What's Inside the Mars Rovers

Captain Zion writes "Space.com has a story about the hardware and software of Mars Rovers Spirit and Opportunity. Basically, they're radiation-shielded, 20MHz PowerPC machines wirh 128Mb RAM and 256Mb of flash memory, running VxWorks. I wonder if I could make a nice firewall with one of these for my home network..."

Comments

What's the bus speed on that thing?

[}InFuZeD{]

Does a 20mhz processor really need 128mb of ram? I mean, with a bus speed that low, you can probably put the data to flash ROM just as fast. What are the chances of you using all 128mb of ram?

Re:What's the bus speed on that thing?

[gerf]

Does a 20mhz processor really need 128mb of ram? I mean, with a bus speed that low, you can probably put the data to flash ROM just as fast. What are the chances of you using all 128mb of ram?

I imagine they can use all the storage they can, since there's no hard drive. So, the RAM acts as a cache for everything that is transmitted (which is a lot, actually). The Flash is used for more permanant software, like OS, commands, other files, ect. I'm amazed they can do it all with as little as they have.

Re:What's the bus speed on that thing?

[4r0g]

The memory is probably partitioned, and some banks can be shut down if they fail. Thus the 128MB. OR, you could use it as texture memory (wonder what the display adapter is like ;)

Re:What's the bus speed on that thing?

[Ianoo]

It's a damned good job they do have the RAM, since there wouldn't be much hope for Spirit if they only had the Flash.

Re:What's the bus speed on that thing?

[CaptainAlbert]

> Does a 20mhz processor really need 128mb of ram?

A processor of any speed doesn't need RAM of any size.

The application you want to run needs both processing power and memory. How much of each? Depends on the application.

Re:What's the bus speed on that thing?

[Gordonjcp]

Flash has a limited number of write cycles. Eventually the flash will "wear out", and become unreliable. Now, imagine if you're banging on a patch of memory repeatedly (like, scratchpad RAM)...

Re:What's the bus speed on that thing?

[GoofyBoy]

>What are the chances of you using all 128mb of ram?

Ask Bill Gates.

Re:What's the bus speed on that thing?

[snake_dad]

Man.. I wish I had 128mb of ram in my 1mhz CBM-64!

Re:What's the bus speed on that thing?

[hemp]

But its that special "wirh " RAM...very hard to get these days, except on eBay. :-)

Re:What's the bus speed on that thing?

[pi eater]

Who knows.. but you have to assume that they equipped this thing with a lot more RAM that they could ever forsee using

geeky stuff.. offensive stuff!

Re:What's the bus speed on that thing?

[Anonymous Coward]

True, but what supprise memore: omly 256 Mb of Flash? given the fact that you can always switch them/use them partialy (no additional power consumption), they do not take much space and mass and the fact that data transfer periods/speed is limited, I would expect flash in the range of at least gigabytes. And one reason of Spirit problems is constant overloading. Oh well, Nasa is too poor and can not affort a few extra flash modules..

Re:What's the bus speed on that thing?

[adrianbaugh]

Completely off-topic, but you reminded me of something: given that most modern CPUs have a meg or so of cache built into the chip, would it be possible to build a machine with no plug-in memory modules and have it boot something simple like DOS? Or would the motherboard complain that you were being silly?

Re:What's the bus speed on that thing?

[Anonymous Coward]

Can't help myself......how about a Beowulf cluster of MERs....

Re:What's the bus speed on that thing?

[Anonymous Coward]

Writing to flash takes a long time. The difference between RAM and flash will be very noticable, even for a 20MHz processor. For example, a write to the Fujitsu "high-speed" flash devices takes 12.6uS (per 16-bit word), according to Fujitsu's specs. For a 20 MIPs processor, 1 instruction is 50 nS, so modern RAM can "keep up" with it. In this situation, writing to flash would be more than 250 times slower than RAM, even with very fast flash.

Disclaimer: Even faster flash devices may exist, but I don't know about them.

Re:What's the bus speed on that thing?

[jandrese]

Generally, no. The cache maintains a _copy_ of recently accessed main memory. Logically it is not a seperate memory space, rather it shows up as certain memory locations being faster than others when you access them. The upshot is that you still need at least as much main memory as cache space to actually use the chace.

Re:What's the bus speed on that thing?

[snake_dad]

More specifically, they could only get it to stop rebooting itself continuously by telling it not to mount the flash memory. And from that moment on it seems to be working correctly. That was said one of the briefings, IIRC.

Re:What's the bus speed on that thing?

[brain159]

Flash "disk" controllers take care of wear-levelling automatically; even though you're writing to the same logical block number on the disk, it's actually not the same spot in the flash every time.

Re:What's the bus speed on that thing?

[CaptainAlbert]

Well, I'm no PC expert so don't take my word for it...

If it's possible for the cache controller to be configured to "direct map" the cache into the CPU's memory space, then I'd say the answer was yes, although you might indeed get some raised eyebrows from your mobo.

Trouble is, I don't think PC processors are ever expected to operate like that. (I've worked with embedded processors whose on-board memory can be configured as RAM, I-cache, D-cache or some combination of the three, but it's another world.)

I can imagine some kind of hardware hack which fools the processor into thinking that there is memory behind the cache. Providing that it's operating in write-back mode, and you configure everything just right and don't try accessing outside a particular range of addresses, then you might just be able to use your cache as if it were just RAM. (Very fast RAM, at that!)

Could be a fun project if you don't mind getting your hands dirty...

Re:What's the bus speed on that thing?

[3waygeek]

In other words, the Spirit is willing, but the flash is weak.

Re:What's the bus speed on that thing?

[iamdrscience]

Using Flash ROM as RAM is a baaaaaad idea. There actually is a limit to how many times it can be rewritten, you'd just never get to it with any normal use of Flash (i.e. as storage space), but it's definitely within reach when you use it as RAM (or as swap space). In fact, you'll wear it out pretty quickly (depends on the system design and activity and such, but I would guess maybe a few weeks for this one).

Re:What's the bus speed on that thing?

[deadlinegrunt]

"I'm amazed they can do it all with as little as they have."

All the more reason you can't help but laugh when people bring up arguments like "with computers/RAM/storage becoming so cheap and abundant who cares if {insert some bloated, interpreted, garbage collected language here} - just update your hardware."

One size fits all never does...

Re:What's the bus speed on that thing?

[secolactico]

Thank you, man. You just made my day with that pun. +5 Funny!

Re:What's the bus speed on that thing?

[t0ny]

I wonder if I could make a nice firewall with one of these for my home network..."

Sure, if you want it to keep rebooting in an unrecoverable loop.

Re:What's the bus speed on that thing?

[Anonymous Coward]

Yes. Flash requires a lot of power to write to but none to remember things. RAM requires a small amount of power to write to - but constant power to remember things. Both have there uses.

Re:What's the bus speed on that thing?

[addaon]

However, many processors (the G3/750FX comes to mind) allow you to address cache as physical memory, allowing exactly what the grandparent mentioned. You'll still need an external memory controller, though, to get stuff from ROM for boot purposes.

Re:What's the bus speed on that thing?

[Anonymous Coward]

You're forgetting DMA.

There are a lot of scientific instruments, Cameras and communication rigs/antennaes on the rover, and given the speed of the processor, I'm willing to bet all those devices just write directly to RAM, and the processor just coordinates. 128MB of ram for high resolution uncompressed pictures is not that much, not to mention all the scientific readings they are taking.

Re:What's the bus speed on that thing?

[Anonymous Coward]

awesome. you just got friended.

Re:What's the bus speed on that thing?

[Anonymous Coward]

did you send that one to rec.humor.funny ?

Re:What's the bus speed on that thing?

[WindBourne]

Not only slow, but very power hungry. On each of these machines, power is a very big issue. The more that the computers, the less there is available for other experiments.

Re:What's the bus speed on that thing?

[JonathanBoyd]

I think you completely missed his point - it wasn't about the amount of RAM, but rather about the speed of it. Not that I agree with him either.

Re:What's the bus speed on that thing?

[sxpert]

I don't think they have ANY need for a display adapter on a mars-borne rover that sits 300M kilometers away from the closest human...

Re:What's the bus speed on that thing?

[JDBrechtel]

Only problem with your statement is it doesn't apply to the markets where {insert some bloated, interpreted, garbage collected language here} is used. Updating your hardware there to save 3 months of development time IS a viable solution.

Re:What's the bus speed on that thing?

[Endive4Ever]

That's true, if you're using Flash memory through a 'disk controller' interface where it's emulating disk memory. No such mechanism is designed into Flash when it's built into the memory map, unless your OS or application does it for you.

Re:What's the bus speed on that thing?

[Anonymous Coward]

"I imagine they can use all the storage they can..."

duh, uhh...

Re:What's the bus speed on that thing?

[Endive4Ever]

Does Bill Gates really know anything bloated memory hogs like KDE and Gnome?

Re:What's the bus speed on that thing?

[Anonymous Coward]

For an embedded device such as the rover, 20MHz is not neccessarily slow. I would think one of the biggest concerns is power consumption which is directly proportional to the speed of the processor. Using only Flash instead of RAM is a bad idea, and would simply not work. There is no comparison with the two. With flash you have an erase cycle, which takes much time plus you can only erase a sector at a time which is usually between 8kB and 64kB. It would end up being many many times slower and would have to worry about wear-leveling. RAM is essential, and hopefully they are using SRAM, if not, DRAM with parity enabled.

Re:What's the bus speed on that thing?

[Anonym1ty]

In other words, the Spirit is willing, but the flash is weak.

I guess I'll have to add that to the book of Martian proverbs.

Re:What's the bus speed on that thing?

[Anonymous Coward]

No, no, no. The correct phrase is:

The Spirit is willing but the flesh is soft and spongy.

Re:What's the bus speed on that thing?

[LiquidCoooled]

Microsoft got away with that for years with Windows 98.
Why shouldnt anybody else be allowed to try ;)

Re:What's the bus speed on that thing?

[Anonymous Coward]

Of course, what we'll find out next is that Spirit crashed because of a memory leak that would have been avoided if they had used a garbage-collected language instead of insisting on managing every little detail themselves with C.

Re:What's the bus speed on that thing?

[Anonymous Coward]

wow, thats a joke we dont hear on Slashdot like 50000000000 times per day!

Its mildly amusing, but it kind of ceases being funny after the 10^50 time.

Re:What's the bus speed on that thing?

[wcrowe]

I did a spit-take on that one. Great pun.

Re:What's the bus speed on that thing?

[sharkey]

Does a 20mhz processor really need 128mb of ram?

128 millibits is awfully small. Even my Commodore had 64 kilobytes of memory.

Re:What's the bus speed on that thing?

[LWATCDR]

I am still amazed that people think that a 20Mhz Risc chip with 128 megs of flash is so little. Look at Voyager. It used a tiny little 8 bit chip.

Re:What's the bus speed on that thing?

[Guppy06]

"Does a 20mhz processor really need 128mb of ram?"

That depends on how much of the RAM is still usable after years of exposure to hard radiation.

Re:What's the bus speed on that thing?

[kcelery]

Cache in French means hidden. If you can physically address it, then it shouldn't be called a cache.

Re:What's the bus speed on that thing?

[SEWilco]

Yes, it is possible.

A processor chip could be created which has onboard RAM, using the same technology as onboard cache.

Indeed, there are such chips. I don't know the maximum amount of RAM presently available. You might find them by searching for "System On Chip", perhaps along with the name of an embedded operating system.

Such system on chip devices are at least able to be used on tiny boards. With RAM and peripheral interfaces built in, they often need only connection to proper sockets.

Re:What's the bus speed on that thing?

[Anonymous Coward]

Whoosh!

Re:What's the bus speed on that thing?

[Urkki]

Well, presumably, if it's being addressed directly, it's not being used as cache. And if it's being used as cache for external memory, it's not addressed directly.

Re:What's the bus speed on that thing?

[leandrod]

> some bloated, interpreted, garbage collected language

Garbage collection ain't bloat at all.

I hope the flash memory was not commodity hardware

[uid100]

How redundant is this setup? With the failure of the flash, it begs the question of "what if the CPU flakes out" and such.

Wrong!!!

[toupsie]

Does it support Ogg?

Radiation hardness

[swordboy]

Does anyone know what the deal was with the flash memory that caused the outage? I heard something about a "solar event" that caused a problem with the flash memory that led to the outage. It was subsequently resolved by disabling the flash. If so, BAE Aerospace has a possible solution with their upcoming line of rad-hard memory.

Re:Radiation hardness

[stevesliva]

It sounds more like it was actually just poor filesystem management, not a hardware failure or soft error.

There have been some incriminating statements along the lines of never having tested the FS for long periods, having to delete extra data from the flash, etc.

Re:Radiation hardness

[Anonymous Coward]

If the memory was ECC and I sure hope it was then a few twiddled bits wont matter.

Re:Radiation hardness

[TwistedGreen]

It appears to be a software error, and not hardware-related. It actually looks like it ran out of swap space and the OS crashed. This article explains what they think happened, and this article has more information on their recovery plans.

A quote:

It is now believed that the rover's flash memory had become so full of files that the craft couldn't manage all of the information stored aboard. Spirit bogged down because it didn't have enough random access memory, or RAM, to handle the current amount of files in the flash -- including data recorded during its cruise from Earth to Mars and the 18 days of operations on the red planet's surface.

Raises some interesting questions about software reliability, I think. Did nobody think about running out of disk space?

Re:Radiation hardness

[britt]

The problem is that they ran out of inodes. Filesystem freaked out and started to cause panics.

Re:Radiation hardness

[noselasd]

Latest theoriy is it had to many files in the flash memory. To many files can pressure the ram rather high depending on the OS and filesystem. I'd also take it an embedded OS such as vxWorks isn't optimized for _lotsoffiles_.
Some more here

Re:Radiation hardness

[Anonymous Coward]

The problem is that they ran out of inodes. Filesystem freaked out and started to cause panics.

Should've used reiserfs. ;-)

Re:Radiation hardness

[parc]

They didn't run out of "disk space". They ran out of RAM. My guess is that when reading the "directory" structure of the flash, a data structure needed to be allocated in RAM. When the malloc for that structure failed, it counted as a hard failure and the system rebooted. Presumably, a malloc failing is a symptom of much larger problems in your system, problems that you won't be able to overcome, so the best thing is to wipe everything and start fresh.

Re:Radiation hardness

[Anonymous Coward]

MOD PARENT UP
(I can't believe the mods have missed this!)

Re:Radiation hardness

[The_K4]

As side effect this means that Opertunity probably has the SAME bug! :(

Re:Radiation hardness

[BladeRider]

Raises some interesting questions about software reliability, I think. Did nobody think about running out of disk space?

I don't think this is an issue of software reliability, it is an issue of system analysis and design. If the system was not designed properly, to handle the amount of data needed to be stored, or to have proper housekeeping - the fault lies with the system analyst not the software.

Re:Radiation hardness

[cdrudge]

Which is why as soon as they were communicating with it they started deleting files immediately.

Re:Radiation hardness

[ehiris]

Whatever happened to load testing?
Seriously, it seems like they treat these robots like they are made out of glass.
I'm sure you could have had monster garage techs build something more reliable than these "scientific instruments".

Re:Radiation hardness

[Anonymous Coward]

Jesus christ, learn to spell.

Re:Radiation hardness

[TwistedGreen]

That makes more sense. I would have thought that the memory would be swapped out, eventually filling any free flash memory until there's nothing left. Though I admittedly am not too familiar with how VxWorks works.

Re:Radiation hardness

[Anonymous Coward]

Did nobody think about running out of disk space?

why? 128MB should be more than anybody would ever need...

ducks then runs

Re:Radiation hardness

[Tablizer]

Whatever happened to load testing?

I read somewhere that the "excuse" given was that the deadline was tight WRT launch window, so that they had to cut back on a lot of integration testing that they otherwise wish they could do.

Congress and the presidents tend to be fickle as a group in their funding and program selection, resulting in deadlines that otherwise might not be so rushed. IMO, congress is micromanaging NASA's robot probes too much based on short-term political whims. Congress should give them a budget and let NASA and/or a science panel decide what probes fund. It is hard to keep ramping up and ramping down repeatedly and still have enough lead time for testing, etc.

Re:Radiation hardness

[AKAImBatman]

They hired scientists, not computer engineers. Most scientists have the smarts to learn computer programming and can build off the shelf hardware (there's a big satellite equipment market that makes space-reliable stuff). However, they're not computer professionals. For what they paid for this little rover, I can think of a radiation hardened design with far more redundancy and capacity than they have in the rover.

For example, why didn't they consider IBM micro-drives? Those existed years ago when they started this project. They could have had a RAID array of three of them. Radiation hardening is still a problem, but that could be better solved by shielding the entire core computer instead of bothering with individual components.

Similarly, I would have used two computing units in a cluster. Yes, your specs would have to go up. My preference would have been an ARM processor, half a gig of RAM (possibly two 512 DIMMs per board that will act as redundant memory), a RAID array of 3 micro-drives (a Gig or larger per if I could get it) and redundant I/O ports via ethernet and XBand wireless. The whole rig would run QNX instead of VxWorks.

The only problem with my setup is power. The poor little craft wouldn't have any left to run its engines! No problem! Go for an RTG power source. About 10 pounds of plutonium dioxide should provide more power than the craft will ever need, for longer than the engines will last. Since the computer would be a sealed and shielded unit, it could possibly last as long as the RTG itself. It could keep attempting to collect and transmit data until either all the sensors fail, or the antenna goes.

Re:Radiation hardness

[kj0rn]

This is my biggest gripe about the rovers - hell, lets just use the amazing amounts of money spent on them to fund a bunch of amatures who have to complete a set of tests (dropping the robot from low earth orbit etc), show it on the telly, choose the best one. Let the robots 'evolve' as the builders learn more about their machines.

Re:Radiation hardness

[mikeee]

Probably not; VxWorks is a real-time OS - it's designed to guarentee interactive performance (eg, something like 'The navigation app *will* get 50% of CPU time over *any* 5ms interval').

I'd bet it doesn't support VM at all; too hard to know how long swapping will take. That's probably also why they apparently kept filesystem metadata in RAM - it's a lot easier to be sure how long file access will take if you don't have to go jumping all over the place looking for it.

Re:Radiation hardness

[barawn]

I would have thought that the memory would be swapped out

Swapped out? To where? You're being a bit recursive here. The operation failing is the initialization and reading of the flash. You can't swap to flash while you're initializing it.

However, I doubt they implemented virtual memory on the thing. It's far too much overhead in the OS to actually implement paging, etc. They were probably just very, very careful about the amount of memory in use at any one time, as with most embedded systems. Someone, however, didn't check that in a massively fragmented flash filesystem, the directory read wouldn't take up the entire RAM. Oops.

Re:Radiation hardness

[parc]

I imagine they hired computer engineers for the parts that required computer engineers, and software engineers for the parts that required software engineers. The scientists would be the guys responsible for supplying requirements to the engineering group.

I imagine thetemperatures involved in the rover (both in flight and on the ground) are higher than allowed by IBM microdrives.

Then again, maybe your trolling. It's hard to tell.

Re:Radiation hardness

[Hays]

They hired scientists, not computer engineers

Uhhh.. No, NASA definitely hires lots of computer engineers, computer science majors, etc. They don't ask the geologists scientists to code a file system. This should definitely be an embarrassment to the JPL code monkeys. This stuff should have been tested and reviewed. I can't believe their longest mission simulation was something like 9 days.

And if the module in question wasn't actually written by JPL, it should still be an embarrassment to them for a) buying it b) not testing it

Re:Radiation hardness

[shotfeel]

Was it a flash memory problem? Maybe I'm looking in the wrong places, but I haven't really heard anything even remotely concrete about what the problem was, or even what the current hypotheses are.

What worries me though, is they continuously state that the two rovers are twins and use identical computers. Maybe its just me, but if one "twin" develops a problem that may be genetic, I'm worried about the other one too!

Re:Radiation hardness

[Ben Hutchings]

You shouldn't swap to flash anyway - it'll wear out too quickly.

Re:Radiation hardness

[AKAImBatman]

I imagine thetemperatures involved in the rover (both in flight and on the ground) are higher than allowed by IBM microdrives.

I'm going to assume you mean lower, not higher. The operating temperature of micro-drives are 0 to 65C. The non-operating temp is -40 to 65C. Thus we need to keep the drive above freezing during operation. That's not really a problem though. Since I'm talking about a sealed computer case for shielding, the heat from processor operation alone, should keep the device nice and toasty.

Microdrive Specs

Looking at standard flash cards, their operating temperature actually seem to be worse. (0 to 55 C)

Example Flash memory

Some of the ones I've seen will go higher than 55C, but I haven't yet seen one that can operate below freezing. Nor have I seen one that can be stored below -20C.

Perhaps someone has a link to flash that can operate at extreme temperatures?

I imagine they hired computer engineers for the parts that required computer engineers, and software engineers for the parts that required software engineers. The scientists would be the guys responsible for supplying requirements to the engineering group.

These are not the same software/hardware engineers you find on the open market. Most of them are scientists or the type who work in scientific fields. They address the problem differently than you or I. That's probably a positive thing for space craft design, but for the "cheaper, better, faster" mantra, you need a very different type of engineering.

