the latter is more than twice the clockspeed!" SMP often doesn't get you much of an improvement at all. A 50% speedup is considered very good. What Apple needs to do is stop playing tricks, and ship machines that are just plain faster on benchmarks that people care about (3d rendering, compiling, games, etc). Then people will take notice.
The increased length of the pipeline is often not as important on floating point code. Often, in 3D graphics, for example, it's a matter of just streaming as many FP values through the vector units (hence the name streaming SIMD) as memory bandwidth allows.
Or at least, France knows it doesn't want to lose several billion in oil contracts. >>>>>>>>>>>> Right. So when the US goes to war just to support its interests (the Gulf War, for example) it's all right. When France does it, it's not?
*> Before anybody tries to reply, check out Boost's compiler status pages. Visual C++ 7.0 has 58 failures the EDG-based Intel C++ 7.0 has 43 failures, and GCC 3.2.1, has only 16!
Unfortunately, the big third candidate last election was from the "don't run with scissors" party. The "consumer's are so dumb they need to be protected from their own stupidity" party. The "let's do all sorts of communistic things so stupid people don't kill themselves" party. And I'm a freaking liberal!
Depends on what you're doing. If you're going to need a very fast > 4-way machine, then the Alpha's interconnection architecture (especially the EV7) will be much better. It's in this realm, large N-way machines, the heavy duty machines from Sun and IBM can really flex their muscles, even though their CPUs aren't usually as powerful individually. If you're looking at 4-way machine, then x86 will be faster, thanks to more memory bandwidth and high clock speeds. Right now, a 2.8 GHz P4 is about 25% faster than a 1.2 GHz Alpha in floating point code, and more than 50% faster in integer code. If you consider the cost differential between machines using the two CPUs, then x86 wins by a mile. Now, in the case of the Alpha, it's sheer dated-ness that's holding it back. Even with the huge differential in clock speed, the Alpha puts up quite a fight. I'd really like to see what it could do if as much effort was put into it as is being put into Itanium or Power4.
The original post argued that once people needed to go past 2GB of RAM, 32-bit processors would be insufficient. The person who replied to him said that the number was 4GB, because a 32-bit processor could address 4GB. I pointed out that it doesn't matter if the processor can address 4GB, because it needs some address space for other devices, which limits the total RAM to 2GB. The point isn't how much the processor can address, because we're talking about RAM limits here. And that RAM limit is (for practical purposes) 2GB.
p4 so they could juice up the multiplier and sell "faster"mhz cpu's at double the price is more than enuf for me to stop watching them. >>>>>>>>> Actually, the high multiplier is there because SSE code is extremely regular, and it is pretty easy to keep the system's SSE pipelines filled. The high multiplier (and fast memory bus) allows the P4 to just churn through FP code. Check out the P4's spec FP marks sometime.
I was thinking about this ( this thread). The biggest reason I can see for not doing this is access latency. Accessing a cache line requires a lookup operation that takes several clock cycles (7 on a P4's L1 cache). Register access is one clock cycle. If you have a tiny L0 cache, made up of register-like memory, you need to use fully associative memory to make lookups as fast as possible. Fully associative memory is expensive, and would probably be very hard to make run at high clock speeds.
What you're missing is that x86 chips have a ginormous amount of internal rename registers (128 in a P4). The bump to 16 *visible* registers in the Athlon-64 is to allow the compiler optimizer to give more information to the CPU about variable usage. I'm guessing that AMD found that more than 16 visible GPRs really didn't help the compiler's allocation routines any.
Actually, no. All of Intel's non-server chipsets top out at 2GB. The reason is that 4GB isn't possible without PAE thanks to stuff (BIOS, APIC, AGP memory, etc) that have to be mapped at the top end of the 4GB. Since 3GB is a rather weird size for main memory (hard to fill on dual-channel chipsets) 2GB is pretty much the reasonable max. PC2100 DDR (optimal for dual channel P4's) run about $50 for 512MB, or $200 for a GB. That's about what I paid for 64MB for my PII-300, and is quite a reasonable amount to budget for memory in a high end or even mid-range machine.
Slashdot Mirror
The thing weighs 7 pounds (less than my 8200) and no, it probably won't last through a 2.5 hour DVD. My 8200 gets about 2 hours of battery life, tops. But, can you run C&C Generals and Solid Edge (3D CAD) on your Portege?
I own and Inspiron 8200 as well, and the thing has yet to blue-screen on me. Of course, this is due to the fact I run Linux, which has no blue screen. However, I used it for the first month of its life with Windows XP (it's a trauma it hasn't fully recovered from yet, no matter how much Gentoo I feed it!) and I never had a blue-screen. Ever. Maybe you should ditch the bundled drivers and D/L NVIDIA's from their website?
:)
The 8200 really is a great machine. Fast, ugpradable, orgasmic screen, etc. However, it's also very flimsy, hot, heavy, and sucks battery life like anything. But most desktop replacements are like that
Actually, browsing porn sucks on a high res display. All the images are meant for 1024x768 displays, and appear too small. Scaling just makes things ugly.
