Most of the problems with AMD systems ARE due to motherboard/chipset or Power Supply. Ever since AMD started putting up their recommended parts list, I generally tell people to visit them when they're building AMD based systems. In fact, my latest system's about the only one I've built or consulted on where the motherboard wasn't on AMD's recommended list (Aopen AK75 with Kingston PC2700 ram).
I will say that I generally tell novice builders to be wary with Athlons because there's always a chance they can crush the die by incorrectly attaching their heatsinks. I'm hoping Hammer'll fix this with a nice nickel slug over the die, but beyond this, I've been using AMD processors since the 5x86 100mHz days and I've always had good luck in my machines and the ones I've built and helped build.
AMD bought out and now produces the Alchemy chips based on the Mips architecture for their embedded space alongside their Elan x86 embedded chips. I guess the situation's a little like DEC--Alpha and Arm, one for the desktop (workstation and servers too), and one for embedded apps.
IIRC, diamond's quite hard to grow in single crystal form (as how we use silicon).
The diode drop that you noted is a function of doping, IIRC, and not so much the intrinsic bandgap (keep in mind, intrinsic diamond and silicon don't conduct).
As for the physical properties, diamond's already the perfect substrate, if only we could grow it reliably in single crystal form. By the way, my semiconductor processing professor once mentioned that IC's made for the military are generally fabricated on sapphire--you get less desirable qualities then silicon, but you get great ruggedness.
Didn't he make the equivalent of releasing 5 to 7 albums with a record label by selling just ~350k copies of a compilation he did directly to his fans?
Sounds like a win to me, more so if they keep on coming back for more.
Diamonds are great for industry. Incredibly hard due to their bonding properties, they make great heatsinks (look up the thermal conductivity of a diamond, there is nothing better). However, karat for karat (am I spelling that right?!?), rubies are rarer then diamonds. De Beers is the ONLY reason why diamonds are worth what they are, otherwise, they're only useful for a bunch of researchers trying to do oddball stuff with a rather cool looking rock.
If you want to see what makes a company great, look at SAS (a business software company). It's privately owned (doesn't owe a damn thing to greedy daytrading stockholders), and management KNOWS that it's strength lies in it's employees (treated like GOD in my opinion--if they only hired EE's I'd drop my resume there in a second, hell I'm thinking of doing it NOW).
Ironically enough, MBAs and anybody in administration are generally the first ones fired.
When was the last time an MBA worked in design and development? This alone makes them VERY good targets for getting fired. For the company to survive you need a product people will buy, and frankly, all the MBA's in the world won't convince people to buy crap (because they can dig their own out of the toilet if they want it bad enough).
This is blatantly highlighted and made into a joke in the episode of Dilbert concerning the merger with Brainsuck. "The synergy will get you."
So this is why I can't get a job at the USPTO! They only hire new IDIOTS!!!!
I'd be relieved to know they didn't hire me because I'm overqualified, but apparently whoever's doing the hiring just doesn't get the kind of harm they're doing to the process.
Although I can picture a company paying off someone higher up to push a patent through if I turned down everything that crossed my desk the first time.
We don't have our dicks far enough our congress critter's butt to get that kind of thing passed.
Besides that. We don't need gun control, we need bullet control.
"I'd put a cap in your ass if I could afford it!!!"
Do you really know what 64-bit gives you? It doesn't give you performance. It gives you extra precision and larger addressable memory size.
Performance in this case is more a matter of how well IBM's compilers work and the inner workings of the instruction decode stages and execution stages work.
The G5 already exists, it's the PowerPC 8500 series. Motorola already has the spec. sheets up and it is nothing more then an embedded host controller for network equipment utilizing the RapidIO bus.
As an EE Professor once said to me a class in Microprocessor Organization, "If you need that many registers, you better clean up your code." Also, have you LOOKED at a classical RISC architecture, for instance, Mips (from Paterson and Hennesey (?)). They say 32, but it's not really 32, it's more like 8 permenant and 8 temporary, and all the rest are normally researved for other stuff. In truth, more registers just allows the compiler to be sloppy.
You have to remember that compilers (much like hardware synthesis code) is driven by a rule set, generally not half as complex as what you or I would place on deciding how to rearrange code to fit an algorithm. A prime example would be a corrolation between how an autorouter routes a PCB board vs how a compiler compiles a high level program. The more layers you add (or registers in this case) the BETTER job it does.
Pity it was only released in TV in japan from what I hear, so it'll never make video release here. On the upside, all the other Miyazaki fans I know have it on fansub, although I want to hunt for it on a digital format.
^_^ Gotta love the John Denver spin off into "Concrete Road."