Re:Radiation hardness

[AaronW]

If anyone saw my earlier posts on VxWorks they would see I am not at all surprised about the problems NASA is having.

As someone with first-hand experience with VxWorks let me say that VxWorks' memory handling code sucks. Their malloc implementation has got to be the worst one ever designed. It fragments horribly and when fragmented has unusable performance. A malloc call can take many milliseconds when memory gets fragmented. Our box used to crash due to the fragmentation. I replaced Wind River's code with DLMalloc and that fixed the memory issue. We went from many tens of thousands of fragments to only a few dozen. Our reliability significantly increased as well after dumping Wind Rivers brain-dead malloc code. BTW, glibc uses a variant of Doug Lea's malloc code, so it's been widely tested.

Furthermore, in VxWorks there is no way to identify what process malloced a block of memory. There is no memory protection either (think DOS). If a task has a memory leak, VxWorks does not provide any method of tracking down the culprit. I had to add that support to the DL Malloc code I ported so we could find memory leaks and general memory usage by task.

Since malloc is critical to the working of VxWorks, if you run out of memory you are basically completely dead. Often the only way to recover is to wait for the hardware watchdog timer to kick in and reboot the system.

If a task dies in VxWorks with a bad pointer there is no way to recover other than reboot. The OS will not clean up after a task (i.e. free memory, close files, release semaphores).

As far as flash goes, VxWorks supports the FAT16 file system. As you know, FAT16 sucks. It only supports a limited number of files in the root directory. It's relatively easy to corrupt, and when corrupt it tends to corrupt itself even worse. There is no wear leveling support. If the FAT table or root directory gets corrupted, you're screwed. VxWorks is even worse since there arn't tools to fix a corrupt file system.

VxWorks is not a scalable OS. The OS gets slower as the number of tasks increases. Realtime support sucks. Although it has support for things like priority inheritance to prevent priority inversion, the best guaranteed realtime latency is half the system tick rate (the tick rate is usually 10ms).

Also remember that unlike open source operating systems, the source code to VxWorks is not available unless you pay some major $$$. Without the source you're basically working blind.

VxWorks is an old RTOS, and its age is definitely showing. It is not a robust OS.

As far as turning VxWorks into a firewall, you'll need to write all your own code. The VxWorks TCP/IP stack is an archaic vulnerable version of the BSD stack. TCP sequence number guessing is trivial. There is no built-in support for firewall support, NAT, or anything else. I have heard many many complaints about the VxWorks networking code. Although the box I'm working on is a router and broadband remote access server, we don't use the VxWorks TCP/Ip stack much. I am sure the VxWorks stack is vulnerable to many of the current DOS attacks as well.

Some of you might say why not use VxWorks AE, which adds memory protection. Think hacked-in memory protection far worse than Windows 95. It's slow, very buggy, and poorly supported. We tried AE and after many months we dumped it and went back to the non-AE version. Very few companies actually went with AE.

Today there are many other choices for an RTOS that are better than VxWorks. For our next major project we're looking at both QNX and TimeSys Linux. QNX is a true microkernel design with the core being around 70K. Every driver is protected and can be restarted if it crashes. I think you can even buy a medical grade version of QNX. TimeSys Linux is also pretty cool, with excellent real-time support and all the advantages of Linux. For something like the Mars rover, QNX would be better due to the limited amount of memory and greater robustness.

Wi

Re:Radiation hardness

[shotfeel]

Just found http://www.space.com/missionlaunches/spirit_sendsp ic_040129.html which says,

"Engineers have since determined that the problem appears to involve too many files overloading the onboard computer memory"

to answer my own question.

Re:Radiation hardness

[wideBlueSkies]

>>For example, why didn't they consider IBM micro-drives?

SLAM. Bounce, Bounce, Bounce, Bounce, Roll, Bounce, Roll, Roll, Bounce, Roll, Roll, Roll.

Probably too much shock for the drives.

wbs.

Re:Radiation hardness

[AKAImBatman]

SLAM. Bounce, Bounce, Bounce, Bounce, Roll, Bounce, Roll, Roll, Bounce, Roll, Roll, Roll.

Probably too much shock for the drives.

Surprisingly, no. From everything I've heard about these drives so far, making them smaller has made them more reliable. It seems that they're able to take quite a bit of shock. At least as much as a normal flash drive.

Of course, triple redundant drives (notice that's the only component I went triple redundant on) would allow you to switch off any that were damaged.

Re:Radiation hardness

[Ryan Amos]

Yeah, except if it were built by monster garage techs it would have a 5000 watt stereo system and flamethrowers on the exhaust. Which is not necessarily a bad thing...

Re:Radiation hardness

[QEDog]

I'm currently working designing an experiment that is going to be run remotely in a hazardous enviroment. The experiment's DAQ demands that we have something fast enought, in the order of a microsecond. We were considering using some PowerPC or 68k board running VxWorks (industry standard) for this work, and use the TCP/IP to communicate through a sat-phone. After reading your post, I'm willing to reconsider other approaches (we are still at a very early stage). Any suggestions, links or more info?

Re:Radiation hardness

[Anonymous Coward]

For example, why didn't they consider IBM micro-drives?

I'm quite sure they aren't rad hardend. It's pretty funny how so many people on Slashdot thing they are smarter than JPL. I doubt the rest of your pipe dream would fit inside the rovers size and weight limits.

Re:Radiation hardness

[AKAImBatman]

Um, wouldn't that be why I was talking about a radiation hardened *case* instead of radiation hardened *components*?

* AKAImBatman smacks Anonymous Coward along the backside of his head.

Re:Radiation hardness

[Anonymous Coward]

I'm not defending VxWorks, but my preference for real-time design is to NOT use dynamic memory allocation (malloc) or dynamic anything else except during startup. If I need to protect a region of memory, I'll setup the MMU myself (statically with no page swapping) since their basic MMU package is not very robust.

BTW: VxWorks 5.5 does support FAT32.

--Robert

Re:Radiation hardness

[Anonymous Coward]

If you are allowed to pour you hart out about how vxworks sucks for your router/broadband project then can I offer some clues and guesses as to why NASA might have chosen it?

The article stated that NASA needed to be able to patch code while it was running, a very reasonable demand as one cant just get the rocket back to earth and start the whole thing over.

Qoute: "In addition to VxWorks' reliability, the system allows users to add software patches -- such as a glitch fix or upgrade -- without interruption while a mission is in flight. "We've always had that [feature] so you don't have to shut down, reload and restart after every patch,""

Iirc from the nat. geographic documentaries the spirit code was still being developed while the rover was on its way to the planet. (How is that for missing a deadline, they must have learned this one from the game industry ;-) ) I may be mixing things up with the orbiter though. This would be a perfect explanation as to why there apears to be a bug/situation which would have been found with running long tests collecting lots of "science data" on earth.

Now you go on about the vxWorks malloc implementation (dynamic memory managment features). NASA may actually have little need for having the OS manage the memory. It apears most of the colected data is stored in a filesystem on flash ready for pickup by a comunication system sending it back to earth (By any available orbiter). Now the problem with fragmentation is that it is likely to get worse as soon as it starts, but with 128 MB of ram and *very* few procecsses I dont see how you could get a problematic or fatal amount of fragmentation unless there are multiple procecses collecting a lot of data at the same time in very high resolution mallocinging tiny amounts (less then a page) at a time with little memory left to spare. Neither of these conditions seem to be likely to be met in any big way, but I havent seen this codee ofcourse.

Now you mention tcp/ip is bad, and I fully believe you. (eventhough my motorola cable modem claims to be running vxworks, consumer broadband must be at least possible with this rtos ;-) ). But, obiously, NASA is not gona use tcp/ip as a protcol and the drivers for the radio hardware are not gona be a standard part of vxworks. I think these pieces of code for sending stuff over high frequency duplex radio data links have been laying around at nasa (or defence contractors) tested and proven on many applications just like many other components of this machines software. This bring me to my other guess as to the choicec of vxworks, the rover has to pretty much plot its own course over the planet. It seam possible that the rover now on mars uses the excact code for that that early prototypes used while spending months in american deserts. VxWorks might have been the os of choice by the prototye engineers, but porting this code to another OS might introduce bugs which goes for all the code nasa already had (tested, and perhaps even proven in space)

Now unlike VxWorks I have run QNX, (as has everyone who has had the time to go over here, have a look If only to brag about using another os on your desktop). While I agree that it is a very very nice os, I could very well see it not having any features over vxworks that would make it worth it for nasa to port all their current vxworks stuff over only to have to test it again while still not knowing for sure if it will run *excactly* like on vxworks. And ofcourse any features vxworks did lack (or rather did have while not being supposed to ;-) ) could have been fixed by nasa coders in no time just like you did only with the help of the source.

Re:Radiation hardness

[Anonymous Coward]

Your comments about VxWorks show me that you don't have a well establish practical Real Time OS (RTOS)experience and are trying to to compare an RTOS to a Server or Desktop OS. That is quite a bit misleading. For example,
consider the use of malloc()/free(). In an RTOS it does not much sense to dynamically allocate and free memory at run time. Same thing applied to other sytem resources, like tasks/processes. This takes quite a bit of time from a Real Time point of view to create new tasks, semaphores, objects, memory allocation, etc.. Most RTOS engineers, who know what they are doing, pre-allocate all the memory they need at boot time, start all the processes they will need, etc. In order words you allocate all your system resources at boot time so you keep your real time characteristics. An RTOS system is static. no dynamically process allocation, memory allocation (i.e. unexpectantly running out of memory), etc. You start everything up and account for your needs. A very different model from probably what you are used to in a Desktop/Server OS. Think about it as static appliance. Give me one good reason why you can't preallocate memory and all the tasks you will need?

Yes memory protection has its benefits. Speed is not one of them. Remember this is an RTOS. How many RTOS do you know of that support memory protection? hmm.. 1 I believe. Which is VxWorks AE. Which you say is slow. gee. If you want a low latency fast RTOS, memory protection quickly falls off your "must have" list.

So VxWorks supports FAT16. So that makes it suck?
It also supports FAT32, NFS, a Flash Filesystem, etc. So all those points make it such too. I see your reasoning. ;-P

VxWorks, every other true RTOS, does not use the system clock exclusively to multitask a you say. In fact, round robin multitasking which uses the system clock is turned off by default Unlike a server/desktop OS, a context switch is not generated strictly via an interrupt (system clock, device interrupts). context switches can occur at any time as your code generates events. For example, giving a semaphore to unblock a test, sending message to a message queue which unblocks a waiting receiver, trying to get hold of an object which is in use (this would block you). This does not require a high latency and it almost immediate. This is because all code runs in processor supervisor mode. i.e. no traps to communicate from one task to another. This goes along with the no memory protections reasoning. It has benefits. Let me guess you didn't read the manual or take a vxWorks 101 class, right?

Nasa has been using vxWorks primarly on most of their projects since 1980! I think that says alot by itself. Thank you for your "expert" opinion in helping people judge the quality of an RTOS.

Re:Radiation hardness

[Anonymous Coward]

If anyone saw my earlier posts on VxWorks they would see I am not at all surprised about the problems NASA is having.

Yes, and we've also seen other opinions from other people.

I sure hope NASA spent a lot of time and replaced major parts of VxWorks. I also hope NASA has gotten better support out of Wind River than I've experienced. Wind River's support has been virtually non-existent for us, even though we pay a lot of money for it.

At a previous startup we were having problems with memory corruption because their functions to disable the cache did not flush it as they said it should. When they finally did respond and I had already fixed it, they asked me how I fixed it. Damned if I would tell them.

One suspects your attitude has much to do with theirs. I wouldn't want to support you either.

Re:Radiation hardness

[Anonymous Coward]

Your comments about VxWorks show me that you don't have a well establish practical Real Time OS (RTOS)experience and are trying to to compare an RTOS to a Server or Desktop OS. That is quite a bit misleading. For example,
consider the use of malloc()/free(). In an RTOS it does not much sense to dynamically allocate and free memory at run time. Same thing applied to other sytem resources, like tasks/processes. This takes quite a bit of time from a Real Time point of view to create new tasks, semaphores, objects, memory allocation, etc.. Most RTOS engineers, who know what they are doing, pre-allocate all the memory they need at boot time, start all the processes they will need, etc. In order words you allocate all your system resources at boot time so you keep your real time characteristics. An RTOS system is static. no dynamically process allocation, memory allocation (i.e. unexpectantly running out of memory), etc. You start everything up and account for your needs. A very different model from probably what you are used to in a Desktop/Server OS. Think about it as static appliance. Give me one good reason why you can't preallocate memory and all the tasks you will need?

Yes memory protection has its benefits. Speed is not one of them. Remember this is an RTOS. How many RTOS do you know of that support memory protection? hmm.. 1 I believe. Which is VxWorks AE. Which you say is slow. gee. If you want a low latency fast RTOS, memory protection quickly falls off your "must have" list.

So VxWorks supports FAT16. So that makes it suck?
It also supports FAT32, NFS, a Flash Filesystem, etc. So all those points make it such too. I see your reasoning. ;-P

VxWorks, every other true RTOS, does not use the system clock exclusively to multitask a you say. In fact, round robin multitasking which uses the system clock is turned off by default Unlike a server/desktop OS, a context switch is not generated strictly via an interrupt (system clock, device interrupts). context switches can occur at any time as your code generates events. For example, giving a semaphore to unblock a test, sending message to a message queue which unblocks a waiting receiver, trying to get hold of an object which is in use (this would block you). This does not require a high latency and it almost immediate. This is because all code runs in processor supervisor mode. i.e. no traps to communicate from one task to another. This goes along with the no memory protections reasoning. It has benefits. Let me guess you didn't read the manual or take a vxWorks 101 class, right?

Nasa has been using vxWorks primarly on most of their projects since 1980! I think that says alot by itself. Thank you for your "expert" opinion in helping people judge the quality of an RTOS

Re:Radiation hardness

[JackHart]

If I had mod points I would mod you up! Grandparent is comparing apples and oranges.

Re:Radiation hardness

[JackHart]

Read comments at 0 to see some more insights... Grandparent is comparing apples and oranges.

Re:Radiation hardness

[tezzer]

Why the hell would you be doing dynamic memory allocation in an RTOS? Bad coder. Bad, Bad, Bad.

Re:Radiation hardness

[Anonymous Coward]

...and your system is what you get when a computer engineer designs a system when they don't have much experience designing for harsh environments. For starters, the read/write heads of a microdrive, like any other hard drive require an atmosphere somewhat more substantial than Mars's to maintain their clearance above the platter. Microdrives have issues once you start to get over 10k ft asl on earth. Perhaps operating in an atmosphere of CO2 (heavier molecules than air)will allow operation at slightly lower pressure, but that effect is unlikely to matter at the 7 or so millibars on Mars. Oh well, just build a pressurized chamber, it'll only add an otherwise useless few kilos.

Oh yeah, with the lower pressure, your uber system is going to have a lot more difficulty dissipating heat now that it's lost most of its ability to lose heat through convection. I suppose you can get rid of heat by directly exposing heatsinks to the cold near vacuum to radiate it away. You also get to crack your motherboard in half in the bargain when you go through your first thermal cycle the night after landing.

Re:Radiation hardness

[Johnny Vector]

For example, why didn't they consider IBM micro-drives?

Vacuum and radiation. The drives may be semi-hermetically sealed, but you need a whole nother level of reliability for space missions. If the drive area leaked to vacuum they would head-crash. But more importantly, it's highly unlikely they could survive the radiation environment.

Radiation hardening is still a problem, but that could be better solved by shielding the entire core computer instead of bothering with individual components.

Nope. Basically, you can't shield against Single-Event Upsets. The solar protons are too energetic; you need a prohibitively thick shield (inches, or feet, of metal). A flyable amount of shielding will reduce the total dose, which helps with radiation-induced aging, but won't stop those random bit-flips. Your only hope for them is to build the circuitry specifically to handle them. Which generally means some combination of bigger, slower, and more power-hungry. Since you can't afford more power, you get bigger and slower. Micro-drives were not designed with that tradeoff in mind.

About 10 pounds of plutonium dioxide should provide more power than the craft will ever need, for longer than the engines will last.

Plus all the associated radiators and electronics necessary to run an RTG. But admittedly the biggest obstacle to them is political, not technical. Please go convince the noisy bit of the public that RTG != Nuclear Bomb. Thanks.

Re:Radiation hardness

[Anonymous Coward]

For example, why didn't they consider IBM micro-drives? Those existed years ago when they started this project. They could have had a RAID array of three of them. Radiation hardening is still a problem, but that could be better solved by shielding the entire core computer instead of bothering with individual components.

• Radiation is the least of your problems, can you point me to a drive that is *proven* to work even after beeing shaking by a *huge* rocket for minutes at liftoff (160 db noise and vibrations) and beaten by a landing with the parachute not working as calculated (the martian athmosphere is very thing, how does a chute work there compared to here? Let hope better the original chute design which failed in tests, The new ones have only been tested in a windtunnel). Now the airbags end up leaking (heaponed during test, these are improved versions with an extra layer iirc) or even without any of these very possible failures? Now what if something else goes wrong? What if the whole thing gets dropped during development, are you gonna open the extra radiation protected shielding open spending weeks reasembling the whole thing just to make sure the drive is still ok?
• Radiation in silicon can be solved not only by shielding but also by silicon design, which mean less weight in shields.
• Why microdrive? If nasa had any use for gigabytes up there (Yeah I know they do now.... but thats a bug ;-) ) they could use a drive that is designed to have bigger changces of surviving big bad hits and shakes *all the time* no mather what size.
• Are you ganna pay for having extra shielding + extra bataries + solar panals to protect, heat and power this drive flown over to mars?
• Are you gona explain to all these profesional satalite and spaceship builders why you want to be the first to even try this?

Similarly, I would have used two computing units in a cluster. Yes, your specs would have to go up. My preference would have been an ARM processor, half a gig of RAM (possibly two 512 DIMMs per board that will act as redundant memory), a RAID array of 3 micro-drives (a Gig or larger per if I could get it) and redundant I/O ports via ethernet and XBand wireless. The whole rig would run QNX instead of VxWorks.

• A "cluster", well actually having two computers checking each other (cluster being something in which interconnection is done in software), no. A system with hardware comparing the output of computers like used in planes, yes please, I guess the rover will just have to spend its power on this extra hardware.
• It looks like the rover will spend this power on high speed ARM chips to. To bad really, I guess a chip that has been in so many satalites and deep spacee probes cant beat portting and debuggin yet more excisting code again (Ofcouse testing if radiation resistant ARM chips will really live through years on mars after months unprotected from the sun in a rocket will be more then just sticking them in the microwave for a few minutes, lets just put this item on our todo list shall we?)
• Wow it seems we have some computing and storage power to spare.. I know lets make videos Ofcourse there is little moving on mars but hey *if* there are green men we can invite then in a interplanetairy talk show ;-) (Note: upgrade uplink (128kb/s wondt do for video), dont forget bigger antenna)
• Hm we still cant fill al this bandwith and diskspace lets start collecting surface temperatures in a gazillion numbers behind the comma, what no room for a thermomether? Ah never mind the real scientist will be pissed but hey they will be to bussy building this thing. I hear their job is no longer in the choosing the data to be collected and then hopefully analizing it but to build spacecraft. I guess some of them will know what they are doing. The engineers will get a brake then? ;-)
• QNX you say, wise choice indeed, I hear you are gona be the one to port all the

Re:Radiation hardness

[Anonymous Coward]

As bad as that sounds, that is common with RTOS. I worked on a pSOS based project, and we had similar memory management problems.

Generally, you have to remember that most RTOSs were orginally designed for very small embedded systems with limited amounts of memory. So things you are used to on Linux or other systems might not be there in order to keep the size down so it can fit into smaller and cheaper hardware to make a product more competitive.

I have no experience with QNX because they did not support PowerPC. Did they ever add support? I'll go check after my post.

You do make a good case for being able to see the source code though because of the problems.

Re:Radiation hardness

[AKAImBatman]

*chuckle* Thanks for that light look at space craft design. I'll try to address the most important ones:

1. As I said before, these Microdrives are surprisingly resilient. Still, the drives should be deactivated and heads parked until takeoff is complete. Certifying that this will work would add expense, but you could pay IBM or an independent firm to do this. Worst case, you can switch back to "proven" Compact Flash without any major impact on development (same form factor). The only downside is that you may have less disk than you had hoped for.

2. You're pointing to power concerns. As I said, 10 pounds of plutonium would do it. My solution to handling the environmental whiners is to keep sending up RTGs until they get used to the idea.

3. Shielding in silicon is exactly what I'm *not* looking for in this design. I'm thinking a sealed lead case. Yes, that will add a few kilograms, but not much more. Some weight will also be saved on not using the "hardened" components. Even better would be lead layered with some of that new plastic shielding. That would help reduce the weight.

4. Don't be looking for new uses for the extra computing and storage power. The point of overspecing the machine is so that errors like the one on Spirit don't happen. It also allows for defective areas of the hardware to be turned off without impacting the overall power of the machine.

