This architecture allows for some interesting optimizations not feasible in conventional CPUs.
"Crusoe's Code Morphing software not only keeps track of which blocks of code execute most often and optimizes them accordingly, but it also keeps track of which branches are most often taken and annotates the code accordingly. That way, Crusoe's branch prediction algorithm knows how likely a branch is to be taken, and which branch it should speculatively execute down. If a branch isn't particularly likely to go one way or the other, then Crusoe can speculatively execute down both branches.
Contrast this with speculative execution done on a normal CPU, where hardware limitations like buffer and table sizes limit the amount of information you can store about a particular branch and its execution history. Since Code Morphing keeps track of the branch histories in software, it can record a more finely grained description of the execution patterns of a wider window of code, and therefore assess more accurately whether or not a specific branch is likely to be taken."
High performance on the desktop is also interesting: "So you see, they made the Code Morphing software extremely modular. They can implement whatever parts of it they like in hardware to get whatever degree of performance gain they want. Crusoe should be viewed more as a proof of concept than as the ultimate outcome of 5 years of work. Crusoe represents one extreme of a spectrum that stretches from "implement the bare minimum in hardware" to "implement everything in hardware." Now that Transmeta has a technology that's proven to work in the most difficult case (where 2/3 of the transistor logic has been moved into software), they can go back in the other (easier) direction and start putting stuff in silicon.
Furthermore, since there's a software layer between the ISA of the binary and the machine's native ISA, Transmeta is free to beef up the execution engine (or any other part of the core) however they like, because the only thing that will require a recompile is the Code Morphing software. A case in point is the two chips in its product line. Each has a slightly different core (the Windows chip has special instructions in it that help speed up Windows), but they both are fully x86 compatible. There's nothing to keep them from stuffing new functions and features (SIMD anyone?) into the silicon, to help scale the product has high up as they want to go with it.
I'd say that it's only a matter of time before we hear an announcement of another product line from Transmeta. It won't be named Crusoe, because it won't be aimed at the mobile and embedded markets. It'll be a workstation and server class x86 CPU that runs Linux like a fiend, and it'll compete directly with Intel's IA-64. I can't wait."
I, for one, am really excited about the possibilities. -- My opinions may have changed, but not the fact that I am right.
The multitasking restriction is that IDE cannot issue more than 1 io request at a time, but this doesn't matter for single disk systems.
The only essential difference between a SCSI HD and an IDE one is the drive electronics, they are physically identical.
This last point brings up an interesting facet of this announcement: the only reason SCSI disks are more expensive is because all the manufacturers make both IDE and SCSI disks, so they market the SCSI as "power-user" and "enterprise". The only reason we don't have 10,000 RPM IDE drives today is marketing. -- My opinions may have changed, but not the fact that I am right.
Hmm... I know that optimizing for speed is best left until near the end of the project, but I doubt that optimizing for space is similar. The reason is because this is intimately tied to your data structures, and these are relatively hard to change at the last minute. OO design can help here, but some architectural decisions are hard to undo.
personally, I think that the mozilla engineers will succeed. -- My opinions may have changed, but not the fact that I am right.
Moving between the keyboard and the mouse is the slowest part of the operation. Of course, considering that you almost never need to use a keyboard when browsing, the point is moot. This argument is most applicable to editors, where you have text input, icons/menus (hopefully with keyboard accelerators), and complex multi-panel/tree view windows (e.g. netscape's preferences and explorer's "Internet options", respectively). To operate most efficiently, you need to use both input methods. This is where the input-type switching latency kills you. -- My opinions may have changed, but not the fact that I am right.
Just replying to myself to note that the reason why I hadn't read the paper yet was because I opened it in a new browser window. Isn't it ironic? Don't ya think? -- My opinions may have changed, but not the fact that I am right.
I haven't read the paper in question yet, but I just wanted to note that, AFAICR, one of the enhancements in mozilla is a re-write of the network layer to explicitly take into account downloading from multiple browser windows, and giving the active browser window more of the pipe. I don't know if this is in current builds or not. -- My opinions may have changed, but not the fact that I am right.
What you're saying is bogus. OK, I do agree with your first point that most (office and internet) apps aren't CPU-bound. However, you never mentioned price. How much do you want to pay for a computer? Is it worth spending an extra $200 for a 10% speed increase in some component that translates into a 2% speed increase in your application? Most consumers say no. -- My opinions may have changed, but not the fact that I am right.