I have one of those laptops. I use KDE, and I just set X to run at 133 DPI, which is the native DPI of the screen. Fonts are *extremely* crisp, and the same size as they would be on any other display. 12pt means 12/72 (1/6) of an inch, not 12 pixels. If you set things correctly (pass the -- -dpi 133 option to X, or change the DPI setting in Windows) pretty much everything should scale just fine, and you get a huge payoff in readibility.
While I certainly appreciate the flexibility here, I can't help but wonder if this will lead to some bad music listening habits. An album is a complete artistic whole. Even if you don't like particular songs, they can often be integral to the body of the work. Further, particular songs can easily grow on you as you listen to the complete album a few times. When I bought the latest Tori Amos CD (Scarlet's Walk) I didn't really like a couple of the songs. But I've listened to the CD more than a dozen times now, and I must say I now like many of the ones I didn't the first time through.
Well, in conjunction with OpenAL, SDL, and most recently OpenML, OpenGL definately does qualify as a multimedia platform.
Firewire 1394b includes the 3.2 gigabit/sec mode.
2 gigaBITS per channel, 32 *dual* channels. That's 128 gigabits/sec, which is 16 gigabytes per second. However, initial implementations will have up to 16 dual channels, in the case of the AGP replacement. And when I said that PCI-E will replace AGP, I meant that there would be no more seperate graphics bus like there is now. In order to do that, PCi-E will have to have at least as much bandwidth as AGP did.
PCI is (in theory) 133 MB/sec. PCI express is from 250MB/sec (2 channel) to 16GB/sec (32 channel). Also, PCI Express is designed to more than replace PCI, it's also designed to replace AGP. Lastly, there are several things bottlenecking on PCI these days. Gigabit ethernet (which Apple ships by default), HDDs that can burst > 80 MB/sec, RAID arrays, IEEE 1394b (400MB/sec!), etc.
Have you ever tried compiling something in Linux. Even on a 2GHz P4, KDE takes 8 hours to do a full build. Renders still take a couple of hours. Even some simple video encoding can take several minutes. Before making overly general comments like "there is no need for more speed from systems" try actually *doing* something with your system.
Also, there is no such thing as a $20 CPU. The only reason you can buy them on PriceWatch for $20 is because they're subsidized by sales of the $400 CPUs. And a PC with eight $20 CPUs would be dead slow. Most code is not that parallizable, and any interconnect that can handle 8 CPUs is going to be a whole lot more than a few hundred dollars.
If anything, I'd like to see an addon vector processor for high speed math. G4
motherboards have them under an Altivec core instruction set.
>>>>>>>>
It's not a component of the motherboard, but of the CPU. And Intel CPU's *do* have vector units, just like AltiVec. These units (SSE units) probably aren't as powerful as the G4's AltiVec units, but both are essentially limited by bus bandwidth, of which the P4 has 3x as much.
I would also want
the ability to directly program (in chip asm) to do misc functions.
>>>>>>>
They do. It's called a "function call." If you're worried about the extra-overhead, don't be. Features like the Trace Cache in the P4 remove a whole lot of the decode overhead associated with x86 CPUs.
If anything, I'd want
an architechure where EVERYTHING's on a very high speed, very high bandwidth
quad-plane bus
>>>>>>
And I want to be the Queen of England.
with basic controllable logic.
>>>>>>>>>
No such thing as "basic" control logic anymore. Read the tome that is the Infiniband specs and weep.
To those who say this isnt possible, I
believe the Altair 8800 used this similar architechure
>>>>>>
Life was easier back then. Today, we're hitting a whole bunch of limits, both engineering and physical. We can't engineer a bus that fast to those tolerences and have it work in existing environments. Not enough precision, too much environmental interference, etc.
Grognard. It's not 1992 anymore and PC's aren't lumbering beasts.
Thin clients: How are people going to use this at home? Over their 28.8 dialup connection? With the work I do, I can peg pretty much anything you throw at me. You think they want user's like me on shared systems? You think I want other users slowing me down?
PC architecture: A modern PC has more resources than most RISC workstations that are 5x the price. Ever since the P4 came out, PC memory bandwidth (one of it's traditional weak points) has skyrocketed. By the end of the year, it will be up at 6.4 GB/sec, which is an impressive number even for an SGI or Sun machine.
Bottlenecks: What do you do where the HDD is the bottleneck? After an hour of use or so, my Linux system pretty much runs out of RAM. On workstation tasks, the HDD is often not the benchmark. It's not the benchmark for the 3D rendering I do, the scientific sims, the gaming, the programming, pretty much everything. In fact, I thought it was going to really suck moving to a P4 laptop, because of the slow 4200 RPM hard drive. Ever since I put 640MB of RAM in there, I don't noticed any slowdown at all.