5. Part of my point was that QNX would have been a better choice from the get-go. VxWorks is not what I would tend to think of as my first choice for such a mission. My idea is to make a cheap machine that tolerates failure. "Cheaper, better, faster." After all, that's the thinking of computer languages like Java. Very simply, "Software and hardware are too complex. Thus software should adapt against failure instead of trying to eliminate all failure during development." Besides, NASA bought or used the built-in solution for most of the drivers they used. No reason why you can't buy similar drivers for QNX.

6. I figure my design would add probably about 60 pounds to the design. I'm not a rocket scientist, but I think that can be accounted for when fueling the rocket.

7. I'm not looking to prove these components to the same degree that previous missions have. "Cheaper, better, faster", remember? When 90% of your cost is certifying your parts for situations that may never happen, how is that cheaper or faster? Thus you're stuck with underspeced hardware that proved itself before the general populace even used the Internet. How is that better?

8. RTGs are heaters unto themselves. Also, the processor will provide a certain amount of heat. Since it's sealed in its radiation hardened case along with the microdrive, the heat will actually build up. Thankfully, we decided to use lead (or at least some lead) which will act as a cooling surface. Thus the inside of the case will stay warm during operation.

I'm not saying that I'm way better than NASA at this, but I do have a bit of a different viewpoint than they do. Perhaps my ideas wouldn't work. Or maybe they would. Either way, it's not a bad choice for investigation. NASA has to be looking to develop technologies that will allow an eventual colony in space. One way to help that happen is if we can look at ways of making off the shelf components workable in space.

Re:Radiation hardness

[Anonymous Coward]

omg thank you, I was wondering when someone would point that out.

Re:Radiation hardness

[AaronW]

While I would agree with you that avoiding malloc and preallocating memory is the way to go, but it is not always possible. In my case, we are using various 3rd party libraries, and changing them to use static memory allocation would be prohibitave. In at least one case, the third party library source code was not available. Also, in many cases the dynamic nature of some algorithms requires dynamic memory management. You cannot statically allocate everything, especially in a limited memory environment.

I know we're not the only ones to have been burned by Wind River's malloc. I know several major companies that also had to replace Wind River's code.

As far as being able to dynamically replace code, VxWorks isn't alone in that. Numerous other RTOSes out there can do the same thing, including QNX. QNX even supports the concept of a hot standby process to take over if the main process dies.

To give you an idea about how Wind River's malloc works, they keep a sorted linked list of fragments from the smallest to the largest. When you try and allocate a block, it walks the linked list until it finds a block large enough. Likewise, when you free a block it checks if it can coalesc the block with a neighboring block. It then goes through the linked list looking for a slot to insert the free block.

Yes, VxWorks may have been around since the 80's, but that's part of the problem too and it is showing its age. In the 80s embedded processors typically did not have MMUs. Now MMUs are quite common in the more powerful embedded processors.

You say you can't have low latency and memory protection? QNX proves that you can. It is low latency and *very* robust. If your driver dies, no problem, restart it. Timesys Linux also has a very low latency, although not as low as QNX. Timesys also has an interesting feature where you can guarantee CPU and networking resources. I can schedule a task to be guaranteed 5.8ms of execution every 8.3ms and it will guarantee that that task will get the CPU time allotted to it with the desired resolution. This is without increasing the system tick rate (usually 10ms). Timesys can also schedule a task to be higher priority than an interrupt. I'm not as familiar with QNXs scheduler, but it's also quite flexible from what I've heard.

As far as FAT, it is not a robust filesystem. It never has been. If the FAT gets corrupted or a directory entry gets corrupted it's difficult to recover. Other than possibly having 2 copies of the FAT cluster table, any corruption can be difficult to repair. If the FAT table gets corrupted, which table is corrupt and which is not? If a directory entry gets corrupted, it can be impossible to fix. For flash memory, unless you are using a device with special wear-leveling, FAT is about the worst choice since any file write that changes the size of a file requires a write to the directory entry and possibly the FAT table. If the table gets corrupted and you don't run a repair operation (which often ends up leaving orphaned files as lost clusters), the file system can happily corrupt itself to death. Why do you think every time DOS/Windows9x/ME crashed it had to repair the disk with scandisk? FAT is a poorly designed file system that was originally designed for 160K floppies and scales poorly. FAT32 is an improvement, but it's still not very robust. For flash, something like Linux's journalling flash file system 2 (JFFS2). More information on VxWorks file system support can be found here.

Basic VxWorks information can be found http://www.slac.stanford.edu/exp/glast/flight/docs /VxWorks_2.2/vxworks/guide/.

Re:Radiation hardness

[Cap'n Canuck]

As great as QNX is compared to VxWorks, it
A) Would never have been chosen as the OS of choice

B) Will never BE chosen to replace VxWorks

Why? It's a great company, but it's based in Ottawa (that's Canada for you Yankees), and NA$A Bucks do not flow over the border.

Re:Radiation hardness

[Anonymous Coward]

Has anybody used the SMX RTOS (true microkernel)? It has a very small, fast, footprint. I used it on am embedded project, and was very impressed.

Re:Radiation hardness

[AaronW]

Take a look at QNX. I'm not very familiar with many other RTOSes. RTEMS is another interesting one I looked at (is open source). If you do decide to go with VxWorks, make sure you get the source code. A word of warning, though, VxWorks TCP/IP support isn't all that great and performance is mediocre. It's based on an ancient BSD stack.

QNX has a lot of reliability features built in that are not present in VxWorks, Linux, or RTEMS. It can take periodic snapshots of a task and recover if a task dies.

Also, field debugging of VxWorks can be a real pain. It does not generate a core file you can debug later.

Timesys has very good real time support, but the context switching time is on the order of 10us depending on the processor, QNX is better.

QNX is more more like a unix environment with real threads and processes. QNX relies mostly on effecient message passing between tasks. QNX also has much better TCP/IP support than VxWorks, including support for IPv6. From what I can tell, development and debugging in QNX is much easier than VxWorks. QNX does not have as many priority levels as VxWorks. QNX Neutrino supports 64 priorities. VxWorks supports 256. Timesys Linux can support thousands of priorities.

QNX also has built-in support for distributed processing as long as there is an interconnect. It also supports SMP (VxWorks does not).

VxWorks, on the other hand, is like a single task with many threads. A "task" in VxWorks is more like a thread in that all memory is globally accessable. Any function can call any other function.

I havn't worked with QNX or Timesys personally. I've spent the last 4 years developing for VxWorks Tornado II. The group where I work dealing with OS related issues has more or less decided to use either QNX or Timesys for our next generation software. Timesys looks good from the tools perspective and the wide Linux experience out there, plus the fact that currently our product runs on a mixture of Unix and VxWorks, plus many of the drivers we need are available for Linux and would need to be ported to QNX. Some of the key software we are using has been ported to QNX, though, and the reliability of QNX is also of interest. QNX has also won some nice design wins like Cisco. Since we make routing equipment we tend to pay attention to things like that.

Also, beware that Wind River's support isn't very good from our experience. It basically boils down to "what support?" even with a support contract. Wind River has layed off a lot of people lately and I'm sure that hasn't helped the company. They've been bleeding money like crazy and are hurting pretty bad. They've been losing market share to other embedded RTOSes and Linux. Hell, Wind River has ported their development tools to support embedded Linux (I wonder why?).

As far as real-time Linux is concerned, I looked at several of them. RT-Linux requires that RT tasks talk to the RT kernel and use a separate API and drivers than the Linux tasks. It's basically Linux running on top of a RT kernel. Monta Vista does not provide hard real-time, and when I last spoke with them they had no solution for priority inversion. Timesys modified the stock Linux kernel so RT tasks don't need to use a separate set of APIs from non-RT tasks and Timesys can guarantee timing, both for the CPU and networking. For soft real-time, Monta Vista seems to be pretty popular. The people I spoke with who used it said it was quite reliable but couldn't comment on support since they never needed any.

-Aaron

Re:Radiation hardness

[neomorph]

You would have to look on ebay for 68k VxWorks board. Nobody has had a new design on 68k for a half a decade now. You are going to need a PowerPC running well over 200 MHz for that kind of response time.

If you really have 1 microsecond response time requirements, you're going to have to either find:

a) An RTOS that doesn't disable interrupts, or
b) Queue it up in hardware.

If you go A, that rules out QNX, and all other Linuxes, real-time or not.

Most companies go B, and then still use an RTOS with a good response time, since the RTOS still has to suck it out of the hardware before the queue fills up.

One RTOS, similar in architecture to QNX, that doesn't disable interrupts, is INTEGRITY

Re:Radiation hardness

[KewlPC]

The problem was that there were a TON of files on the flash drive, and the flash management software wasn't able to handle them all. It would freak out and crash, causing the watchdog timer to reset the computer.

The solution so far has been to disable the flash drive until they find out exactly what part of the software failed and can fix it.

Re:Radiation hardness

[Cujo]

Hey, you can qualify a hard drive for space use, let me know. So far, no one's done it AFAIK. Some of the issues:

• Vaccuum. I don't think there's any hard drive that will work in a vacuum.
• Radiation
• EMI/EMC, including Magnetic cleanliness
• Temperature range. Standard range is -34 deg C to +71, although that's usually negotiable. The less hardware you have to baby, the better.
• the cost of a qualification program
• Reliability and life.

6. I figure my design would add probably about 60 pounds to the design. I'm not a rocket scientist, but I think that can be accounted for when fueling the rocket.

60 pounds is a hell of a lot. Mass is always tight, and there is no easy fix by "fueling the rocket." Would that there were!

Re:Radiation hardness

[PinkyGigglebrain]

Actualy VxWorks does have a way of generating a "core dump" of sorts. It doesn't do it by default like a desktop OS. You have to use Windview (a profiling tool), and it does affect OS overhead, hence it being off by default. It generates an image of what tasks/code where executing when the system crashed. Not quite a true core dump but if its setup right it can tell you where in your code the crash happend, sometimes even down to a specific block of code.

Re:Radiation hardness

[AKAImBatman]

Vaccuum. I don't think there's any hard drive that will work in a vacuum.

By sealed, I meant "hermetically sealed". Since there's nothing living inside the case, the air can be sitting.

Radiation

As I said, shielded container. Although, someone pointed out that it wouldn't stop some of the most energetic stuff. The big question is whether a few inches of shielding would be enough to substantially reduce the radiation. We don't need it perfect, just good enough.

EMI/EMC, including Magnetic cleanliness

I didn't think that EMI would be a big problem. What would generate the interference? Hmm... perhaps a Faraday cage could be built into the casing?

Temperature range. Standard range is -34 deg C to +71, although that's usually negotiable. The less hardware you have to baby, the better.

Operating temp is 0C to 65C for microdrives. I originally stated that by sealing the device in lead, you would have a passively cooled system. (The lead would heat and bleed off the extra into space or atomosphere.) However, the more I think about it, the more I like the idea of having an active liquid or gas cooling system. This would also be built into the lead casing, and could be controlled by the computer itself. You'd need a temperature sensor inside the case, but that's not difficult or expensive.

the cost of a qualification program

Personally, I'd like to cut as much of this out as possible. "Cheaper, better, faster" remember? Probably the best way to handle this is to build the entire unit, and stress test the entire thing (Vibration, radiation, etc.). That would be cheaper than testing every component, and would give an good idea as to how sturdy shielded components are when compared to individual hardened components.

Reliability and life.

So far, Microdrives have been shown to be highly reliable and generally last. Given that flash cards are slowly destroyed by rewriting, microdrives may have a longer life time.

Here's an idea. Instead of triple redundancy, go for two microdrives and one CF card. RAID these together. Now if there truely is a failure in all drives, the CF can take over. Personally, I don't think it would be a problem, but that way you have a compromise between proven and unproven.

60 pounds is a hell of a lot. Mass is always tight, and there is no easy fix by "fueling the rocket." Would that there were!

Mass may be tight, but I find it difficult to believe that it's impossible to add 60 pounds to the craft. The craft is chosen based on the requirement for the mission, right? And the amount of fuel burned by the rocket is calculated for exact insertion into Mars orbit, right? So it's conceivable that the rocket could have extra capacity. (What are they using for interplantary missions these days? Titans?) I suppose I'd need to check the specs on the rocket.

The weight could be reduced by eliminating a computing unit. Instead, give the single unit dual processor capabilities. Both processors will attempt the same calculation in sync, which will be then checked by a simple bus controller. If the results don't match, the instruction is recalculated. That would reduce the power requirements as well. Say 30 pounds now? (Most of the weight is in lead shielding and plutonium.)

It would be interesting to be on one of these projects and have access to the mission limitations. I'm willing to bet that they could be pushed pretty far with a little creative thinking. :-)

Re:Radiation hardness

[dellis78741]

As far as support goes, I suspect that Wind River is providing NASA with 24/7 on-the-spot support from their very best people.

Re:Radiation hardness

[Anonymous Coward]

A couple of other comments as much of your info is dated:

If you do need to do dynamic memory allocation and want to monitor how much, which tasks, etc. you can use MemScope from RTI.
http://www.rti.com/products/scopetools/memsc ope/in dex.html
It plugs in to VxWorks with no code changes, but as pointed out previously you need to be careful with dynamic memory allocation in an RTOS.

With regards to filesystem choices all Wind River platform products now ship with TrueFFS
http://www.windriver.com/products/true_ff s/index.h tml
There are also available from a 3rd party a transactional filesystem and other flash file system. Check out: http://www.datalight.com/product_detail.asp?p_id=2 2&archive=0 and http://www.datalight.com/product_detail.asp?p_id=2 1&archive=0

Wind's TCP/IP stack has been updated recently and what ships in their PNE product is now a router stack instead of an end station stack.

Finally, if you purchase any of Wind's Platform products http://www.windriver.com/platforms/index.html you get full source code to VxWorks including the entire kernel. This is their response to ease of source code access available with Linux.

20 MHz is not enough

[dimss]

20 MHz? Not enough for firewall if you have something faster than modem. Really.

Re:20 MHz is not enough

[kyknos.org]

I still run my 7Mhz Motorola 68000 Mac and it is good even for text browsing :o) 20 Mhz is hell fast!

__

Re:20 MHz is not enough

[hh1000]

The Sinclair ZX81 ran at something like 3.25Mhz. Due to a desire to create the lowest cost computer possible for students, the Z80 generated the video signal in software!

http://www.honneamise.u-net.com/zx81/zx81faq.htm l

Clever programming counts for a lot!

Re:20 MHz is not enough

[32bitwonder]

No? My 25Mhz Macintosh LC III is enough to host a webserver running Linux & Apache. It even runs PHP. The 20Mhz PPC CPU used in the rovers is most likely quite a bit faster than the 68030 in the LC III.

Re:20 MHz is not enough

[Ignorant Aardvark]

My 25Mhz Macintosh LC III is enough to host a webserver running Linux & Apache.

Post a link to your website and then we'll reconsider your statement ...

Re:20 MHz is not enough

[32bitwonder]

32bitwonder.org

Re:20 MHz is not enough

[Ignorant Aardvark]

Very nice, I didn't think you'd go through with it. Nice site, and it seems to be keeping up with the "load" perfectly so far (though I imagine if it got a +5 rating, you might be in some trouble).

Re:But The Question is..

[danheskett]

No, it runs vxWorks.

It says so right in the overview of the article...

You doin' okay?

Re:But The Question is..

[Quill_28]

...I'm sure it could run Linux, but currently it is running VxWorks.

What's inside the rover?

[Chilltowner]

Darn. Interesting articles, but I was hoping that inside it was filled with creamy nougat center. Oh, wait. I'm thinking Mars bar. Nevermind.

Wait a second...

[deitel99]

The machines aren't as slow as the top post says... they don't run at 20MHz, they are "capable of carrying out about 20 million instructions per second". Depending on the complexity of the instructions, the processor actually runs several times faster than 20MHz.

Re:Wait a second...

[danheskett]

That's an excellent point. A lot of people are thinking instruction = 1 cycle. The real world is that it's not unusual for an instruction to take 2, 4, 10, or even 100 cycles. The reality of the matter is that instructions can be anything from a single two bit sum to a floating point division. I see this mistake a lot... bravo for applying what you read against the supposition of a simplification.

Re:Wait a second...

[Zog The Undeniable]

Actually, they're probably slower :-P

Modern superscalar (pipelined) processors have a lot more MIPS than megahertz.

Re:Wait a second...

[Urban Garlic]

Although since they're PowerPC chips, of course, even if they ran at 20 MHz, they'd still be a *lot* faster than comparable Intel. You can't just go by the specs. You have to benchmark it in real world applications and think about overall system performance...

Re:Wait a second...

[AppyPappy]

If they had used an Amiga, they wouldn't be having problems right now.

Re:Wait a second...

[kinnell]

they don't run at 20MHz, they are "capable of carrying out about 20 million instructions per second". Depending on the complexity of the instructions, the processor actually runs several times faster than 20MHz.

And you got this information where? If I remember correctly the PowerPC specification states that the processor can complete an instruction every clock cycle. Even modern CISC architectures with deep pipelines are capable of accomplishing this providing the pipeline can be kept full.

Re:Wait a second...

[Rootbear]

Actually they are running at 20MHz. I've seen several write ups which clearly state that. The RAD6000 can apparently run at up to 33MHz, with a claimed 35MIPS. The rovers are "underclocked", probably due to power budget concerns.

Go to
http://www.iews.na.baesystems.com/space/rad600 0/ra d6000.html
and click on the rover picture to get a PDF brochure, which gives the 33MHz/35MIPS figure.

Rootbear

Re:Wait a second...

[deitel99]

And you got this information where?

The quote's from the article.

I'm just pointing out the difference between the article and the post linking to it. Although true that it is possible for the processor to run one instruction per cycle, it doesn't claim that in the article, and assuming it is the case just because it can be is a little careless.

Re:Wait a second...

[zhenlin]

The POWER chips initially required that all instructions took 1 cycle to execute. (They didn't have floating point, as such)

I don't know if this attribute carried down to the early PowerPCs, but I know the newer ones certainly don't.

Re:Wait a second...

[Have Blue]

Superscalar refers to multiple execution units, not pipelining.

Re:Wait a second...

[Anonymous Coward]

Don't you mean you need to benchmark in out of this world applications.

Re:Wait a second...

[Anonymous Coward]

Yeah, in if your daddy had pulled out, we wouldn't be having this problem of reading your stupid posts.

Re:Wait a second...

[cdrudge]

Ah nostalgia coming back...My first assembly class was on a RISC Sparc Station IIRC (did they make CISC Sparc Stations?). Every op code took one cycle. (or two in the case of jmp, but you got a free op before the actual jmp took place.) One of our first real projects was to implement floating point arithmetic in as few operations as possible. Shudder.

Re:Wait a second...

[edwdig]

NASA doesn't use very modern chips. They use old designs because they want stuff that's been well tested. If you're talking 20 MIPS, then we're probably talking about a chip that's 10-12 years old or so. I doubt there is much, if any, of a pipeline in the processor. This is probably around 486 level technology.

Re:Wait a second...

[Thuktun]

deitel99: The machines aren't as slow as the top post says... they don't run at 20MHz, they are "capable of carrying out about 20 million instructions per second". Depending on the complexity of the instructions, the processor actually runs several times faster than 20MHz.

danheskett: That's an excellent point. A lot of people are thinking instruction = 1 cycle. The real world is that it's not unusual for an instruction to take 2, 4, 10, or even 100 cycles. The reality of the matter is that instructions can be anything from a single two bit sum to a floating point division. I see this mistake a lot...

You both assume that the one who wrote the article didn't make the same mistake in the opposite direction.

In this article about the Stardust probe, the RAD6000 is said to be "a radiation-hardened version of the PowerPC chip used on some models of Macintosh computers" which "can be switched between clock speeds of 5, 10 or 20 MHz".

Re:Wait a second...

[Endive4Ever]

they'd still be a *lot* faster than comparable Intel.

Is the whole of existence still a titan 'Apple versus IBM (oops, Apple versus Microsoft now, isn't it?) struggle?

I can visualize two hundred years from now that there will be regular rockets, and Apple rockets, and still a few people left who refuse to embark to the moon base on the regular rockets, because of the 'Industrial Design' of the Apple rockets, which burn twice as much fuel.

Re:Wait a second...

[shotfeel]

And considering the PPC has several funtional units, it could actually carry out several instructions/cycle. Unless this PPC for some reason only has an integer execution unit and no floating point or others. For all I know it may even be altivec enabled -be a big help for manipulating images. Although I'm guessing based on the "early 90's" reference its probably G1 or G2 at best (601-604 range).

Anybody know any more specifics on the processor?

Re:Wait a second...

[Anonymous Coward]

Don't forget that we're talking of a RISC CPU. If (as I suppose) the author is talking about machine code instructions, then the equation 1 instruction = 1 clock cycle is not that far from being true.

Re:Wait a second...

[Octorian]

Well, they never did make CISC SPARCstations, because the "SPARC" is a RISC processor. However, before developing SPARC, Sun did make workstations using Motorola 680XX chips.

In any case, often the chip doesn't actually execute 1 instruction per clock cycle. What it does do is pipelining, which means on average it may fetch one instruction per cycle, and complete one instruction per cycle, but they're usually not the same instruction (where an individual instruction takes several clocks to complete).

Re:Wait a second...

[shotfeel]

And they'll be Apple zealots quick to correct your errors. Yes its idustrial design means a higher initial cost, but it actually uses much less fuel (compare the power use of a PPC vs. Pentium).