Try 'best operating system', 'worst operating system', 'best search engine', 'worst search engine', 'worst computer hardware', 'biggest hot dog', and 'worst e-commerce partner'. Google rules. -- My opinions may have changed, but not the fact that I am right.
The portability between the various flavors of linux is source-level: you need to recompile to run on a different arch. Most programs that do things right and use size_t s instead of longs will compile on alpha with no difficulty.
Microsoft has a lot more difficulty, because: (1) They are committed to binary backwards compatibility, and (2) there are a bunch of types that are specified as being 32 bits. -- Man is most nearly himself when he achieves the seriousness of a child at play.
XML, for one. IE5 also supports webDAV. Unfortunately, webDAV probably won't be Netscape 5. -- Man is most nearly himself when he achieves the seriousness of a child at play.
Apparently there will be some sort of official beta release, now that mozilla is architecture-complete. I haven't been able to find any explicit information about this. Anyone have any info? I would like to know projected dates, projected features, and projected stability. -- Man is most nearly himself when he achieves the seriousness of a child at play.
The three really usable linux browsers are netscape/mozilla, lynx, and KFM. -- Man is most nearly himself when he achieves the seriousness of a child at play.
The term 'portal' was never clearly defined. The term first appeared when search engines were trying to differentiate themselves from their competitors. They added local weather displays, news, stock market tickers, etc. All these things were customizable (i.e. the weather in your area, or the stocks you own). These portals claim to provide everything that you want to know.
Basically, a portal is something that presents you with all the information you need/want. So it definitely makes sense to create a 'genetic information portal'. -- Man is most nearly himself when he achieves the seriousness of a child at play.
Language, yes. Programming language, no. XML = eXtensible Modeling Language. -- Man is most nearly himself when he achieves the seriousness of a child at play.
OH MY GOD! I have not updated this page in ages. Crazy stuff.
So many things have happened since the last time I wrote stuff:
My university is still packed with idiotic students. This time the striking students have decided what can be researched and what not, depending on whether it is helping "imperialists" or the "population of mexico". Someone described this as facist. The first time someone uses the word facist correctly.
I went to Boston and to Atlanta (for ALS). We got a secret investor firm to invest in our supper-dupper free software company to develop free software and provide kick-ass applications for users all around the globe.
So, I am quitting my current job and moving to the US to pursue this new venture with my friend Nat. Exciting times.
-- Man is most nearly himself when he achieves the seriousness of a child at play.
Slashdot Mirror
- This architecture allows for some interesting optimizations not feasible in conventional CPUs.
- High performance on the desktop is also interesting: "So you see, they made the Code Morphing software extremely modular. They can implement whatever parts of it they like in hardware to get whatever degree of performance gain they want. Crusoe should be viewed more as a proof of concept than as the ultimate outcome of 5 years of work. Crusoe represents one extreme of a spectrum that stretches from "implement the bare minimum in hardware" to "implement everything in hardware." Now that Transmeta has a technology that's proven to work in the most difficult case (where 2/3 of the transistor logic has been moved into software), they can go back in the other (easier) direction and start putting stuff in silicon.
I, for one, am really excited about the possibilities."Crusoe's Code Morphing software not only keeps track of which blocks of code execute most often and optimizes them accordingly, but it also keeps track of which branches are most often taken and annotates the code accordingly. That way, Crusoe's branch prediction algorithm knows how likely a branch is to be taken, and which branch it should speculatively execute down. If a branch isn't particularly likely to go one way or the other, then Crusoe can speculatively execute down both branches.
Contrast this with speculative execution done on a normal CPU, where hardware limitations like buffer and table sizes limit the amount of information you can store about a particular branch and its execution history. Since Code Morphing keeps track of the branch histories in software, it can record a more finely grained description of the execution patterns of a wider window of code, and therefore assess more accurately whether or not a specific branch is likely to be taken."
Furthermore, since there's a software layer between the ISA of the binary and the machine's native ISA, Transmeta is free to beef up the execution engine (or any other part of the core) however they like, because the only thing that will require a recompile is the Code Morphing software. A case in point is the two chips in its product line. Each has a slightly different core (the Windows chip has special instructions in it that help speed up Windows), but they both are fully x86 compatible. There's nothing to keep them from stuffing new functions and features (SIMD anyone?) into the silicon, to help scale the product has high up as they want to go with it.