50% speedup means just that. If you've got a single CPU system, adding a second CPU can make things 50 percent faster. If a render takes 3 minutes, speeding things up by 50% will make the render take only 2 minutes. This is about in line with what is shown for 3D rendering. Games have less of a speed up, around 30% (for Quake, pretty much the only SMP game). In most cases, SMP has much less of a speedup. Specifically, it is widely acknolwedged that having 2 x GHz CPUs is nowhere near as good as 1 2x GHz CPU, because the base case speedup is the same, while the average case speedup is much lower (due to most code not having enough parallelism).
Unfortunately, the 2.5 + 2.5 > 4 is even more bullshit than the "the G4 can match a P4 even though the latter is more than twice the clockspeed!" SMP often doesn't get you much of an improvement at all. A 50% speedup is considered very good. What Apple needs to do is stop playing tricks, and ship machines that are just plain faster on benchmarks that people care about (3d rendering, compiling, games, etc). Then people will take notice.
The increased length of the pipeline is often not as important on floating point code. Often, in 3D graphics, for example, it's a matter of just streaming as many FP values through the vector units (hence the name streaming SIMD) as memory bandwidth allows.
Or at least, France knows it doesn't want to lose several billion in oil contracts.
>>>>>>>>>>>>
Right. So when the US goes to war just to support its interests (the Gulf War, for example) it's all right. When France does it, it's not?
A better compiler*, with source code, free. What more do you want?
*> Before anybody tries to reply, check out Boost's compiler status pages. Visual C++ 7.0 has 58 failures the EDG-based Intel C++ 7.0 has 43 failures, and GCC 3.2.1, has only 16!
Unfortunately, the big third candidate last election was from the "don't run with scissors" party. The "consumer's are so dumb they need to be protected from their own stupidity" party. The "let's do all sorts of communistic things so stupid people don't kill themselves" party. And I'm a freaking liberal!
Depends on what you're doing. If you're going to need a very fast > 4-way machine, then the Alpha's interconnection architecture (especially the EV7) will be much better. It's in this realm, large N-way machines, the heavy duty machines from Sun and IBM can really flex their muscles, even though their CPUs aren't usually as powerful individually. If you're looking at 4-way machine, then x86 will be faster, thanks to more memory bandwidth and high clock speeds. Right now, a 2.8 GHz P4 is about 25% faster than a 1.2 GHz Alpha in floating point code, and more than 50% faster in integer code. If you consider the cost differential between machines using the two CPUs, then x86 wins by a mile. Now, in the case of the Alpha, it's sheer dated-ness that's holding it back. Even with the huge differential in clock speed, the Alpha puts up quite a fight. I'd really like to see what it could do if as much effort was put into it as is being put into Itanium or Power4.
To quote the original poster:
"...but it's easily bypassed, and often MUST be to do normal operations."
I'm asking when the second part of the statement is every true. When "MUST" you bypass the type system to "do normal operations."
The original post argued that once people needed to go past 2GB of RAM, 32-bit processors would be insufficient. The person who replied to him said that the number was 4GB, because a 32-bit processor could address 4GB. I pointed out that it doesn't matter if the processor can address 4GB, because it needs some address space for other devices, which limits the total RAM to 2GB. The point isn't how much the processor can address, because we're talking about RAM limits here. And that RAM limit is (for practical purposes) 2GB.
p4 so they could juice up the multiplier and sell "faster"mhz cpu's at double the price is more than enuf for me to stop watching them.
>>>>>>>>>
Actually, the high multiplier is there because SSE code is extremely regular, and it is pretty easy to keep the system's SSE pipelines filled. The high multiplier (and fast memory bus) allows the P4 to just churn through FP code. Check out the P4's spec FP marks sometime.
I was thinking about this ( this thread). The biggest reason I can see for not doing this is access latency. Accessing a cache line requires a lookup operation that takes several clock cycles (7 on a P4's L1 cache). Register access is one clock cycle. If you have a tiny L0 cache, made up of register-like memory, you need to use fully associative memory to make lookups as fast as possible. Fully associative memory is expensive, and would probably be very hard to make run at high clock speeds.
What you're missing is that x86 chips have a ginormous amount of internal rename registers (128 in a P4). The bump to 16 *visible* registers in the Athlon-64 is to allow the compiler optimizer to give more information to the CPU about variable usage. I'm guessing that AMD found that more than 16 visible GPRs really didn't help the compiler's allocation routines any.
Actually, no. All of Intel's non-server chipsets top out at 2GB. The reason is that 4GB isn't possible without PAE thanks to stuff (BIOS, APIC, AGP memory, etc) that have to be mapped at the top end of the 4GB. Since 3GB is a rather weird size for main memory (hard to fill on dual-channel chipsets) 2GB is pretty much the reasonable max. PC2100 DDR (optimal for dual channel P4's) run about $50 for 512MB, or $200 for a GB. That's about what I paid for 64MB for my PII-300, and is quite a reasonable amount to budget for memory in a high end or even mid-range machine.