Not only that, they're easier to fly and cheaper to maintain.

Oh, and it doesn't get bogged down by all those viruses and worms. ..and its connected to the iTMS so you can plug in your iPod and purchase music to listen to on the way there (other players not supported).

I could go on, but....

Re:Wait a second...

[sootman]

>The rovers are "underclocked", probably due to power budget concerns.

pfft. the guys at Ars and HardOCP overclock stuff on a budget all the time. :-)

Re:Wait a second...

[Anonymous Coward]

If there is any kind of pipelining (which most RISC processors have) it will be able to execute different stages of several instructions in parallel so in reality it might be nearer 4 or 5 instructions per cycle. The whole point of having every instruction take 1 clock was to allow for this to happen. Also I don't know if these chips have any kind of branch prediction (I suspect not as this adds lots of complexity and messes up the ability to predict execution times) but if it does this will further boost the number of instructions per cycle in some cases.

Re:Wait a second...

[Endive4Ever]

Much less expensive fuel, but you have to ship it back to Apple to get it refueled, or hire a technician to do it for you.

Easier to fly, but you've gotta buy 'Approved for Schwinn' parts for it (remember the Schwinn snobs back when we were all riding around on our Huffys as kids?) which are expensive far beyond their somewhat-higher-than-the-cheapest-available quality. Definitely NOT cheap to maintain.

Re:Wait a second...

[daniel23]

But at least both the fuel and the technicians come in 5 different friendly colors you can choose from. So you can fly a rocket that matches your gf's eyeliner!

Re:Wait a second...

[Uplore]

Actually, the chips they use on the rover are RISC chips (Reduced intruction set) and are in no way comparable to the standard desktop CPU's you or I have on our desktop pc.

Self-warming

[Faust7]

To survive the frigid Martian night, MER computers are housed in warm electronics boxed heated by a combination of electric heaters, eight radioisotope heater units as well as the natural warmth from the electronics themselves.

Just leave off the heatsinks and fans, and everything should be fine.

Re:Self-warming

[Cyclopedian]

To survive the frigid Martian night, MER computers are housed in warm electronics boxed heated by a combination of electric heaters, eight radioisotope heater units as well as the natural warmth from the electronics themselves.[Emphasis added by me]

If obsessed environmentalists don't like NASA sending up probes with any radioactive material ('it might blow up, ohh..'), then how did this little tidbit get by them? Do they consider it non-radioactive? If they're only concerned by radioactive propulsion systems, then I think they're a bunch of hypocrites. Radioactivitiy is radioactivity whether it's propulsion or heating.

If they don't mind it, then let's send up those dune buggies with RTG and 18-inch wheels and cover a lot more of Mars.

-Cyc

Re:Self-warming

[Zog The Undeniable]

If only they'd used an Athlon, the planet would have been habitable in Bermuda shorts by the end of the week.

Re:Self-warming

[Tango42]

Hang on, they have radioactive heaters inside the radiation shielding? I know cosmic rays etc are much worse that a few alpha/beta particles, but isn't that still a little silly?

Re:Self-warming

[JDevers]

If I'm not mistaken, virtually all probes have some sort of radioisotope heater...

Radioactivity is NOT radioactivity when you are considering things like this. Saying the people who don't want nuclear powered rockets should hate this as well or else they are hypocrites is tantamount to saying that the people who don't like oil spills should bitch about how some motor oil ALWAYS stays in the plastic container it is shipped in. Not quite the same problem. Afterall, these things aren't much more radioactive than a Coleman lantern wick or a smoke detector element...

Re:Self-warming

[Waffle Iron]

If they don't mind it, then let's send up those dune buggies with RTG and 18-inch wheels and cover a lot more of Mars.

I don't know, but I'd guess that for rovers of this size and weight, it might be easier to use solar panels than to figure out how to dump all of the waste heat that an RTG would produce. These devices are highly inefficient in terms of converting thermal energy to electrical energy, so they are probably spewing kilowatts of thermal energy at all times, and there's no way to shut them off.

That would pose two immediate problems: with the el-cheapo balloon-based reentry system used, how do you keep the rover from frying when it's wrapped up in its airbag cocoon? The only way would seem to be a heat sink, but that would add significant weight to the spacecraft.

The second problem would be how to radiate all of that heat during operation. The outer planet probes have spindly rods that hold the RTG a few meters away from the main body. A rover couldn't do this and maintain balance, so you would need some kind of closely attached radiator with liquid cooling and/or heat shielding to protect the main unit from the intense heat. Once again, this could end up being heavier than simple solar panels.

Re:Self-warming

[kinnell]

If they're only concerned by radioactive propulsion systems, then I think they're a bunch of hypocrites.

They're probably concerned that nuclear propulsion might irradiate outer space. No, really!

Re:Self-warming

[The Fun Guy]

Radioisotope thermoelectric power units need to be hot enough to allow for electricity to be generated by thermocouples placed between the unit and the heat sink (space). A quick Google search gives 200-500 watts of power generated from multiple interleaved stacks of plutonium-238 or strontium-90, average radioactive source strength of around 50,000 curies, depending on design.

Radioisotope heaters use much less material, as they only need enough heat to keep the warm electronics box above -40F or so. From the Environmental Impact Statement in the Federal Register ([wais.access.gpo.gov][DOCID:fr10de02-54]):

"Each rover would employ two [calibration] instruments that use small quantities of cobalt-57 (not exceeding 350 millicuries) and curium-244 (not exceeding 50 millicuries) as instrument sources. Each rover would have up to 11 RHUs that use plutonium dioxide to provide heat to the electronics and batteries on board the rover. The radioisotope inventory of 11 RHUs would total approximately 365 curies of plutonium."

Nothing you'd like to swallow, but still, much smaller than a radioisotope power unit.

Re:Self-warming

[A55M0NKEY]

Irradiate outer space, who the hell cares right?

There is TONS of information to be learned from the levels of isotopes and natural radiation in our solar system, but spewing radiologically hot stuff out the backs of rockets would ruin that data for all time. On Earth, carbon 14 dating can't be used to date materials from after the first atomic bomb blast. Who knows what information you'd ruin with a dirty rocket?

Re:Self-warming

[A55M0NKEY]

Someone said a while back that a house made of aerogel would need only a candle to heat it. Surely surrounding the electronics in and Aerogel box would trap so much heat they would melt themselves. I understand that they have to shut the electronics off at night, so the radioisotope heaters make sence, but why all the other heaters?

Re:Self-warming

[useosx]

Better yet, they should just overclock it at night. Nothing like a 20MHz PPC overclocked to, uh, 100MHz. Can they solder pins from 600,000 miles away?

Re:Self-warming

[WormholeFiend]

hey dont make fun of Athlons that's how I keep my apartment warm during the frigid canadian winter

Re:Self-warming

[Anonymous Coward]

and there is more radioactivity in your backyard then a RTG :-\

Re:Self-warming

[Anonymous Coward]

I do not believe that Coleman lantern wicks
are radio active anymore - they used to be.

Re:Self-warming

[Anonymous Coward]

...heated by a combination of electric heaters, eight radioisotope heater units as well as the natural warmth from the electronics themselves.

we'll need to send quite a few more of those radioisotope heater things up with the astronauts when we go there in 15 years to make sure nobody freezes to death... :-p

Re:Self-warming

[Tablizer]

I don't know, but I'd guess that for rovers of this size and weight, it might be easier to use solar panels than to figure out how to dump all of the waste heat that an RTG would produce.

Maybe drag it around like a big, heavy tail? :-)

Re:Self-warming

[HeghmoH]

He's not talking about people opposed to Orion. He's talking about the people who protested Cassini, both the launch and the Earth flyby gravity-assist maneuver. It had a little radioisotope generator onboard which is what got everybody so up in arms. But nobody pays attention to it on a Mars probe. Is it suddenly not evil if it's not generating electricity?

Re:Self-warming

[Anonymous Coward]

Who cares what the environmentalists think? I mean, the Viking probes had RTG's and that's why they functioned for so long. When they say they don't like stuff like that, it's aprime example of the victory of ignorance over intelligence.

Re:Self-warming

[Rich0]

I believe Viking used RTG's...

Obviously there must be some way to handle it.

If you can dissipate 50W of heat from a 1cm^2 computer chip, I'm sure you can dissipate a few kilowatts from a MUCH larger RTG. Add fans or something like that - remeber - once you go with an RTG your need to be energy-efficient drops quite a bit. You can probably use some waste heat to replace the various heaters that keep the thing from freezing up at night...

Re:Self-warming

[baryon351]

I've been wondering about the effects of these machines on mars itself. I'm sure the energy they're putting out is pretty minimal, and isn't going to do much, and I'd hazard a guess there's more radioactive spots around than the rover carries. But what about life that gets through? Are there any little organic compounds sitting on the rovers/landers/airbags that could make a big difference? Seeing how some lichens survive at phenomenally low temperatures, could a few little spores stuck to parts of the airbags eventually build up a colony, then move to nearby rocks, and from there on... who knows where else?

Not that I'm quite worried in 10 years we'll be attacked by lichen-looking martians, but I'm curious

Re:Self-warming

[AKAImBatman]

Considering that you're sitting next to a massive nuclear furnace that's putting out 5 hundred billion times anything even the largest nuclear rocket could put out, I'd say you're full of it.

The sun has already done radiological damage. Deal with it.

Re:Self-warming

[Endive4Ever]

My brother in law has a 'Rocket Simulator' game that he runs on his fan-ridden Athlon machine. I asked him one time about the noise the game was making. He replied 'what do you mean? the speakers are unplugged.'

Re:Self-warming

[Endive4Ever]

Some of my PPC Macintoshes have a jumper that's taped over 'Warranty Void if Removed.' I think it's the Beige G3 boxes. Is that a clock multiplier jumper?

(would I void out the 'warranty' on the Beige G3 boxes, which I paid $15 apiece for?)

Re:Self-warming

[A55M0NKEY]

It's not putting out Uranuim, Thorium, Plutonium or other heavy elements. Those are produced in supernovae, and can tell a story that would be erased by a nuclear rocket.

Re:Self-warming

[AKAImBatman]

It's not putting out Uranuim, Thorium, Plutonium or other heavy elements.

You do know how plentiful those are in our solar system, don't you? We have entire asteroids made of the stuff. Anything exhausted by nuclear rockets would be much "hotter" and easy to identify. Not to mention that there wouldn't be much of it. Some engines (e.g. GCNR) don't even exhaust any of those radioisotopes.

Re:Self-warming

[shotfeel]

I was just going to say,

"In Soviet Russia, space irradiates you."

but it made too much sense.

Re:Self-warming

[A55M0NKEY]

You do know how plentiful Carbon 14 is on earth don't you? But the amount relative to Carbon 12 is what is important for dating which is why when the amount changes you can no longer use the radioactive decay rate of Carbon 14 to date things.

Of course the exhaust of a nuke rocket would be hotter than the other stuff in the solar system, but think about what 'hotter' means: Hot means 'rich in *highly* unstable isotopes'. Those short lived isotopes with half lives of only a few years are not the problem - they do not exist naturally because they would have completely decaysed by now, but other unstable isotopes with long half lives that might exist naturally in the solar system would also be emmitted. There is no way to distinguish a natural U235 atom from one emmitted by a nuclear rocket. They are precisely identical.

Re:Self-warming

[Waffle Iron]

I believe Viking used RTG's...

Some quick googling revealed that the Vikings RTGs generated 40 watts of power and weighed 26 kg. The Spirit's solar cells generate up to a peak of 140 watts of power. The Viking probes were not mobile, and they weighed about 3X as much as the rovers without rocket fuel. Moreover, the Vikings were not a "low-cost" mission, and they used fuel-hungry retro rockets to do a soft touchdown on the surface (further increasing launch weight). The retro rockets meant that they did not need to be wrapped up in a bunch of balloons for extended periods of time.

It seems that the RTGs did not provide more energy or enable less launch weight than the solar cells. In fact, if high-performance solar cells had been available in the 1970s, perhaps the vikings wouldn't have used RTGs either.

Re:Self-warming

[AKAImBatman]

There is no way to distinguish a natural U235 atom from one emmitted by a nuclear rocket. They are precisely identical.

1. It simply isn't enough to make a difference to the ecology of the solar system. (It takes DAYS for light from the sun to reach the outer parts of the solar system.)

2. Some engines don't even emit Uranium. Most of the engines currently developed fall into this category.

3. Radioisotopes can be tracked by the engine exhaust vectors. Any minute amounts in the exhaust vector can be discounted.

4. You're changing things by existing. Are we supposed to stop breathing because we're having a butterfly effect on the weather? Scientists are going to have to sort things out as best as they can.

You do know how plentiful Carbon 14 is on earth don't you? But the amount relative to Carbon 12 is what is important for dating which is why when the amount changes you can no longer use the radioactive decay rate of Carbon 14 to date things.


That has ZERO to do with space engines. The extra radiation pulses from nuclear blasts increased the rate of decay due to spontaneous fission. As a result, an already highly inaccurate and unstable dating method was made that much more inaccurate. It's not like the prehistoric uranium fission fires screwed that up at all. Nooo.

So, if a little bitty nuclear explosion screws up carbon dating inside the radiation belts, who in the world would try to use it in space, where every solar flare changes the rate of decay?!

Re:Self-warming

[el-spectre]

Yup, it's entirely possible that some anerobic microbes would survive the trip (perhaps frozen and then thawed upon landing). As I recall, we've retrieved satellites from low orbit (like 400 miles up) that got irradiated fairly badly, and found living bacteria on them.

Re:Self-warming

[Cujo]

Meters away isn't necessary if you do the thermal design right. This spacecraft for example, will hold its RTG fairly close. We're not talking about a huge amount of heat, and thermal isolation isn't risky technology.

I'm confident that the waste heat from an RTG is anything but a showstopper - it would be especially beneficial in the cold Martian nights.

Re:Self-warming

[DerekLyons]

If obsessed environmentalists don't like NASA sending up probes with any radioactive material ('it might blow up, ohh..'), then how did this little tidbit get by them?

Mainly because the environmentalists are not that obsessed, and because the tiny heaters don't get as much talked about in advance as the much larger generator units. IIRC the heaters don't require the formal/legal process, (enviromental impact statements, safety planning, etc...), that the generators, so that also helps slip them beneath the radar.

Re:Self-warming

[Detritus]

Pu238 is an alpha emitter, which requires minimal shielding.

Re:Self-warming

[KewlJedi]

NASA is waiting until we send humans to crack out the athlons. How else are we going to melt the polar caps for water?

Re:Self-warming

[WhiteWolf666]

The EXTREMELY weird thing---is the estimated lifetime of Spirit&Opportunities solar panels.

3 months

Seriously.

Apparently, dust will settle on the solar panels, and in 3 months this will reduce their effectiveness enough such that the rovers will not be able to operate....

Bummer, eh?

I'm hoping it doesn't work out that way....Why couldn't they push some kind of cleaning brush on the rovers?

An RTG would not have burned out in 3 months. I don't remember off hand, but weren't the Pioneer probes RTG powered? I think they transmitted data for decades.

Re:Self-warming

[Anonymous Coward]

that's planned for 2009 mars science laboratory
(#2)

If you've got money to burn, sure

[Powercntrl]

But I'd take a Linksys over a hacked Mars Rover anyday... Billions cheaper, ya know.

Re:If you've got money to burn, sure

[FroMan]

Wouldn't that be hunderds of millions cheaper? Cost for this was ~$850,000,000 I thought, and that got us two of the litte buggers.

Re:If you've got money to burn, sure

[Endive4Ever]

No. A lot of that money was spent the way it is buying vintage boat-anchor hardware on eBay: it went for 'shipping' costs.

Radiation Shielding

[kyknos.org]

How is it done? Some external armor, or even insides of the chip are different?

---

Re:Radiation Shielding

[pi eater]

A modded alienware case perhaps?

geeky stuff.. offensive stuff!

Re:Radiation Shielding

[the real darkskye]

The CPU is fabricated to withstand the radiation, a brief summary can be found here or by googling

Re:Radiation Shielding

[PhuCknuT]

Both.

They have extra shielding on the outside, and the electronics on the inside are designed to disipate sudden charges created by radiation hits.

Re:Radiation Shielding

[shawnce]

A lot is done with extra shielding but often radiation hardened chips use larger feature sizes then modern equivalents. The larger the features the more resilient they can be to particle/energy hits. Basically they are harder to damage permanently.

Re:Radiation Shielding

[stevesliva]

There's more to it than just feature size. You need the right substrate doping and circuit structures than can dissipate the charge introduced by a particle without having it course through the parts of the chip most susceptible to being fried, or perhaps just having their logic state flipped.

Re:Radiation Shielding

[gl4ss]

you know, instead of who*cough* catching some free karma you could have read the article..

*RAD6000 microprocessors are radiation-hardened versions of the PowerPC chips that powered Macintosh computers in the early 1990s, with 128 megabytes of random access memory (RAM) and capable of carrying out about 20 million instructions per second. A critical feature of the spaceworthy chips -- developed jointly by BAE systems, JPL and the Air Force Research Laboratory -- is the radiation shielding, which uses a series of resistors and capacitors to ground harmful radiation before it can damage onboard electronics.*

Re:Radiation Shielding

[shawnce]

True. I didn't say otherwise, was just pointing out one common aspect.

Re:Radiation Shielding

[Anonymous Coward]

On the FPGA's I understand that the Logic is replicated multiple times, and checked against itself. And some more process magic.

Re:Radiation Shielding

[sirsex]

Larger feature size (especially thicker gates) and special processing can help against radiation that would cause permenant damage. For lesser events, usually only a bit will get flipped or noise is introduce onto a node. So you can build three of every critical circuit, and vote on the outputs. Majority wins.

Re:Radiation Shielding

[Naito]

besides the larger features, they also apparently have extra grounding points compared to standard processors.

Re:Radiation Shielding

[UrgleHoth]

From the article:
Before AFRL researchers stepped in a few years ago, Gordon explained, the Defense Department and NASA paid from $50 million to $100 million for each processor in development and manufacturing costs. Now, after AFRL involvement, the price of a typical processing module dropped to between $500 thousand and $2 million and is available as off-the-shelf hardware.

To me, that is an incredible reduction in cost. Makes me wonder where the money was going, duplication of efforts between development teams (time, materials, facilities)? high beaurocratic overhead?

Re:Radiation Shielding

[bash_jeremy]

Constant bombardment by radiation, however, generates unwanted electrical charges inside transistors, building to the point that the transistor, or switch, can no longer control the electron flow

Yeah, I have that problem too sometimes.

Redundency Check?

[shlomo]

I read somewhere that the shuttle spacecraft, has 6 computers for redundancy checks.

Your Average plane have a triple backup system, I spoke to some engineer and he said preflight checks are usally just making sure two of the systems are still working

you'd think they could at least send up some more hardware with these little critters. The extra weight would pan out, when things go bad...case in point see what they are dealing with now :)

Re:Redundency Check?

[pi eater]

Even I have a backup computer just in case mine breaks down and I can't play games.

The stakes are too high.

geeky stuff.. offensive stuff!

Re:Redundency Check?

[snake_dad]

That way of thinking would make the cost of robotic space exploration approach that of human space exploration. Plus, the rover will not crash with loss of life in case of a minor computer failure. There is a much bigger margin for troubleshooting "in the field" than with aircraft or manned spacecraft.

Ofcourse NASA did implement a measure of redundancy by sending two rovers instead of just one.

Re:Redundency Check?

[Mycroft_514]

The original shuttle systems had 5, not 6 computers. Later that number was reduced.

Re:Redundency Check?

[Anonymous Coward]

well the $800+ million wasted could have saved quite a few lives right here on earth.

Re:Redundency Check?

[vofka]

If I recall correctly, the Shuttle has 5 GPC's (General Purpose Computers), three of which are "online" at any one time.

The online GPC's each carry out the same set of calculations (potentially each uses code designed to do the same thing, but written by different programmers), and they compare each others results. If any single GPC is considered to be too far wrong, the offline GPC's submit their answers. The three GPC's that are in closest agreement then become the new online GPC's, and the remaining two go offline. The GPC's can reboot themselves if they are too far out of whack, if they fail in one of the "results elections", and of course when they are told to do so by the crew.

Also, whenever a GPC is sent offline by one of the others, a specific caution indicator (and potentially the master caution indicator and klaxon) is activated, and the relevant error codes are shown on one of the forward CRT's. The error codes, along with other information such as the currently running program and the current mission phase, determine the crew's actions. Actions can be as simple as disabling the master caution klaxon for the current alert, all the way to hand-checking certain results and manual GPC restarts.

This is all from memory (from about 5 years back), so some of this may have changed recently, particularly on Atlantis with the "glass cockpit" upgrade that happened 18 months or so ago, but the general gist should be about right (and I'm sure I'll soon know if it isn't!!)

Re:Redundency Check?

[jpellino]

Before they went to the new glass cockpit (going from about 30 old displays to about a dozen digital ones) , the Shuttles had five identical general purpose computers (IBM or Honeywell, IIRC), four worked in parallel to compare the results of any given operation, with the fifth as a tie-breaker.
One of those 30 was a mechanical cue-ball, too - Tony England tells the story of being in the Shuttle simulator one day and being interrupted because one of the cue-balls on the flying shuttles went out, and they had only the one in the sim to rip out and swap into a working shuttle, presumably they'd fix the dead one and put it back in the sim when they could.