Slashdot Mirror
Most of the problems with AMD systems ARE due to motherboard/chipset or Power Supply. Ever since AMD started putting up their recommended parts list, I generally tell people to visit them when they're building AMD based systems. In fact, my latest system's about the only one I've built or consulted on where the motherboard wasn't on AMD's recommended list (Aopen AK75 with Kingston PC2700 ram).
I will say that I generally tell novice builders to be wary with Athlons because there's always a chance they can crush the die by incorrectly attaching their heatsinks. I'm hoping Hammer'll fix this with a nice nickel slug over the die, but beyond this, I've been using AMD processors since the 5x86 100mHz days and I've always had good luck in my machines and the ones I've built and helped build.
AMD bought out and now produces the Alchemy chips based on the Mips architecture for their embedded space alongside their Elan x86 embedded chips. I guess the situation's a little like DEC--Alpha and Arm, one for the desktop (workstation and servers too), and one for embedded apps.
IIRC, diamond's quite hard to grow in single crystal form (as how we use silicon).
The diode drop that you noted is a function of doping, IIRC, and not so much the intrinsic bandgap (keep in mind, intrinsic diamond and silicon don't conduct).
As for the physical properties, diamond's already the perfect substrate, if only we could grow it reliably in single crystal form. By the way, my semiconductor processing professor once mentioned that IC's made for the military are generally fabricated on sapphire--you get less desirable qualities then silicon, but you get great ruggedness.
Didn't he make the equivalent of releasing 5 to 7 albums with a record label by selling just ~350k copies of a compilation he did directly to his fans?
Sounds like a win to me, more so if they keep on coming back for more.
Diamonds are great for industry. Incredibly hard due to their bonding properties, they make great heatsinks (look up the thermal conductivity of a diamond, there is nothing better). However, karat for karat (am I spelling that right?!?), rubies are rarer then diamonds. De Beers is the ONLY reason why diamonds are worth what they are, otherwise, they're only useful for a bunch of researchers trying to do oddball stuff with a rather cool looking rock.
If you want to see what makes a company great, look at SAS (a business software company). It's privately owned (doesn't owe a damn thing to greedy daytrading stockholders), and management KNOWS that it's strength lies in it's employees (treated like GOD in my opinion--if they only hired EE's I'd drop my resume there in a second, hell I'm thinking of doing it NOW).
Ironically enough, MBAs and anybody in administration are generally the first ones fired.
When was the last time an MBA worked in design and development? This alone makes them VERY good targets for getting fired. For the company to survive you need a product people will buy, and frankly, all the MBA's in the world won't convince people to buy crap (because they can dig their own out of the toilet if they want it bad enough).
This is blatantly highlighted and made into a joke in the episode of Dilbert concerning the merger with Brainsuck. "The synergy will get you."
How much does Pro/E run vs AutoCAD? Why is it that I expect AutoCAD to be cheaper???
Holy Cow!!!!
So this is why I can't get a job at the USPTO! They only hire new IDIOTS!!!!
I'd be relieved to know they didn't hire me because I'm overqualified, but apparently whoever's doing the hiring just doesn't get the kind of harm they're doing to the process.
Although I can picture a company paying off someone higher up to push a patent through if I turned down everything that crossed my desk the first time.
We don't have our dicks far enough our congress critter's butt to get that kind of thing passed. Besides that. We don't need gun control, we need bullet control. "I'd put a cap in your ass if I could afford it!!!"
Do you really know what 64-bit gives you? It doesn't give you performance. It gives you extra precision and larger addressable memory size. Performance in this case is more a matter of how well IBM's compilers work and the inner workings of the instruction decode stages and execution stages work.
The G5 already exists, it's the PowerPC 8500 series. Motorola already has the spec. sheets up and it is nothing more then an embedded host controller for network equipment utilizing the RapidIO bus.
As an EE Professor once said to me a class in Microprocessor Organization, "If you need that many registers, you better clean up your code." Also, have you LOOKED at a classical RISC architecture, for instance, Mips (from Paterson and Hennesey (?)). They say 32, but it's not really 32, it's more like 8 permenant and 8 temporary, and all the rest are normally researved for other stuff. In truth, more registers just allows the compiler to be sloppy. You have to remember that compilers (much like hardware synthesis code) is driven by a rule set, generally not half as complex as what you or I would place on deciding how to rearrange code to fit an algorithm. A prime example would be a corrolation between how an autorouter routes a PCB board vs how a compiler compiles a high level program. The more layers you add (or registers in this case) the BETTER job it does.
Pity it was only released in TV in japan from what I hear, so it'll never make video release here. On the upside, all the other Miyazaki fans I know have it on fansub, although I want to hunt for it on a digital format. ^_^ Gotta love the John Denver spin off into "Concrete Road."