I'd say that it's only a matter of time before we hear an announcement of another product line from Transmeta. It won't be named Crusoe, because it won't be aimed at the mobile and embedded markets. It'll be a workstation and server class x86 CPU that runs Linux like a fiend, and it'll compete directly with Intel's IA-64. I can't wait."
--
My opinions may have changed, but not the fact that I am right.
- UDMA cpu usage is comparable to SCSI.
- The multitasking restriction is that IDE cannot issue more than 1 io request at a time, but this doesn't matter for single disk systems.
- The only essential difference between a SCSI HD and an IDE one is the drive electronics, they are physically identical.
This last point brings up an interesting facet of this announcement: the only reason SCSI disks are more expensive is because all the manufacturers make both IDE and SCSI disks, so they market the SCSI as "power-user" and "enterprise". The only reason we don't have 10,000 RPM IDE drives today is marketing.--
My opinions may have changed, but not the fact that I am right.
personally, I think that the mozilla engineers will succeed.
--
My opinions may have changed, but not the fact that I am right.
Moving between the keyboard and the mouse is the slowest part of the operation. Of course, considering that you almost never need to use a keyboard when browsing, the point is moot. This argument is most applicable to editors, where you have text input, icons/menus (hopefully with keyboard accelerators), and complex multi-panel/tree view windows (e.g. netscape's preferences and explorer's "Internet options", respectively). To operate most efficiently, you need to use both input methods. This is where the input-type switching latency kills you.
--
My opinions may have changed, but not the fact that I am right.
Just replying to myself to note that the reason why I hadn't read the paper yet was because I opened it in a new browser window. Isn't it ironic? Don't ya think?
--
My opinions may have changed, but not the fact that I am right.
I haven't read the paper in question yet, but I just wanted to note that, AFAICR, one of the enhancements in mozilla is a re-write of the network layer to explicitly take into account downloading from multiple browser windows, and giving the active browser window more of the pipe. I don't know if this is in current builds or not.
--
My opinions may have changed, but not the fact that I am right.
I can vouch for them too.
--
My opinions may have changed, but not the fact that I am right.
So is Kip your real first name? It is in my case.
--
My opinions may have changed, but not the fact that I am right.
What you're saying is bogus. OK, I do agree with your first point that most (office and internet) apps aren't CPU-bound. However, you never mentioned price. How much do you want to pay for a computer? Is it worth spending an extra $200 for a 10% speed increase in some component that translates into a 2% speed increase in your application? Most consumers say no.
--
My opinions may have changed, but not the fact that I am right.
... when you get a drinking game made up for you. That link is hilarious!
--
My opinions may have changed, but not the fact that I am right.
Try 'best operating system', 'worst operating system', 'best search engine', 'worst search engine', 'worst computer hardware', 'biggest hot dog', and 'worst e-commerce partner'. Google rules.
--
My opinions may have changed, but not the fact that I am right.
.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
Microsoft has a lot more difficulty, because: (1) They are committed to binary backwards compatibility, and (2) there are a bunch of types that are specified as being 32 bits.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
Do you know the subject line of the thread?
--
Man is most nearly himself when he achieves the seriousness of a child at play.
.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
If the subject was "first post", then could you read it?
--
Man is most nearly himself when he achieves the seriousness of a child at play.
Time to crash.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
XML, for one. IE5 also supports webDAV. Unfortunately, webDAV probably won't be Netscape 5.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
This is easier:
bash$ factor 1999
1999: 1999
--
Man is most nearly himself when he achieves the seriousness of a child at play.
Apparently there will be some sort of official beta release, now that mozilla is architecture-complete. I haven't been able to find any explicit information about this. Anyone have any info? I would like to know projected dates, projected features, and projected stability.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
The three really usable linux browsers are netscape/mozilla, lynx, and KFM.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
Basically, a portal is something that presents you with all the information you need/want. So it definitely makes sense to create a 'genetic information portal'.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
Language, yes. Programming language, no.
XML = eXtensible Modeling Language.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
So many things have happened since the last time I wrote stuff:
So, I am quitting my current job and moving to the US to pursue this new venture with my friend Nat. Exciting times.
--
Man is most nearly himself when he achieves the seriousness of a child at play.
And I thought he liked programming in perl! For shame!
--
Man is most nearly himself when he achieves the seriousness of a child at play.