Yikes.

Re:Redundency Check?

[GileadGreene]

you'd think they could at least send up some more hardware with these little critters. The extra weight would pan out, when things go bad...case in point see what they are dealing with now

There's a limit to how much mass we can send to Mars, driven by the throw capability of our launch vehicles. MER was launched on a cheaper (read less throw capability) launch vehicle (the Delta II), and was pretty tight on mass from the get-go. There simply wasn't room for an extra computer. In fact there was barely any room for redundancy at all, so most of the design is single-string. The argument was that they had "system-level redundancy" (i.e. two rovers), but that only works if the problem isn't a design flaw...

Re:Redundency Check?

[Naito]

not only is the 5th a tie breaker, it was apparently designed and programmed by a completely separate company from the other 4, to make sure that there was no possible way that a common fault in the 4 GPCs would affect the 5th one if it came to that.

Re:Redundency Check?

[DunbarTheInept]

but that only works if the problem isn't a design flaw...

If the problem is a system-level design flaw, then redundnacy of that system in one craft wouldn't buy you anything *either*, so I fail to see the validity of this complaint. ("Hey, it seems our software managing the flash filesystem has a flaw - But hey lucky for us we're running three copies of it and getting the same error three times instead of just one. Isn't that a wonderful improvement?")

Re:Redundency Check?

[MikeyToo]

Both of these previous comments are basically correct. There are five GPCs (General Purpose Computers) in the original configuration of the STS. I've heard them compared to a 286 before. During a mission, three run the primary flight software, one is a hot spare and one has the BFS (Backup Flight System) loaded. The BFS is a no-frills ascent/descent/abort program.

During normal operations, the three primary computers perform their calculations and compare the results. If one of the computers disagrees, it is taken offline and the hot spare is brought online. The crew receives a master caution alarm and the commander is tasked to handle the problem since the GPC panel is over his head. If a second failure occurs then the system dumps to the BFS and you start looking at abort situations. They computed the probability of a double failure at four in a billion. However, YMMV.

If you want to know the whole story on the computers go to Computers in Spaceflight Chapter 4.

I read an article about the the flight software. The reliability they achieve is phenomenal. Orders of magnitude over what any of us will ever see unfortunately.

Re:Redundency Check?

[Linker3000]

I look forward to my next fault-tolerant PC; which will comprise an Intel, AMD and Motorola processor. One CPU will run Windows, one Linux and the last one a Mac OS. I will fire up my WP front end, which will spawn copies of Word, Star Office and Nisus Writer Express (Mac). Boy my documents will be soooooooooo perfect!

Re:Redundency Check?

[Anonymous Coward]

Boy my documents will be soooooooooo perfect!

Why not just use WordPerfect?

Re:Redundency Check?

[Detritus]

The hardware is identical. The software (Backup Flight System) was written by Rockwell. The primary computers run software (Primary Avionics Software System) written by IBM.

Re:Redundency Check?

[GileadGreene]

What you say is true of software. However, hardware is a different ballgame. For example, if the design flaw involved specifying hardware items that turn out to have inadequate lifetime then per-rover redundancy may mitigate the problem, but system-level redundancy won't (e.g. redundant inertial measurement units might allow the guidance system to operate after one IMU has failed, by switching to the previously unused backup - one IMU per rover wouldn't solve the problem).

Re:Redundency Check?

[DunbarTheInept]

In one case, you have one rover with two of the devices, and in the other you have two rovers with one device each. In both cases, in order for a failure of the device to cause the end of the mission, you have to have BOTH instances of the device failing. So long as one works, the mission can continue using a single rover. The difference is in what happens when *BOTH* work - in one case your mission continues with one single rover, and in the other it continues with two.

Re:Redundency Check?

[GileadGreene]

I disagree. In one case you have a rover with one IMU that works for half of the EDL sequence, and a second that works for the other half. In the other case you have two rovers with IMUs that only work for half of EDL, and both make a big black mark on the surface of Mars.

Re:Redundency Check?

[DunbarTheInept]

In one case you have a rover with one IMU that works for half of the EDL sequence, and a second that works for the other half.

That's a problem that has nothing to do with being a systematic failure in the design, but is instead a non-systematic glitch (otherwise it would be the same exact failure in the same exact place in both rovers). You claimed you were talking about sytematic failures, not random ones.

Re:But The Question is..

[millahtime]

Something like this can't run GNU/Linux. There are real time implications. Linux may be fast but not as fast as a real time OS. Linux is in terms of milliseconds while a true real time OS is in terms of hundreds of microseconds. For this it does make a difference.

Is there a Linux is space yet?

[Anonymous Coward]

I was just wondering if any device is/was running linux in space?

Re:Is there a Linux is space yet?

[vasqzr]

NASA has been evaluating it...

On a related note, they use NVIDIA/Linux setups to view the Mars Rover data.

Article

20mhz?

[pi eater]

Just goes to show how much can be accomplished with such a slow processor.

Impressive!

geeky stuff.. offensive stuff!

Re:20mhz?

[Mysticalfruit]

More imporantly.... it shows how little can be achieved with copious ammounts of processing power!

Re:20mhz?

[c_oflynn]

Remember this *probably* isn't running an OS like we would think of it. I'd wager that everything is designed specifically for the mission, so the speed isn't a real issue. I mean its not like NASA is re-compiling its own kernel or something. And time isn't that imporant - everything HAS to go slow pretty much.

Re:20mhz?

[Anonymous Coward]

It is a realtime operating system it is running. Time is everything.

Just image...

[F4Codec]

... A beowulf cluster of those - does 2 make for a cluster, and did they bring a big enough piece of cable, or do you suppose Wifi would work.

What would Beagle have run had it run?

Re:Just image...

[Anonymous Coward]

For searching...

Price

[sparklingfruit]

They cost way too much to be a router. $150-200k according to howstuffworks.com. Mainly due to radiation shielding or something.

Nasa should've gone to transmeta.

A nice Firewall?

[blorg]

"I wonder if I could make a nice firewall with one of these for my home network..."

You could, but the latency would be a bitch.

Re:A nice Firewall?

[noselasd]

In the vxWorks version I have here, I see no protocol/packet filtering options.

Re:A nice Firewall?

[Skater]

Not to mention the gov't tracking you down for rooting one of the rovers...

--RJ

Re:A nice Firewall?

[Anonymous Coward]

Funny.

Units, units, units!!!

[Cutriss]

Basically, they're radiation-shielded, 20MHz PowerPC machines wirh 128Mb RAM and 256Mb of flash memory, running VxWorks.

Mb = Megabits
MB = Megabytes.

The article writes out megabytes, so MB should be used, not Mb!

Re:Units, units, units!!!

[xk]

Make that MiB.

Re:Units, units, units!!!

[Mycroft_514]

I glad enough when they don't say Megs. You don't add the plural "s" to an abbreviation.

Re:Units, units, units!!!

[Anonymous Coward]

I thought about using mibbibytes instead, but really, that's hardly something that we'd get confused about, and the difference is negligible.

Megabits versus megabytes, however, has an order of magnitude of difference.

Re:Units, units, units!!!

[sparkie]

I actually worked with a kid who went ballistic on some guy for saying 'megs' ... it was in a school so he took him to the library and proved definitively that the word is 'meg' not 'megs' ... ever since then I cringe whenever I see 'megs'

Re:Units, units, units!!!

[pixelbeat]

Flash uses the same units as disks (powers of 10).
So it should be 64MiB and 256MB.
I've some more info here

Re:Units, units, units!!!

[Anonymous Coward]

Sure you do. Mod, .. I mean, Mods, you know what to do.

Re:Units, units, units!!!

[asuffield]

Saying it won't make it so. I've seen at least three contradictory "definitions".

Re:Units, units, units!!!

[iantri]

And would those be Binary megabytes or Decimal megabytes?

It's a battle you just can't win.. I personally perfer "Mbit" for megabits and the conventional binary megabyte.. but I've been known to write it MB, Mb, or mb.

In general, I think it's a bad idea to make such a distinction based on the capitalization of one letter.. it's too easy to screw up and too difficult for most people to understand..

Re:Units, units, units!!!

[hungsolo]

Looks like they got away with unit conversion failure for a while (unlike Polar Lander). That explains why they started deleting file like mad.

Re:Units, units, units!!!

[Anonymous Coward]

In general, I think it's a bad idea to make such a distinction based on the capitalization of one letter.. it's too easy to screw up and too difficult for most people to understand..

Well teach them the difference. They need it to understand mathematical formulas, the metric system, the unix file system and a lot of programming languages.

Well...

[Anonymous Coward]

Basically, they're radiation-shielded, 20MHz PowerPC machines wirh 128Mb RAM and 256Mb of flash memory,

I don't think anyone would want a Beowulf cluster of those things.

Should it have some redundancy?

[mikesmind]

"It's quite unusual to have a single computer for the whole mission," Scuderi said, adding that many missions tend to have redundant systems as a guard against failure.
Now, while having two rovers is a form of redundancy, wouldn't it be wise to have some redundancy on each individual rover? I understand that there are concerns like weight and budget, but wouldn't some redundancy be a good form or risk management?

Re:Should it have some redundancy?

[pi eater]

Yeah.. or build a computer with two or more of everything, including the bus lines.. redundant ram, etc..

Basically a computer with redundant parts that would take over if something went wrong.

geeky stuff.. offensive stuff!

Re:Should it have some redundancy?

[scallions]

Sounds like a job for the Daystrom M5.

Not a firewall,

[Morologous]

but you could probably make a pretty nice remote controlled rover to explore the outer reaches of your back yard.

Re:Not a firewall,

[the real darkskye]

Would a PGP Signature do for ID?
Could you send one out for smokes and beer?

Beer with out leaving the arm chair, BONUS!

Ouch

[savagedome]

In addition to VxWorks' reliability, the system allows users to add software patches -- such as a glitch fix or upgrade -- without interruption while a mission is in flight. "We've always had that [feature] so you don't have to shut down, reload and restart after every patch," Blackman said, adding that some commercial desktop systems require users to reboot their computers after a patch

The bold emphasis is mine but that is a big Ouch for Microsoft.


RAD6000 microprocessors are radiation-hardened versions of the PowerPC chips that powered Macintosh computers in the early 1990s
Shouldn't Apple be using this in their commercials somehow to further boost their reliability. I am sure the PR market can put it in a way that non-geeks watching tv can relate, right?

Re:Ouch

[Anonymous Coward]

The bold emphasis is mine but that is a big Ouch for Microsoft.

Way to karma whore/groupthink, dude!

Re:Ouch

[Quasar1999]

AH dude... Microsoft is totally different... VxWorks is based on the fact that all memory is linearly accessable.. there is no memory protection... that works great when all your apps are written in-house, and you know the timer ain't gonna trash the motor controller...

But try that on an OS for a desktop system, and your email program just may blow up your paint program... (remember Windows 3.1's stability? Make that 10x worse)... You can't use VxWorks for the desktop as Windows is used today... it needs a lot of protection... The ease of upgrading is due to the lack of protection...

Re:Ouch

[natelr]

Concidering one of the rovers has had some serious issues may not help Apple to convince people to buy thier boxs.

Re:Ouch

[confused one]

The ease of upgrading is due to the lack of protection...

Nothing of the sort. The ease of upgrading is because you can upload a new version of the code, shut down the old process and start the new one without a reboot/reset. Typical Unix like stuff.

Re:Ouch

[maroberts]

Its a real time OS - people often replace the default memory management by a hand written one which meets their own needs, including possibly protection. IIRC, the last time we replaced malloc by routines which were faster, detected leaks and made efforts to avoid memory fragmentation.

Re:Ouch

[Quasar1999]

Nothing of the sort. The ease of upgrading is because you can upload a new version of the code, shut down the old process and start the new one without a reboot/reset. Typical Unix like stuff.

You are probably more correct than I, as that is a method of doing a patch that always works. I have worked on a system that was upgraded at runtime without killing any processes, we simply augmented our flash while everything was running, then when that succeeded, we changed a table in RAM to replace our interrupt handling code to point to the new flash region. This happend with no downtime... I guess it's a specific case, but it is doable that way... And memory protection would have made it impossible without stopping a task and respawning it.

Re:Ouch

[dasmegabyte]

1) When I perform a patch on my Macintosh, I have to reboot as well. Is this a big Ouch for Apple and BSD? I often reboot after patching Linux installs...because while I could just shut down every reliant system manually (something which would be MADNESS on a commercial system), I want to be sure that everything in the system is going to come up in the right order. I've emerge'd packages and had that prevent PAM from starting after an outtage...better to be safe, right?

2) Yes. Apple should mention in their commercials that processors use in their computers 14 years ago were installed on a NASA rover that had a computer failure. That would be marketting gold.

Re:Ouch

[Anonymous Coward]

(remember Windows 3.1's stability? Make that 10x worse)

I remember MacOS 9.x and earlier's stability. About the same thing.

Re:Ouch

[Anonymous Coward]

Actually - it may not be ideal but you can get away with no memory protection on a "desktop" OS if the developers are all really really good. The old Amiga OS did not have memory protection. On my Amiga system I used to have Lightwave running (rendering) in the background for weeks at a time while running other foreground apps such as an IRC client, word processors, etc. I never really had problems with this - the apps were obviously well behaved.

Again, I'm not saying this is ideal, just that it can (and was) done and it did work! These days everyone has just been conditioned to accept poorly written software.

Re:hummm

[Zog The Undeniable]

What about a solar-powered Sunscreen firewall running on Sun Solaris?

Nah...that would be silly.

Flying VxWorks to Mars

[Hiroto. S]

I googled across following presentation with a little more details.

Flying VxWorks to Mars

If you're really interested in this type of HW

[Morologous]

In a machine similar to this, eBay for IBM RS/6000 43P models. They're uni-processor PowerPC devices 250-332Mhz.

They do make nice firewalls for distributed sites (actually currently used in production at the largest government entitlement program).

Re:I hope the flash memory was not commodity hardw

[Rootbear]

There is very little on the Rovers that is "commodity" in any sense. The CCD image sensors, the computers, everything, is all custom made. Everything has to be made to withstand the rigors of flight and the harsh environments of space and Mars. The CPU does not have a backup, which is a bit unusual for NASA (I'm a contractor at NASA/Goddard, but not involved in any flight missions). However, the particular computer used on the rovers (the RAD6000) has a very good record. There are something like 150 in use on various spacecraft and they've all worked very well.

And the flash memory has probably not failed. It seems to have been a software problem, not hardware.

Rootbear

Who else ...

[Muad'Dave]

...read the URL to VxWorks as WinDriver.com instead of WindRiver.com?

Re:Who else ...

[Omni-Cognate]

Me, for one. They should hyphenate that or something.

Internet conneciton

[vpscolo]

And it can still send back at 128 kbits/sec which is faster than my connection can managed. Just waiting for it to start getting spam advertising pr0n and viagra.

Spirit Rover: Staying up longer and harder

Rus

Re:Internet conneciton

[TALlama]

What do you think is filling up all that flash memory?

Re:Internet conneciton

[DrewBeavis]

Mars Rover used as Spam Relay!!

Just some satire that fit...

Re:Internet conneciton

[Anonymous Coward]

It can only send 128 kbps to the Mars Global Surveyor and Mars Odyssey satellites for very short times (around 10 minutes). At best, they can relay a little less than 10 MB (megabytes) of data during one overpass of a satellite.

The main direct-to-Earth communications device, the High Gain Antenna, can send at bit rates only up to about 11 kbps. That can be operated for longer times, so it's still quite useful.

The backup omnidirectional antenna (Low Gain Antenna) transmits at up to 0.1 kbps.

memory overflow caused Spirit shutdown?

[peter303]

A leading hypothesis is that flash memory overflow caused Spirit to be shut down for two weeks. Either it was a failure in the memory chips or OS software garbage collection. They are purging and patching now. A few days of testing and perhaps Spirit is active again.

The lockup happend just as they were going to drill into the rock they've been sitting in front of for nine days. Perhaps there was drill issue too. When the rover memory crashed, it tried to reboot its computer at least a hundred times.

Re:memory overflow caused Spirit shutdown?

[britt]

It wasn't an overflow. They ran out of inodes. They had lots of space, but no more inodes, which is why deleting files should fix things up.

The first intergalactice filesystem bug

Re:memory overflow caused Spirit shutdown?

[Anonymous Coward]

A leading hypothesis is that flash memory overflow caused Spirit to be shut down for two weeks.

Yeah - everything was going fine until the rover contacted Microsoft for its weekly update and choked on all the security patches.

Re:memory overflow caused Spirit shutdown?

[Anonymous Coward]

*** PEDANTIC ASSHOLE ALERT ***

Perhaps interplanetary filesystem bug. But not interstellar or intergalactic.

Re:memory overflow caused Spirit shutdown?

[LordNimon]

It can't be intergalactic if it's in the same solar system. It could be intragalactic, but a better word is interplanetary.

Re:memory overflow caused Spirit shutdown?

[princewally]

Except for the fact that it's a Macintosh.

Re:memory overflow caused Spirit shutdown?

[Orion442]

Which explains a lot more

Re:memory overflow caused Spirit shutdown?

[paganizer]

Don't even think it.
It's a custom chip developed by IBM, based on the PPC architecture.
Well, maybe an Amiga.

Re:memory overflow caused Spirit shutdown?

[KewlPC]

It isn't a Macintosh. The article was wrong when it said that the RAD6000 is a radiaton-hardened version of the same CPU that used to be used in Macs.

While the RAD6000 is a PowerPC processor, it's based off of IBM's RS/6000 CPUs. Macs, on the other hand, used Motorola CPUs until the G5.

Re:memory overflow caused Spirit shutdown?

[princewally]

Ok. I was wrong. That's what I get for RTFA.

Would everybody please quit refuting me?

driving the rover

[QEDog]

The question that we all want to know is, how do they drive it? I imagine that they have 3 identical car cockpits, with steering wheel, brakes and gas pedal. 3 different engineers drive it, voting on their actions for redundancy. If one of them dies, or goes to the bathroom, or simply starts honking like a mad man, still the other 2 could respond.

Re:Martian Geeks Will Overclock It

[zakath]

Last time I checked MS didn't make processors...maybe you meant Intel?

Re:hummm

[gl4ss]

well iirc Solaris was set on a planet that was mostly a gigantic ocean?

Not 20Mhz

[kuyttendaele]

But 20 MIPS

Re:But The Question is..

[Anonymous Coward]

you seem to assume that real time OS are faster than non real time Oses, that's usually not the case. the difference between the two is that a RTOS has a guarenteed response time, not a faster one...

it is not uncommon on so hard real time system to disable processor cache, it makes the processor slower, but the response time to an interrupt is easier to calculate. In RTOS interrupt latency must be PROVEN not to be longer than a constraint.

Something strange...

[J. Charles Holt]

I noticed something odd in the latest shots from the Rover. Just on the horizon:

http://idisk.mac.com/charlesholt/public/DuckDodg er s.jpg

What does transmeta have to do with it?

[rdunnell]

A transmeta processor would still have to be properly shielded against radiation and other environmental hazards.

Hmmm...

[southpolesammy]

I wonder if I could make a nice firewall with one of these for my home network..."

I'm guessing due to the latency, you wouldn't even need to setup any rules.

Beagle-2

[TheSurfer]

An interview with one of the Beagle-2 software developers can be found here: http://linuxdevices.com/articles/AT7460495111.html

Re:Beagle-2

[Anonymous Coward]

Yeah Linux worked out REAL well for them. :P

Save HST! Sell Sojourner knock-offs

[bhima]

What NASA should do is to hire a Taiwanese firm to build inexpensive knock-offs of Sojourner. They already have the design, I'm sure a few bright minds could cut the chassis price down significantly; after all we don't need all the exotic materials. I'm sure IBM still makes a PPC variant that would make a new cheap board layout easy. As far as the OS: of course we don't need VxWorks (Nor could the project afford it) we have NetBSD!

The profits from Slashdot alone could extend the life of HST or launch the James Web Space Telescope early.

I thought about the current rovers, but I think they are a bit large to be successful!

Re:Save HST! Sell Sojourner knock-offs

[jCaT]

Yeah, and we could use duct tape instead of titanium screws to hold the whole thing together.. it'll be great! The cost savings will be HUGE!

You know, there are some times when it is better to not be a cheap bastard, and this is one of them. When it's your $29 dvd player from wal mart that quits working, you can just return it and get a different one... you can't exactly do that on mars.

Re:Save HST! Sell Sojourner knock-offs

[CXI]

For a second, like the other reply, I thought you were talking about making a fleet of knock-offs to send to Mars. However, an RC car style robot kit would be pretty cool. I did see that NASA is releasing a lego kit of the current rovers. I would think a reduced scale MER would also be nice to have, especially if you could get a basic big track level unit cheap and upgrade it with cameras, sensors, wireless and a navigation system as add-ons.

Re:But The Question is..

[Anonymous Coward]

I do stuff for space program and they always run RTOS (VxWorks or OSE). It's really small and super reliable. Trust me they aren't going to be switching to your run of the mill linux distro.

Good thinking...

[artemis67]

If we had sent a Wintel PC, it might have kicked off an interstellar conflict with the Martians...

Re:Good thinking...

[ls-lta]

They were going to, but the Martians just blushed when we asked them if they could help out with the control-alt-delete. It seems that control-alt-delete is an obscene gesture to the Martians. I, on the other hand think, oh, wait...

Mac users..

[JayPee]

You know what's annoying about this story now that it's making the roungs.

Mac users everywhere take this as "Oooohh.. there's a Mac inside of those things!" "There are Macs on Mars!" Bleah.

And before you mod me down, realize that I'm an unrepentant Mac user and an Apple Authorized Service Tech.

Re:Mac users..

[nbahi15]

Don't foget that it also sounds like a TiVo with wheels and a science arm. I can't help but think of that cute little TiVo guy sunbathing on mars.

Re:Mac users..

[Anonymous Coward]

Not really.

It's cool that it's a PowerPC, but this thing is such a distant relative of the ones used in Macs it might as well be a MIPS chip.

Besides, everyone knows that what makes a Mac a Mac is the operating system.

--an unrepentant Mac user

Re:Mac users..

[natelr]

Well that will teach them. Had they installed OS X they wouldent be having flash memory problems! heh.

Re:Mac users..

[Endive4Ever]

It's my understanding that there was an 8085 processor in the earlier Mars Rover from a number of years back.

The 8085 is in my TRS-80 Model 100. "So there's a Model 100 on Mars!" Whoo-hoo.

Of course, there are also probably pieces of wire in the Mars Rover that are the same brand wire as used in my Kenmore Washing Machine. Whoo-hoo, etc. etc.
(ad nauseum)

To take care of Rover's heating problem...

[notetoi]

they should have used an AMD processor. Heck, a couple of AMD processors would be more than enough to raise Mars' temperature to room temperature.

How come they use Motorola/IBM CPU's ??

[Anonymous Coward]

and not Intel or AMD ?

Re:I hope the flash memory was not commodity hardw

[Jarnis]

Their redundancy plan is simple - they *DID* send out two rovers. Even if one craps out, they still get tons of science data from the surviving one.

This is actually the smart way to be redundant - why build huge redundancies into single device when you can just manufacture two and toss 'em both at the job. If you luck out and nothing fails, you get double the scientific data too!

Monopoly issues?

[sammyo]

Shouldn't each rover be required to have a different OS? I mean this must just quash the Martian software industry.

space shuttle uses 1969-vintage ibm 360 computers

[peter303]

The space shuttles run on five AP-101 computers, originally designed in 1969. The started with 32 kilowords of magnetic core memory for radiation protection, since upgraded to semiconductor memory. These computers were chosen due to their success in the Apollo, Skylab, and B52. For science and personal work the astronaut specialists usually bring personal laptops which are thousnds of times more performant.

Re:But The Question is..

[millahtime]

The response time to an interupt is the big thing though. You have to have the garunteed interupt response time. With a real time OS you get that after every iunstruction. With Linux 2.4 that wasn't really there which they remedied in 2.6 but it isn't true real time. For machines like this, planes, space shuttles and cars you will find a real time OS. Would you want your drive by wire response time to be 200 micro seconds or 4000 microseconds. In reality that makes a difference.

Processor is *not* a PowerPC

[GileadGreene]

radiation-shielded, 20MHz PowerPC machines

No, they're not.

The processors in MER are RAD6000's, which are radiation-hardened versions of the RS/6000, the predecessor to the PowerPC (see this for details). The RAD6000's younger brother, the RAD750, is indeed a rad-hardened PowerPC.

As an aside, there is a big difference between a radiation-shielded processor and a radiation-hardened processor. Shielding implies just sticking some kind of rad-absorbent material between the processor and the environment. A rad-hardened processor is actually manufactured in a different way - different gate layout, different design rules, often different materials (Silicon-on-Insulator is popular). These things are done to minimize or prevent the effects of single-event upsets (when a bit is flipped by high-energy particles) and single-event latchups (which basically turn a couple of gates into a glorified short-to-ground). The materials changes may also improve the overall total dose tolerance of the processor. The work required for redesign is one of the reasons that space-qualified rad-hard processors lag the commercial market. The NASA Office of Logic Design has some good papers on space processors available online if you're interested in learning more.

Re:Processor is *not* a PowerPC

[addaon]

The Rad750, btw, is a deeply cool chip. Once it's mature enough to start using for scientific-level stuff, it will be a real revolution in what we can do. One of the limitations with Hubble was that it had so little processing, a full data dump needed to be done for even checking orientation; there was no ability to offload processing to the sat. If the 750, or something similar (not that I know of anything too similar) is up there for our next big telescope, it will make a real difference in the efficiency of how it is used.

Re:Processor is *not* a PowerPC

[Naito]

Hubble's got a rad-hard 486 IIRC

Re:Processor is *not* a PowerPC

[GileadGreene]

Amen to that. Thankfully, many of the missions I know of that are in development are baselining RAD750s. Of course, byt the time some of them launch (2010-2012 timeframe) I'm sure there'll be some newer, deeply cooler chip available. Such is life in the space industry... :-)

Re:Processor is *not* a PowerPC

[Tablizer]

The work required for redesign is one of the reasons that space-qualified rad-hard processors lag the commercial market.

There is interest in such hardware for military applications I hear because of other countries working on EM pulse weapons. Thus, probes may have more CPU options in the future.

Re:Processor is *not* a PowerPC

[asuffield]

The RS/6000 is a line of systems (big fridge-shaped bastards). It has had many different processors over its lifespan, which has been rather long.

There have been four basic chip designs used. First was POWER. Later on came POWER2, which was similar but added more instructions.

At some point around here, PowerPC was created. This is not a chip, but a family, defined in terms of the instruction set. It was loosely based on POWER2, but mutually incompatible; it contained some new stuff and didn't include all of the old stuff.

Next came POWER3, the first RS/6k chip to use the PowerPC instruction set.

Then RS/6k was renamed to pSeries, and POWER4 came out, which is another chip using the PowerPC instruction set.

POWER5 is due out this year.

Which chip are you talking about?

Re:Processor is *not* a PowerPC

[GileadGreene]

AFAIK the RAD6K is derived from the original POWER architecture.

Ask Slashdot?

[snake_dad]

While JPL scientists have expressed confidence they can overcome the problem, possibly in a matter of weeks, the hardware and software manufactures behind the computer system are more than ready to help.

"If they ask, we come," Blackman said, echoing the enthusiasm of BAE officials. "I think when something like this happens, the whole community responds to it."

Maybe they should put up an image of the data, the diagrams of the systemboard, and descriptions of the symptoms, and submit it to Ask Slashdot. I'm sure there are a couple of VxWorks experts here who'd love to take a crack at this :)

Re:Ask Slashdot?

[maroberts]

I'm sure there are a couple of VxWorks experts here who'd love to take a crack at this :)

Although the parent comment is (currently) modded as funny, maybe there's an element of truth in it.

Actually, I'm moderately curious to know if NASA would be willing to have an Open Source controlled craft. It would be interesting to see whether problems suffered by Spirit would be foreseen and overcome. It would also be a test of the "given enough eyeballs, all bugs are shallow" principle.

Re:Ask Slashdot?

[Anonymous Coward]

This is a great idea. If they used GNU/Linux, don't tell me there wouldn't be 2^16 Linux geeks RACING to find the solution first. John/Jane Geek saves the Mars rover! Just imagine the bragging rights, and how nice that would look on a resume...

Re:I hope the flash memory was not commodity hardw

[Duke of URL]

Anonymous sources say it was bad sectors on the flash memory, not a total failure. Copy all data off the flash memory, mark the bad sectors and then copy back.

Had to remove some parts...

[mnemotronic]

It's some Ok hardware, given the source (primative carbon-based species). We removed the antennas & cameras, and turned it into a coffee table. The CPU was slower than most MDAs (martianal data assistants), but provided us with a laugh. Junior ate the instrument package before we could stop him. Naughty grzybfyx!
Thanks.

Sincerely, Marvin & Family

QNX

[Anonymous Coward]

They should have read the vxWorks to QNX porting guide and used a real RTS. http://www.qnx.com/mailings/vxporting/

Re:QNX

[Anonymous Coward]

Or, they could forget about that completely and use LynxOS, along with its Linux binary compatibility layer.

QNX won't publish their real-time response, so they will be in the same boat as with Wind River if they went with that.

No, not PowerPC...

[noselasd]

No. Its a RAD6000 CPU. Not entierly your stock PowerPC chip.
RAD6000

I want some real specs.

[yaff]

Hang MPU and the OS. What's this thing using to communicate back to the earth-men? Are they still using a Motorola 400MHz walkie-talkie as Sojurner did to its base station? Or are they up in the GHz range now to get better gain from tiny dishes? More importantly, what frequency is it squawking so I can run out and build a Pringles can phased-array to hear it. The world wants to know these things!

wow

[Anonymous Coward]

Imagine a beowulf cluster of...

Nevermind.

They used to use sapphire...

[emil]

...as the substrate of the chip, rather than a silicon wafer, so the chip was a "sapphire" chip rather than a silicon chip (although doped silicon could then be used to form transistors, as could Gallium Arsenide or Germanium, through the regular lithographic process).

This is the classic "Silicon On Insulator." IBM has a process of embedding a layer of glass beneath the surface of a standard silicon wafer, allowing SOI using silicon substrates. This and their work with copper set them apart from the other large silicon transisitor foundries (TSMC, Intel, etc.).

The processors on the rovers are probably SOI, but I don't know which process is used.

windriver.com

[hefa]

Was is only me who read windriver.com as Win Driver (not Wind River) dot com? Got freaked out there for a while... phew...

Re:But The Question is..

[BSD Yoda]

Actually, I read on Slashdot that VxWorks is dying.....

Login Shell on Rovers

[Anonymous Coward]

It looks like from this presentation that the rovers have a "shell" the scientists can "log into." A fifty million mile telnet. Very cool.
http://avatar.cs.colorado.edu/~siewerts/mar spath/j pl/sld020.htm

Re:I hope the flash memory was not commodity hardw

[RaboKrabekian]

First rule of government spending:

Why build one when you can build two at twice the price?

(And yes, before some ass starts going off that it's not actually twice the price because of R&D and whatnot overhead...I'm aware of that. It's a joke).

That is one pricey firewall!

[Anonymous Coward]

Believe me, if it is running VxWorks, you can't afford it.

RAD6000s seem closest to PPC601s

[questamor]

One chart I've seen of IBM's POWER chips and their derivatives had an entire section devoted to the PPC chips, and the RAD6000 wasn't included in these, but was an offshoot to the side, branching just before the PPC601.

By all other standards however, they seem to be closely related in time and technology to the 601, which powered Powermac 6100, 7100, 8100, 9150, 7200, 8200 and 7500 PPCs, as well as I think, one of IBM's thinkpads.

Those 601s are a very nice chip, and quite underestimated at what they can do at low clock speeds. If the RAD6000 is anywhere similar, I can understand why it was picked.

Re:RAD6000s seem closest to PPC601s

[greed]

There was a 20MHz "RSC" (RISC single chip) processor, used in the RS/6000 model 220. Back when the POWER CPU was a multiple-chip board, this was a single-chip implementation of POWER. Circa 1992.

Some quick googling shows that the RSC series is where the RAD6000 came from, "IBM RS/6000, predecessor to the PowerPC, ...."

And the specs seem to be 25 MIPS at a 25 MHz clock; which is a pretty bad ratio for an RS/6000.... no wonder the model 220 seemed so horribly slow.

The 601 (featured in the model 250) rocked by comparison; they even gave the expensive multi-chip POWER CPUs at the same clock rate (67 MHz) a good run. The big POWERs would win by having 256 bits to memory instead of a measly 64, but it was close. (The POWER2, though... wow, that was a nice box for its day.)

Re:RAD6000s seem closest to PPC601s

[badasscat]

Those 601s are a very nice chip, and quite underestimated at what they can do at low clock speeds. If the RAD6000 is anywhere similar, I can understand why it was picked.

I doubt it, because it probably had little to do with performance. I read an article a while back (tried to Google it now, couldn't find it) that said NASA is generally about ten years behind consumer PC's in terms of the CPU speeds they send into space. These things have to go through so much testing for reliability, etc. that it often takes that long to certify a CPU for space flight. When they do find a CPU they can certify, they'll often continue using it for years. They won't just stick the latest PPC chip in a vehicle, assuming it's the same as the last one, only faster. That's the quickest way I can think of to disaster.

Not much CPU power is really needed for space flight, though, and there's not really much point in taking along more than you need. We went to the moon using slide rules. We can make it to Mars on 20mhz chips - and I'll bet the only reason we even need that much power is for the scientific experiments on the surface. Speed is not really the primary concern on space missions.

Re:RAD6000s seem closest to PPC601s

[Anonymous Coward]

No the 601 was CRAP. The first "power" macs if you could call them that were slooooower than the previous motorola powered ones at half the clock speed. A 33mhz 68040 could out grunt a 601 at 66mhz with its hands behind its back. And I should know, I had one of each. If you had a 20mhz version of that chip thats 10mhz 68k speed which is more akin to the original mac plus.

I doubt the mars rovers are running that slow otherwise NASA would have put a 8mhz space hardened 68000 in there

Re:RAD6000s seem closest to PPC601s

[cosmo7]

That's not necessarily the CPU's fault. When the 601 was considered sexy, PowerMacs were running MacOS 7.5 (or thereabouts) and executing toolbox calls through the 68K emulator (including some of the calls that the emulator itself used), so there was a performance hit.

I agree though, the 601 was kind of primitive.

Re:RAD6000s seem closest to PPC601s

[Cesare Ferrari]

I went from a IIcx (16Mhz 68020 with a 68881 FPU) to a 7500 (100Mhz PPC 601) and got something like a 100x speedup when running some of my floating point based code.

So no, the 601 didn't suck. It was about as fast as you could get in any PC at the time.

I think you are remembering running emulated 680x0 code on the PPC which of course takes a big hit. A wonder of software engineering, but if I remember the emulated machine didn't do FPU correctly, so any code which had FPU optimised code paths didn't get run in the emulator, which caused quite a slowdown.

The 68040 was quite a change from the earlier processors (I think the '040 was the first with a pipelined architecture, and it included FPU on board), and so it had a considerable speedup.

The 68000 was a real CISC processor with all the trappings. I remember those days well, but don't miss it ;-)

Re:RAD6000s seem closest to PPC601s

[Jeremy Erwin]

Sure, if all you had were 68k binaries-- on integer, the PPC601 emulator was half as fast as a comparably clocked 68030. But native applications were quite fast.

A Macintosh (7.8MHz 68000) rates about 0.40 Dhrystone MIPS. A PowerMac 7100 (66 MHz PPC601) gets about 129 MIPS. A 33 MHz 68040, perhaps 23. source

Of course. dhrystone mips are pretty much obsolete as a benchmark.

Re:I hope the flash memory was not commodity hardw

[mo^]

Indeedy! hopefully we can get some cool figures back from one or the other.

I was thinking about this when beagle packed up. If the beagle unit is so "cheap" (i seem to recall 50million ?) compared to the rovers, then i was wondering why nasa and the beagle team couldnt cooperate and literally carpet bomb mars with beagles.. ....

Just random thinking from a brit who REALLY wanted to see the beagle work.

Nice work to NASA though.

Re:But The Question is..

[noselasd]

No, the response time really isn't _the_ big thing. It ofcouse have to be fast but it's how reliable it is. That is, what response times it guarantees. While e.g. Linux sometimes can have much faster response time than some RTOS's, it's still no guarantee that latency is good _all the time_. Having response times ranging randomly from say 0,1ms-0,5s (which might be the case on general purpose OS'es) is alot worse than having a *guaranteed* response time of e.g. 0.5-1.0ms

FPGA's

[retostamm]

There are also Xilinx FPGA's in the Rover. Cool thing because they can be reconfigured if you find a bug while the thing is in transit.

Xilinx radiation-tolerant Virtex(TM) FPGAs are being used in the "main brain" of the rover vehicle, controlling the motors for the wheels, steering, arms, cameras and various instrumentation, enabling the vehicle to travel about the planet.

They also controlled the Pyrotechnical stuff during landing.

[Disclaimer: I work for this great company.]

Re:FPGA's

[sxpert]

and the fpgas in the cameras are from actel, according to the pictures of the boards

Re:FPGA's

[yaff]

The last time I read Xilinx's line on "radiation-tolerance" it was that you could reprogram (or refresh) sections of a Virtex while the rest of the chip was in use. So if a stray gamma particle flips a bit in your configuration SRAM you can reset it.

This seems more like radiation-dodging rather than tolerance.

Re:I hope the flash memory was not commodity hardw

[c_oflynn]

Trust me on this one - NASA probably knows more than all this Slashdot discussion...

Everything in there is space-certified - the chips will al be radiation hardended, each chip is probably around 500 times more expensive because of this, it aint easy. Then that is combined with multiple-redundancy and so on...

The CPU itself won't flake out probably, but what can happen is SEU (Single Even Upset), which is where bits get randomly flipped. This happens in a normal CPU a bit, but radiation makes the problem quite severe.

Re:I hope the flash memory was not commodity hardw

[The_K4]

So now you can have 2 rovers with the SAME software bug!

Mod Parent up

[Anonymous Coward]

Mod parent up... very informative!

Rad6k processor, space processor design.

[Anonymous Coward]

The processor is probably the Rad6000, rad hard powerPC. You can adjust the processor speed while running to save power, but that's about the end of the good things.

To satisify the simpleton editor comment "I wonder if I can make a beowolf cluster/firewall of these things" - Probably not. Generally there is only one brand of 10MBit Ethernet interface for them (which is klutzy) and the pipeline on these stalls more than a '73 yugo. It's a specialized processor, but it's not terribly amazing by terestial standards.

The problem is, the newer faster processor designs rely on things like cache, predictive pipelines, etc. Most of these circuits, in a chip, are pretty sensitive. If you get a cosmic ray hit (the real thing, not a martian ray gun) on most things like cache, then the bit error propegates. When they rad harden a processor, they redesign it so that when you get SEU (Single Event Upsets) they generally do not propegate. This takes a pretty significant redesign of the processor.

Re:I hope the flash memory was not commodity hardw

[Anonymous Coward]

Is the RAD6000 Hal9000's grandfather/mother/entity/beta version?

Re:Ask Bill Gates

[LA_Samurai]

Actually, Bill Gates would be the perfect person to ask. As I remember it, he DID say no one needed more than 640 KB of RAM. Presumably, he could have built the Rover with 640 KB RAM running DOS. Of course, no multi-tasking, no support for flash memory, file name limited to 8 characters with 3 character extension, and other technological innovations DOS is renowned for.

RAD6K

[Anonymous Coward]

I am an engineer that works with the RAD6K processor boards. A couple of observations here.

1. The RAD6K really does run at 20 Mhz. They're creakingly slow. They're spec'd to run up to 33 Mhz, but the customer can get them to clock at lower speeds (I've seen them run at 12.5 Mhz). The only drawback is the PCI bus is also clocked as the same speed as the CPU. This is a mixed bag - but a slower PCI bus helps improves signal integrity and decreases power consumption.
2. The board is PCI, but NOT compact PCI. There is a proprietary PCI connector and a proprietary PCI backplane. You cannot plug commercial PCI products unless you have an adapter to interface to the proprietary PCI connectors.
3. For those who are not aware, there are three types of memory being used on the rovers. There is the SRAM (the RAD6K boards use SRAMs, not DRAMs), the EEPROM, and apparently, FLASH RAM. The EEPROM and the SRAM are on the processor board itself - there is probably more EEPROM memory in the system on another board. The EEPROM usually holds the flight code, and there are usually two copies. An original version that was launched with the spacecraft, and one patched version made possible via uplinks.
   
4. I am amazed at the presence of FLASH RAM's. I am not aware of any rad-hardened FLASH RAM devices for spaceflight use. In addition to radiation hardness, the device must be made reliable with an approved QML-Q or V manufacturing flow. Radiation hardness is just icing on the cake, but the key is that the device must be reliable to withstand temperature extremes, shock and vibration. So, I have yet to see a FLASH RAM device that can be used. I am aware of the Chalcogenide based RAM's which are essientially uses the same substrates on CD-ROMs as memory cells. These products are hard to come by right now and are a high risk because we don't have sufficient data and flight heritage. A catch-22 in flight design is if it hasn't flown before, we don't want to fly it. But at some point, someone has to fly the first generation (someone who is willing to take a huge risk). Anyway, the FLASH RAM's on the rovers are in all likelihood upscreened commercial products. In other words, a mass buy of an entire lot of dies of commercial FLASH RAM's may have been bought, packaged in-house or through a vendor, and then screened for reliablity at extended specifications. This is not the same as the manufacturer who can guarantee the specs by design by designing it from the outset with increased reliability in mind.
5. Radiation shielding? Minimal at best! The RAD6K shields its SRAMs by placing it on the underside of the processor board and orienting it such that the particles hit the processor side of the board instead of the RAM side of the board. There is some degree of radiation shielding for particles of sufficiently low energy. The truly high speed particles are going to get through and the only thing that will truly stop them is shielding whose thickness is measured in feet. That amount of shielding is too heavy for launch. The best we can do is mitigate the effects of radiation by guaranteeing devices can withstand a certain amount of radation dosage (measured in kRads) and design for latchup protection (latchup is a parasitic condition in which an ionizing particle impacts a transistor structure in a way that causes a SCR to be formed and a runaway current condition is initated leading to the device being burned out by high currents). Radiation effects in the form of SEE's (single event effects) such as bit flips can be mitigated by redundancy and voting circuits, memory scrubbing, and error checking using checksums/CRC's.

Re:RAD6K

[baryon351]

The RAD6K really does run at 20 Mhz. They're creakingly slow. They're spec'd to run up to 33 Mhz, but the customer can get them to clock at lower speeds (I've seen them run at 12.5 Mhz). The only drawback is the PCI bus is also clocked as the same speed as the CPU. This is a mixed bag - but a slower PCI bus helps improves signal integrity and decreases power consumption.

This is the part I'm curious about. 20MHz is slow by any measures compared to current desktop machines, but how do they compare to other CPUs of the time, or other CPUs of the same speed? would a 20MHz RAD6000 under or out perform a 486? a 68040? perhaps the first of the real PPCs, a 601, at the same speed?

I know speed and outright performance is a much tinier part of the whole picture with a spacecraft's computers, where the track record of reliability and capability proved true, rates as high as anything else, but I'm curious about whether there are some unique parts to the RAD6000's logic, separate from its radiation hardening, that make it especially worthwhile for the uses they're put to.

Re:RAD6K

[demachina]

"A catch-22 in flight design is if it hasn't flown before, we don't want to fly it. But at some point, someone has to fly the first generation (someone who is willing to take a huge risk)."

Or you fly it as a non mission critical experimental payload which is what we did back in the day I worked on avionics. You fly it as an experimental package so it gets the stress but if it breaks either its not in the mission critical loop, or if it is in the loop you can switch back to proven hardware. I kind of assumed this would be a standard part of qualifying electronics for space flight as well though its obviously a lot more expensive. Its not feasible to test it on Mars due the expense but you could test it in geostationary orbit where it will get lots of radiation and temperature extremes, as well as launch vibration and G's.

Re:RAD6K

[Anonymous Coward]

Maxell Technologies Press

Of course, I'm curious as to whether this stuff was flown before (seems not), and if not it seems to be a bit of a leap for JPL to take. Also, curious as to whether this allows more modern hardware. I have doubts myself.

Re:RAD6K

[stuffduff]

What would it take to generate a magnetic field strong enough to divert particles? If particles can be bent around the sensitive semiconductors this might enhance their survivability.

Just a thought.

Re:RAD6K

[cinnamon colbert]

re point 4..so they ought to be flying 20 pieces of next gen hardware every mission to get flight time experience ?

Re:RAD6K

[gnuman99]

A few things come to mind regarding the size of the magnetic field:

• A magnet the size of Earth
• Part of a nutron star
• Some sort of a plasma system (see plasma drive on JPL site somewhere)

All of these might just be slightly heavy pieces of equipment.

Note that most of the radiation damage that the rover can suffer was during the flight to Mars. For example, a solar flare poited in the right direction might have fried it during flight - it might have a difficulty doing so when it is on Mars

The bottom line is that the major problem is NOT the radiation but software. I wander when NASA will start using Linux - the current system is worse than Windows 3.1!

Re:RAD6K

[JGski]

Too much shielding for SEU actually increases the total dose exposure due to spallation/bremstrahlung. So if you want both SEU and total dose protection you must minimize Z-shielding and go toward circuit-based hardening, just as the article mentions shielding with "capacitors and resistors".

Re:RAD6K

[anubi]

Interesting...

Space is cold.

Superconductors require cold.

Could energy be stored in a magnetic field? So instead of kinda uselessly being stored as chemical energy, maybe the energy could be used to deflect radiation? Of course the energy stored that did not get absorbed remains usable when withdrawn from the shield field?

A lot of missing stuff here... just a concept... but is it workable?

Not Windows

[Quila]

the system allows users to add software patches -- such as a glitch fix or upgrade -- without interruption while a mission is in flight. "We've always had that [feature] so you don't have to shut down, reload and restart after every patch,"

All you needed to read was that to know they weren't using Windows.

Man...

[nphinit]

...imagine throwing Linux on one of those things! You could make a Beowulf cluster of Mars Rovers! You could encode .ogg at blazing speed! ::convulses with glee::

make a firewall?

[SilverGiant]

.

"I wonder if I could make a nice firewall with one of these for my home network..."

1. build computer
2. drop off on mars on way to grocery
3. laugh in face of hacking
4. reconsider when computer mistaken by rover for new type of rock, drilled for mineral content

Off earth is a good place

[Anonymous Coward]

SCO can hardly claim IP protection once out of earth's atmosphere.

I'm being serious.

Re:But The Question is..

[Anonymous Coward]

And for all that, nearly all RTOSs have very fast response times. Because if you need a guaranteed response, most of the time you want a guaranteed FAST response.

But yes, what makes something an RTOS is that it guarantees that something will happen within X time from when you want it to happen.

Re:I hope the flash memory was not commodity hardw

[A55M0NKEY]

Redundancy.. Every computer has parts, and each of those parts has a cross section. If you have 3 computers, then you can use integrity checks to detect computers that have been zapped by radiation and are not running properly. When this happens to a computer it usually means some part has been struck by a xray or something.

Suppose a 'computer' has 3 parts the A-chip the B-chip and the C-chip. There are 3 redundant computers. The A-chip in computer 1 gets fried, the B-chip in computer 2 gets fried and the C-chip in computer 3 gets fried. Now you have 3 broken computers yet 2/3 of your chips remain unfried. You could build 2 more computers with your still functioning parts. I wonder if the parts in the RAD6000 are internally redundant. Is the 'A-chip' in the RAD6000 actually consist of 3 redundant A-chips and an integrity checker? It would seem that the smaller your redundant systems' cross sections were and the finer the grain of your integrity checks were the better your chances of mission success.

If each square cm of board has P = .01 of being zapped by a ray during the mission then a 100 sqcm board has a 37% chance of surviving if there are 3 redundant 100 sq cm computers then there is a 25% chance they all will die during the mission. But if each sqcm has redundant copies of itself which are integrity checked, the probability that every single square centimeter on the 100sqcm board will still have a functioning copy of itself and so the probability that the board as a whole will survive the mission is 99.99% a 0.01% failure rate.

Clearly 0.01% failure is better than 25% failure.

I thought...

[Anonymous Coward]

"Besides, everyone knows that what makes a Mac a Mac is the operating system."

I thought it was the price that made a Mac a Mac...

Re:I thought...

[Anonymous Coward]

nope, its the infant approved 1 button mouse...

No...

[Mr. Underbridge]

Was is only me who read windriver.com as Win Driver (not Wind River) dot com? Got freaked out there for a while... phew...

...the guy who posted that joke first obviously did.

Now I know

[Golobarti]

Why one is missing - ET is running it as a firewall. Now it's his turn to send me a sweet, sweet gaming system

Re:Ask Bill Gates

[Anonymous Coward]

Really? He said that?

Well I never!

Oh, BTW, you deserve to be modded to oblivion for the most tired piece of karma ho'ing this week.

Re:I hope the flash memory was not commodity hardw

[Lord Graga]

Trust me on this one - NASA probably knows more than all this Slashdot discussion...
Phew...

Re:I hope the flash memory was not commodity hardw

[uid100]

How about a simple 2 flash card solution setup in a mirrored configuration - if one card pops, you still have data integrity. I'm sure there are many problems with this, but seems to be plausable on the surface. Minimal cost and a $1B project could be saved by such measures.

Re:But The Question is..

[Anonymous Coward]

Except that according to one poster, most of the real time is hanlded by a Xilinx FPGA. VxWorks is a horrible general purpose OS, it's no wonder they're having file system problems. If there's no real time code it's hard to see why they would have chosen it.

What OS?

[chadm1967]

"The operating systems running on Spirit and Opportunity are based on a flexible commercial platform initially chosen by JPL engineers for its reliability."

What platform? I wish they would state what it is.......

Re:What OS?

[cotodoso]

They did, in the next paragraphs -- VxWorks by Wind River. It's a real-time OS that gets used quite a lot for this type of work.

I've worked with it once or twice. I didn't really care for it -- it seemed like some wierd hybrid of a Unix CLI and DOS.

Re:What OS?

[fishbowl]

>What platform? I wish they would state what it
>is.......

VxWorks, from Wind River.

In my mind, the flexibility that ends up in the finished product is much more interesting than the reliability. It's quite cool that they can reprogram the thing from 400 million km, and be reasonably confident that it will still work. Or even, that, if they break it, they can fix it.
NASA must do a LOT of work to resolve the conflicting goals of reliability and flexibility.

Re:What OS?

[anubi]

Hybrid of Unix CLI and DOS.

Yes.. exactly as I understand it too. From what I see, its much like a multitaking DOS kernel, where much R&D has gone into building a very tight and robust kernel. From what I see, they are trying to have the kernel itself as simple and bare-bones as possible so as to minimize any possibilities of losing control in such a manner as the kernel itself cannot restore order in the system.

This, like the assembler, is more on the order of tools an embedded programmer needs. But then, we are embedding huge things these days, and I am still used to embedding small things.

Mission control software is in Java

[danwiz]

As noted in a previous slashdot posting, the software in the control room was written in Java.

A ZDNet article says Java made communicating between multiple software pieces very flexible and James Gosling, inventor of Java, spent considerable time helping develop the system. Sun also describes how the same application was used for the Pathfinder mission back in 1997.

Re:Mission control software is in Java

[bala_dei]

Why do they want to add one more layer of BUG? Is C++ communication is not worth doing?

Ping times will kill you...

[nsushkin]

"hardware and software of Mars Rovers ... I wonder if I could make a nice firewall with one of these for my home network..."

The ping times to Mars and back will kill ya.

According to SCO...

[holizz]

...their code is in all the Mars Rovers, most of the shuttles and I hear GNU/Linux boxes are popular with the Martians.

How original

[Anonymous Coward]

Deja vu

software and hardware source?

[drwho]

Is the source code available? or at least parts of it? How about the hardware - rad-hard CPUs are useful, though satellites have been known to survive 30 years without and special treatment.

I'd just like to put a Direct Connect Hub in space...yeah!

Re:software and hardware source?

[anubi]

I followed your link... it was about the OSCAR Amateur Radio satellite, which is a helluva good job. I guess they used the RCA1802?

But, to the point, the processor designs for space usage have had a lot of being looked at for reliability.

Thats a helluva good thing.

In practice, a lot of ordinary business economics mandates good enough is good enough, despite the fact it crashes a lot. Look at our biggest software company and consider whether the product derives its value from its reliability, or from its marketing.

If one spends too much time on something the customers see of little value, you won't stay in business long.

Like many of you, the technical knowledge I have illustrates the foolishness of deploying non-robust technology, but most people do not see computers the way we do. Most people seem to see it like cars.

I am currently taking a class in auto mechanics tune up just to see the state of the art of the robotics and computerization in auto electronics. It amazes me how styling often mandates a very non-robust design, making the auto very difficult ( read: expensive ) to work on.

Example: leak-prone air conditioning connectors which are designed for rapid assembly at the factory, not designed for longevity. Ever try to find a slow leak in a freon system where everything is tucked under panels taking hours to access because all the mounting hardware has to be aesthetically designed so as not to be visible? Of course the customer does not see repair costs in the showroom. But there is little financial incentive to do it right. Car sells for the same whether or not the air conditioning seals are designed to last for the life of the car.

NASA and the Space Program provide the financial incentive to build robust stuff. Face it, it takes a helluva lot of R&D over and beyond "what works" to optimize the design for extreme robustness. As a government effort, all this R&D effort now trickles down to the commercial sector, where we can use the advanced technology without having to pay the premium for its development. If you were designing an ATM, and were lucky enough to work for a company who takes reliability seriously, wouldn't you love to take advantage of all the work NASA has funded on these rad-hard processors and fault-tolerant techniques? You know how many people get seriously inconvenienced whenever an ATM they are counting on to work doesn't? Most of the cost of technology is in R&D, not actual production. It would not surprise me at all to see the same advanced techniques developed for the Mars Rovers showing up in vending machines.

Re:software and hardware source?

[drwho]

This is a bit late, but hey thanks for the great reply.

On the terms of robust-ness, there is als o the problem of avoiding the accusations of waste and abuse. There was much disgust over $300 hammers and $800 toilet seats and the like (I don't remember the figure, but you know what I mean), but was the idea addressed that these are not mass-marketed items, and in fact maybe they are unique, and that all the tooling, testing, and paperwork costs have to be counted in?

Quality costs money, it is true, but you never "get what you pay for" - you get what you negotiate. Dollars don't have minds to choose the better product, that is your job, as a buyer. And there's a good number of people that deserve to be fired from their jobs as consumers, for shoddy work.

Re:I hope the flash memory was not commodity hardw

[Rostin]

why build huge redundancies into single device when you can just manufacture two and toss 'em both at the job

I'm not sure that would be better. A rover with no redundant systems could be crippled by the failure of any of a huge number of components. Even though you've got a whole spare rover, the same thing is true of it. So basically two failures (even if they are unalike) could ruin you day. OTOH, if you build in redundant systems, several primary components could fail and you'd still have a working rover. You'd have to break the primary and all the backups for any given system to put it entirely out of commission.

Your point about having twice the data is a good one, though.

Re:I hope the flash memory was not commodity hardw

[fshalor]

With the ability to update code on the fly, most of the changes they had to make to spirit to make it work have been made to oppertunity before go time... Reduncy in the humans is also a part of the program.

Since the've also had more than one path for getting data to and from the machine... it's all good.

Yeah yeah yeah ....

[dsb3]

.... so how many FPS does it get?

chernobyl

[bobsalt]

why? you live near chernobyl!?!?!

thats IT!!!!
1. Build radiation hardened firewall.
2. Movie to Russia where radiation hardens YOU.
3. ?
4. Profit!

Firewall for home...

[xfs]

Use the robots while they're still on mars as your firewall... that's going to be only ... 20 minutes lag?

Re:But The Question is..

[Anonymous Coward]

VxWorks!!!

The fact that NASA put a probe on Mars using a 60's style OS architecture is an amazing tribute to the programmers who worked on it.

- No implicit memory protection (protection domains - ha!)
- Goofy GUI-based development environment
- the list just goes on and on...
- Questionable real-time performance (it works great if you shut down all services)

Overall, I am surprised that NASA continues to use this outdated software. There are far better alternatives in the embedded real-time OS domain that would easily run in the same environment.

And before everyone starts talking about Linux, forget it, it is too big and fat and impossible to rate for high reliability missions like this one. By the time you trimmed it down to size, you would be better writing your own OS.

As a company, Wind River is dying, their market share is down, and people are starting to wake up to the fact that you need good programmers to write software, not goofy GUIs and outdated kernels.

Ask anyone who programs embedded systems who has used VxWorks and you will always get the same response - "Man I hate that OS, and management wants us to use their tools, but they are useless".

Amazing things can be done with a watchdog timer...Anyone who uses Wind River for anything remotely serious knows that.

I just hope NASA picks something else when they send people to Mars, or else here will be the conversations:

Astronaut: "Base, I have to hold my breath for a bit while my suit computer reboots, it must have crashed when I exhaled too much...damned OS crashed again!"
Base: "Roger that, maybe you should reload your suit software? or perhaps breath slower?"
Astronaut: "mmmmfffffff!"

Wind River is like the Microsoft of the embedded world, except a choice still remains.

no inodes per se...

[dvd_tude]

This set of slides hints that they used DOS FAT. Since the Flash is 256 MB, that would imply FAT16.

Still doesn't completely explain the running-out-of-RAM problem, though. Kinda nasty that it reboots - it should degrade more gracefully than that.

By the way, I doubt that they swap to Flash or use VM at all due to the non-deterministic nature of VM (and the fact that it wears out the Flash.)

Intel Inside

[thomas536]

*Intel chime*
Intel Inside

10 years ago

[Baki]

I was using about this configuration (20Mhz powerPC and vxworks) for prototypes of similar robots. At that time, the real thing (to go into space) would run into a few kilobytes without real operating system (just an ADA runtime) on some mil-std obscure CPU.

At those times, we dreamed of being able to put our prototypes into space instead. It is nice to read that that has come true :)

Possibly

[maroberts]

because they wanted a processor that did not guzzle several gazillion watts when running.

I suppose they could avoid any problems with cold by running an Intel/AMD CPU though....

So, could we say...

[Verminator]

Macs are from Mars, PCs are from...Uranus?

Never mind.

Can I pre-mod myself as a Troll?

The CPU is POWER, not PPC

[rshadowen]

The RAD6000 is based on a POWER CPU called RSC (RIOS single chip) - it's not a PPC chip. This was a design that consolidated the 5 - 6 chip RIOS processor complex onto a single lower performance die for low end workstations. I worked on the development team at IBM.

The RSC design played a key role in bringing Apple and Motorola together with IBM to create the PowerPC line of CPUs. The 601 was the first PPC and was basically a redesign of RSC. It supported both POWER and PPC architectures, although there were deviances from PPC since the architecture was actually being defined at the time we were working on the chip.

The RAD6000 version of the design happened because IBM wanted to pursue some government contracts, so had the RSC specially qualified. Another group then took the design and performed the radiation hardening.

After Pathfinder we had some cool IBM/Mars posters hanging around the building, but oddly enough they vanished very quickly...

Same problem three times

[DunbarTheInept]

Redundancy is useful for handling hardware errors, but this problem was in the software that keeps track of the flash filesystem. So having three computers would just mean you get the same problem three times, and your expensive redundancy would have bought you nothing. (They are having to take steps to make sure Opportunity doesn't encounter the same problem.)

Firewall

[Anonymous Coward]

I wonder if I could make a nice firewall with one of these for my home network..."

...as opposed to Beagle; which made a nice fireball.

MOD PARENT UP

[Shinglor]

Very informative, when I saw 25 KiB/sec on BitTorrent I just thought the author was a moron!

Firewall

[mr_z_beeblebrox]

I wonder if I could make a nice firewall with one of these for my home network...

Sure they can take radiation, butblaster....I don't know.

Re:But The Question is..

[Quill_28]

Oh good lord, I don't want it to run linux, just saying it could

Re:NASA is run by idiots...

[karnal]

Don't know if you're trolling or not, but the engineers who send this stuff up there want to make sure it works.

Kind of like how the financial industry is still stuck on COBOL. It just works, and all the time spent seeing if a "new" cpu or programming language would be wasted getting it back to the reliability of the old system.

We don't need the latest and greatest up there, just something that works. As an aside, power consumption is a big issue as well. Would not want something that dissipates 60-80watts of heat (which translates to at least that much power to drive it full tilt) since they are "low" on energy resources as is.

Re:NASA is run by idiots...

[egomaniac]

Are all the millions of dollars spent on full for the rockets?! Why the fuck is my home system more advanced than NASA's? You'd think they'd of at least used a design similar to a Dual G4 or maybe even a Dual Xeon or P4. Can someone explain to me why NASA gets millions of dollars if they are using 1990 equipment?

A) They don't need to play Doom 3 up there. 20MHz is sufficient for almost anything you would want to do on Mars. Why send up more than you need to?

B) Your computer runs far hotter and consumes far more power than the Mars rovers do. Power is at a premium when you're millions of miles away from the nearest electrical outlet.

C) The rovers are radiation-hardened. Your system is not. Your computer would last about twelve minutes in space before it locked up. A big part of radiation hardening is using larger (and therefore slower) transistors.

D) It takes years to certify a particular piece of equipment as spaceworthy. NASA isn't going to just pop in the latest and greatest Athlon and assume it will work "because the last one did". That means that anything flying into space is automatically going to be at least a few years behind the curve.

Case

[Anonymous Coward]

Ok, mb is so obviously wrong that everyone knows it should be MB, but "Captain Zion" lists it as 128 Mb, which would be 16 MB. How am I supposed to know what he means? Do you people expect me to actually read the article?

I mean, with a bus speed that low, you can probably put the data to flash ROM just as fast.

Flash isn't quite as robust as SRAM, but who knows. Maybe the OS is just really inefficient.

What are the chances of you using all 128mb of ram?

Well I dunno, considering that it's dealing with lots of photographs...

The AP is now reporting....

[Anonymous Coward]

...that the problem with the first rover was an "Endian" problem. Apparently, one of the mission controllers, feeling over confident, sent commands from his Macintosh laptop instead of the proper command controller. Macinstoshes use "backwards" endian (high byte in lower address) as opposed to the more efficient "regular" endian used by Intel/Windows and Linux.

This backwards endian screwed up the rover and it had to be reset. NASA has now ordered all Macintosh laptops off the network. Apparently, Apple had donated a bunch of Mac laptops before a Nova special was to be taped, to the Macs would appear on television. (This is common practice for Apple.) A few of the mission specialists liked the machines and used them.

This could be a major scam! NASA executives taking in effect "bribes" to get product placement for equipment!

You forgot the most important step...

[Anonymous Coward]

5. Profit!

Re:space shuttle uses 1969-vintage ibm 360 compute

[greygent]

The B-52G/H models used 3 ACUs each composed of (if I recall correctly) 4 Z-80 processors. i'm not sure about the number of CPUs in each systemn, but I am sure of the processors. These systems were loaded (very slowly) via tape, and they ran the RADAR and bombing computer subsystems.

I'm not sure which computers this article is referring to, but it may have been an earlier revision of the computer system in the B-52.

The only other computer system I can think of in the B-52s was the ECM system, and since it was highly classified, and I did not work on that system, I'm not overly sure.

For a little extra karma whoring, i'll relate a funny story. Our base, which shall remain nameless, would fly early morning ECM sorties while people were driving into work.

Their target signals for jamming? The radar guns the cops used in their speed traps to try and catch our folks driving into work. Apparently, the cops finally caught on and complained at the high number of equipment failures with their radar gun equipment.

Re:space shuttle uses 1969-vintage ibm 360 compute

[jostallin]

So can I imagine a Beowulf cluster of Space Shuttles?

NOT 20MHZ AT ALL!!!!

[Anonymous Coward]

WHAT THE ARTICLE SAYS IS WRONG.

THE WEB OF THE LINK SAYS: "THE PROCESSOR CARRIES OUT 20 MILLION INSTRUCTIONS PER SECOND". EACH INSTRUCTION MAY TAKE AT LEAST 4 CLOCK-CYCLES TO BE EXECUTED. SO THE REAL SPEED OF THE PROCESSOR'S CLOCK WOULD BE 100MHZ AT LEAST.

EVEN WITH THIS CLOCK SPEED THERE ARE COMPUTERS (LIKE MACS) WHICH GET THE PERFORMANCE OF A COMPUTER "APARENTLY" EVEN 4 TIMES FASTER-->CLOCK SPEED IS NOT COMPUTER PERFORMANCE AT ALL, IT'S JUST ONE MORE FACTOR

NOW THINK ABOUT THE PENTIUM-I AND THE FIRST RISC (REDUCED INTRUCTION SET COMPUTER). THIS ARCHITECTURE IS BASED ON THE USE OF FEW INSTRUCTIONS FROM A SET OF SIMPLE INSTRUCTIONS TO REPRESENT A SET OF COMPLEX INSTRUCTIONS WICH MAY TAKE A LOT OF CLOCK CYCLES (EACH ONE WITH A LENGTH OF 1/MHZ SECONDS), TO IMPROVE PERFORMANCE.

OT from Parent and Article:

[deadlinegrunt]

Sorry but you missed it entirely. Perhaps it is my poor articulation and my apologies for that, but reread it until the context sinks in rather then the literal meaning you took from it. Hint: "I am amazed that they can 'save 3 months of development time' with as little as they have" was not the catalyst for the subjective response, it's that you can do quiet a bit with very little resouces and intelligent thinking rather than letting a tool replace what you *should* be doing in the first place. That is also regardless of your 'market'. People that don't get that usually never do though, hence my original OT response.

On a side note I am curious if you are actually inferring that implementations are not carried out on inappropriate platforms for no other reason than because it's possible? If so, that's rich...

Re:OT from Parent and Article:

[JDBrechtel]

No, that's not what I meant. I did take your comment out of context. It appears I did what I assumed you were going. That being, making a general statement not related to the topic at hand (the Spirit).

I was wrong.

Just human

[edxwelch]

"Raises some interesting questions about software reliability, I think. Did nobody think about running out of disk space?"

Just goes to show you we are all human, whether you are the NASA Lead Software Architect, or just a Joe Soap hacker you still will make the occasional blooper.

Better latency, too

[billstewart]

The Linksys not only has a faster CPU, it avoids the delay of shipping your packets to Mars and back....

Re:Sigh

[A55M0NKEY]

The only part of the solar system known to have an 'ecology' is earth. But isotopes can be detected in minute quantities. I'm sure an asteroid that was pelted by the exhaust from a nuke rocket would have very detectable levels of those isotopes. Also, ( and I am not an expert on this ) elements can be detected from a distance. A cloud of exhaust would probably skew things. People aren't going to be able track the orbits of radioactive exhaust for every nuclear powered space ship. They won't be able to be discounted.

As for nuclear powered spacecraft that do not vent radioisotopes, I have no problem with them in use outside the earth's atmosphere

As for point #4, you do change things by existing, but you can try to be intelligent about HOW you change things. If I lived on an island with only 20 coconut trees I would not burn them for light and warmth. I'd burn any other wood I could find even if it was inferior to palm trees.

Carbon 14 dating is very accurate for recent ( less than 10000 years) events. It can get you within a few years. There have been no natural uranium fires in that time period. But used to date events that happened millions of years ago, it is less precise. That is usually fine. It usually doesn't matter if a T-Rex died on a Monday or a Teusday after all. Also, natural uranium fires probably didn't spew crap into the air like a nuke blast, most of what was burning probably stayed in the ground.

Forget about Athlon's...

[GrodinTierce]

Tejas draws 150 Watts.

Imagine a...

[patheticloser]

beowulf clus...never mind.

Mars router?

[Luminary Crush]

"I wonder if I could make a nice firewall with one of these for my home network..."
Maybe, but if you thought the latency on your modem was bad, do you think you'd be happy bouncing your signal off Mars?

Re:Sigh

[AKAImBatman]

I'm sure an asteroid that was pelted by the exhaust from a nuke rocket would have very detectable levels of those isotopes.

*Groan* Uranium is COMMON in asteroids. My engines would use no more than a few tens of pounds of uranium. Worst case, we'll say it uses a few hundred pounds. How is the particulate matter being exhausted by my engines anywhere near the TONS of uranium contained in many asteroids? In fact, most Uranium on Earth is from the tons that burn up in the atmosphere every year.

As for point #4, you do change things by existing, but you can try to be intelligent about HOW you change things. If I lived on an island with only 20 coconut trees I would not burn them for light and warmth. I'd burn any other wood I could find even if it was inferior to palm trees.

But if you lived on an island with hundreds of thousands of palm trees, would you really feel bad about burning a few for a fire? Remember, a ship will burn/expel *pounds* of uranium. (Much of which will be some other substance after use.) How does that compare to the BILLIONS OF TONS in the solar system?

Also, natural uranium fires probably didn't spew crap into the air like a nuke blast, most of what was burning probably stayed in the ground.

The "crap spewed" doesn't stay in the air very long. Uranium is too heavy. Carbon dating was messed up from radiation pulses. Nothing more, nothing less. I'm sorry that makes life harder for geologists, but that's an unfortunate part of life. How many archeological digs were made harder by robbers, tourists, curious natives, etc, who had all previously wandered the area? It's a fact of life. Space geologists are going to have to get used to the fact that engines will alter things slightly. Possibly, many of them won't care since it's so little. The ones who do, will have to figure out how to work around it.

Re:NASA is run by idiots...

[JGski]

This does seem like a Troll, but maybe it's simple ignorance about the operating environment of space.

COTS (commercial off-the-shelf) parts don't last long above LEO (low-earth orbit). I used to be involved in testing COTS and RAD parts for space use; in most cases COTS parts died permanently within seconds to minutes when exposed to "typical" interplanetary or deep space radiation levels, while parts from IBM's rad hard line lasted longer that we were able to test them.

Re:NASA is run by idiots...

[greymond]

um ok yes my system is not space worthy, but what I said was

"at least used a design similar to a Dual G4 or maybe even a Dual Xeon or P4"

They are not using actual PPC processors from Apple, there using RAD6000, which is...similar...to the PPC's

I'm just complaining that they could/should use better or give me my millions of dollars in taxes back. I see them as a money sink hole, that does not produce anything worthy with the amount of money the government pumps into them.

Re:I hope the flash memory was not commodity hardw

[The_K4]

I agree, but had it been a hardware issue having the rovers identical might not have been such a good plan. As it's a software issue it's fixable from remote which is a Good Thing!

What's Inside the Mars Rover?

[AsnFkr]

A very unhappy midget.

Do you spend the rest of your time

[myowntrueself]

writing 419 emails???

If you want more redundancy

[rk]

Write your congressman to give NASA a big budget increase. This isn't the 60s when the money was there for the asking. They talk about these rover missions as an 800 million dollar mission, but it's not 400/400. The first one was about 550 (figuring it out) and the next one was 250. We can probably do more MER-type rovers for 200 million a pop now. In terms of interplanetary missions, that would be chump change.

Redundant systems and extra engineering costs money, though. The budget isn't a concern, it's a HARD limit. "DO X with Y dollars, or else don't do X at all." This isn't the Apollo program anymore.

Images of the inside of Spirit

[DigiShaman]

Go easy now on the server. But here is your tax dollars at work. No wonder these robots are so robust. The simplicity in enegineering is astounding. Link below.

http://www.roombacommunity.com/robosweep/roboswe ep .html

Re:NASA is run by idiots...

[k8er]

Ok then, they can take the x% of your taxes that go to NASA and apply it to something that you feel is more worthy, as long as they let me redirect x% of my taxes from some program that I think sucks towards NASA's budget. Then we'll both be happier.

Air pressure!

[Anonymous Coward]

Hard drives require air pressure to fly the drive head over the platter. Drop to mars ambient presure (1% of earth ambient) and boom, the head crashes and scrapes the bits off the platter.

You can't seal it because the case can't take the pressure.

Also, moving parts are a lot more temperature-sensitive than silicon. They have lubricants which sieze up and thermal expansion which messes up all the tolerances.

CMOS generally works better at lower temperatures, and the usual low-temperature limit is simply what the manufacturer bothers to test because that's the customers want to pay for. You may have seen pictures on the web of overclockers dunking Intel's microprocessors in liquid nitrogen (approx -196C = -321F) which is way outside of their specified temperature range) and they run just fine.

You can find web pages on the design of seriously cryogenic (liquid helium and below) preamplifiers and the like. It does things to standard transistors, but they can still be used.

The main thing you have to watch out for with wide temperature extremes is thermal expansion stresses cracking the IC packages (metal lead-frame expands more than plastic case) or breaking the solder joints to the board.

But basically, sometimes equipment can be run well outside its normal specifications safely if you understand how it's made and what the effects of the variation are in detail. Rated temperatures mean that the manufacturer promises that the equipment will work at those temperatures and hasn't spent much effort ensuring it'll work outside those temperatures, but it doesn't mean the gadget is guaranteed to not work outside those temperatures.

With careful part selection, you can do some pretty impressive stuff. There are folks who build instruments for the bottom of oil-well drills, a high vibration environment swimming in 200C cooling/lubricating "mud", out of carefully selected standard commercial components.

Now the problem with two processors - if they disagree, how do you decide who's right? And how much chance for error does your deciding circuitry add to the system? While processors usually fail stopped fairly quickly, doubling the number of processors doubles the chances that something will fail in a more creative manner.

These guys are not idiots. They thought through the implications very carefully. The RAD6000 processor is very thoroughly tested in space and has an excellent reliability record.

Having said all that, I do wish I could find more on the web describing the design decisions made. There's generally some very good engineering in these things, and the stuff I've read like this History of Apollo Launch Facilities and Operations is fascinating.

And whoever complained that they were still developing the software after launch - well, they had 6 months, and I'd be surprised if they couldn't find one bug or worthwhile enhancement in that time. They can do remote upgrades, and might as well take advantage of it.

Re:Air pressure!

[AKAImBatman]

You can't seal it because the case can't take the pressure.

Lead can't take pressure? Why not?

Also, moving parts are a lot more temperature-sensitive than silicon. They have lubricants which sieze up and thermal expansion which messes up all the tolerances.


Yep. That's why I wanted a sealed case with no airflow. Cooling would happen by heat passing through the lead case. Alternatively, you could better insulate it and build in some sort of active fluid cooling in the case.

CMOS generally works better at lower temperatures

In theory. I'm not sure that's really the case though, Thinking it through, the contraction of the memory is going to make it difficult for the UV rays to properly target the memory. Granted, I haven't paid much attention to the design improvements in flash memory for a very long time.

Now the problem with two processors - if they disagree, how do you decide who's right?

That's a standard problem with clusters. The best answer is to recalculate. Alternatively, you could go triple redundant, but that's starting to get extreme.

These guys are not idiots. They thought through the implications very carefully. The RAD6000 processor is very thoroughly tested in space and has an excellent reliability record.

Don't get me wrong. I think NASA is doing very well. After all, NASA's rovers are roving while ESA's aren't. That says a lot about NASA. I just have different idea on what "Cheaper, faster, better" means. My interpretation of that isn't going to penetrate NASA because that's not their culture. Their culture is that everything works right the first time, which is in direct opposition to the "cheaper, better, faster" idea.

And whoever complained that they were still developing the software after launch

Not me. I'm just glad they were able to recover. :-)

No, not the same CPU that was used in Macs

[KewlPC]

The RAD6000 is basically a radiation hardened version of IBM's RS/6000 PowerPC processors. While RS/6000s are PowerPC chips, it was Motorola who supplied all Macintosh processors until the Mac G5.

The IBM PowerPCs and Motorola PowerPCs are different beasts that just so happen to share the same instruction set.

Question?

[Anonymous Coward]

How can a 20 MHz computer send data back at a 100Mbps rate?

Re:NASA is run by idiots...

[egomaniac]

um ok yes my system is not space worthy, but what I said was

"at least used a design similar to a Dual G4 or maybe even a Dual Xeon or P4"

They are not using actual PPC processors from Apple, there using RAD6000, which is...similar...to the PPC's

NASA used a RAD6000 processor, which is essentially a radiation-hardened version of the POWER chip. It's low-power, reliable, and already certified as spaceworthy. Plus, it's readily available.

Now, you've asked why they didn't use something similar to a Dual G4 or Dual Xeon configuration. I listed four perfectly good reasons, all of which you seem to have ignored. Fine, I'll list a fifth.

5) A radiation-hardened, spaceworthy version of the POWER chip already exists in the form of the RAD6000 processor. Radiation-hardened, spaceworthy versions of the Xeon, Pentium 4, and PowerPC G4/G5, to the best of my knowledge, do not exist.

Now, I realize that you've ignored all of my other arguments, but if I need to explain why they used a chip that exists over one that does not, you're on your own.

Re:NASA is run by idiots...

[greymond]

A) They don't need to play Doom 3 up there. 20MHz is sufficient for almost anything you would want to do on Mars. Why send up more than you need to?

- Why are they getting millions of dollars if they are using chips that are 13 years old...

B) Your computer runs far hotter and consumes far more power than the Mars rovers do. Power is at a premium when you're millions of miles away from the nearest electrical outlet.

- Very true. Why not spend 13 years developing a low power design for processors and other hardware, that could be benificial for both space exploration and the earthen population...

C) The rovers are radiation-hardened. Your system is not. Your computer would last about twelve minutes in space before it locked up. A big part of radiation hardening is using larger (and therefore slower) transistors.

- Very true again, but it's not like they sent an apple 8600 up there, they modified the base design of a particular processor and added some great technology to it to make it space worthy, why could they not figure out how to do this with better technology...

D) It takes years to certify a particular piece of equipment as spaceworthy. NASA isn't going to just pop in the latest and greatest Athlon and assume it will work "because the last one did". That means that anything flying into space is automatically going to be at least a few years behind the curve.

- Correct, but i'm not saying Athlon 64 or P4 Extreme. I'm saying they have crap thats 13 years old, you'd think by 97 they would have said hay technology is moving really fast, lets figure out a way to use our low power technology with incorporating a dual G3 design....

The point i'm trying to make is NASA complains a lot about not having enough money, even though the government has donated billions to them over the past decade, and I and others expect more from NASA than pretty pictures for the amount of time and money put into their efforts.

radioisotope generators and environmental impact

[javaxman]

Believe me, environmentalists haven't missed the fact that NASA and the military are sending up buckets of plutonium strapped to rockets. It's just that they've tried and can't seem to do anything about it.

When I worked at the JPL, I personally stamped "declassified" on a huge stack of Environmental Impact Report documents ( boss' orders, of course ). Why they were ever considered classified I'll leave to your consideration, but they were. These were in preparation for the launch of the Galileo probe.

The short of it is that some dudes in New Mexico played with the RTG units and a rocket sled, then told the feds that these things "most likely" would not rain plutonium dust on Disney's theme parks in Florida, at least not dust small enough to get stuck in your lung and cause you serious health problems. In a technical safety-ratios sort of way. OK. It could happen. Not terribly likely. And the heat tiles on the shuttle are safe, too, won't break. And nobody is going to hijack a plane just to fly it into a building.

Yea. We should all trust guys who work at a nuclear test lab to provide unbiased data about plutonium containment saftey. Is your sarcasm detector working ?

I don't know if your *brain* is working if you're completely unconcerned about plutonium being strapped to giant rockets and blasted into the stratosphere, given the probability of a launch mishap. Of even more concern are low-earth-orbit military birds which are going to "safely" burn up in the atmosphere. Like meteors... oh... wait...

Don't get me wrong, I actually _want_ NASA to launch these things. I'm less excited about the larger, more likely to re-enter ones the military has/is/will be putting over *your* head...

Why does it need an RTOS?

[lnjasdpppun]

There has been a lot of talk of what RTOS to use in other comments. I'm wondering why a rover needs to use a real time OS? From my thinking there is a delay of a few minutes with control commands anyway so whats a few milliseconds ontop of that? I don't think things would be happening that fast even for the rovers automatic movements that a second or so of delay in the OS would cause any problems.

Of course in the Slashdot tradition, IANARoboticist so there must be a reason NASA uses a RTOS. Is it becuase they need a small memory footprint and most OSes for embedded systems are small and just happen to be real time? Ie real time is not so much a requirement as a side effect of choosing a small size kernel?

Can someone enlighten me please?

Re:Why does it need an RTOS?

[Anonymous Coward]

Sure, I'll take a stab. Here are some thoughts on spaceflight software and why realtime is needed. (I'm not a FSW developer but know people that are).

One reason is the always severely constrained environment. You can't just send up a 2Ghz Athlon with gigabytes of memory - explained elsewhere in this discussion - space would physically kill it. So you never have a huge amount of CPU horsepower or memory to work with - the computer is always very primitive, very small, very slow compared to a decent PC. And RTOS's are designed for this kind of environment. But that's not the main reason to use them for space hardware.

As you said, it takes several minutes for commands to be radioed from Earth to Mars, but during those minutes, the rover is doing a whole buncha stuff. It's kind of like your Linux or Windows machine - it's actually doing a whole boatload of stuff even when you're not doing anything and the machine appears to be "just sitting there".

Similarly, the rover's computer is doing many, many things all the time, things that require precise timing and low latencies. It has to concurrently save off data from science instruments, performing I/O between buses and memory/storage devices, listen for commands from Earth, send data to Earth or a Mars orbiter, keep track of which way to point its antennas, run stored command sequences, and run self-test routines on all its subsystems, not to mention operating the hardware to drive itself around, while trying not to get stuck or drive off a cliff, and do all this in a robust and fault tolerant manner.

At any given time, a number of activities like this are occurring. It is the continuous, concurrent nature and tight timing requirements that makes realtime necessary. For instance, I/O between instruments and storage devices. The bus protocol will have certain timing requirements that the computer has to meet while doing all the other stuff.

Keep in mind that there is a fixed amount of operating memory available, so things like I/O buffers (of which there will be many) are statically allocated and limited in size - if the OS or the rover's application software takes longer than expected to do some task, data will start backing up, I/O commands will get retried, data will eventually start getting dropped, and things will go bad if the problem isn't corrected - probably this would end up with a reboot and dropping into safe mode.

The key is, if everything does what it's supposed to, *when* it's supposed to, things go smooth. When operations take longer than they should, you have problems. Processor time, like memory usage, is strictly budgeted and you just can't afford to have _anything_ run that you didn't plan and design into the system.

Example - on a project I was involved with (a huge realtime system), there was a problem with the data collection server that was driving the designers nuts. The processes would periodically pause for brief amounts of time - just long enough to blow the time budgets for the network communication. Really gave them fits. It turned out the problem was the OS's virtual memory paging routine, periodically waking up to flush things to disk. They were using a soft realtime OS, and this use of CPU time by the OS wasn't technically out of spec, but it really gave them trouble. (As I recall, they asked the vendor "can we turn it off?" and the answer was "no").

Hope this helps...

Re:NASA is run by idiots...

[egomaniac]

- Why are they getting millions of dollars if they are using chips that are 13 years old...

Because it takes millions of dollars to develop a lander and rover, strap it to the top of an enormous incendiary device, launch it into space, and bring it safely to rest on a target millions of miles away. Relatively little of that cost comes from the actual hardware of the rover, and of the cost of the rover, the computer chips that run it are certainly a small fraction.

Why not spend 13 years developing a low power design for processors and other hardware, that could be benificial for both space exploration and the earthen population...

Since when is NASA responsible for developing microprocessors? You're complaining about them spending too much money, and now you think that they should be researching new processor designs too, instead of just buying an off-the-shelf processor?

Very true again, but it's not like they sent an apple 8600 up there, they modified the base design of a particular processor and added some great technology to it to make it space worthy, why could they not figure out how to do this with better technology...

No, they didn't. You seem to be missing the fundamental point that NASA did not design or manufacture these microprocessors, any more than Dell designed the Pentium 4. Just because you can crack open a Dell and find a Pentium 4 inside of it does not mean that Dell had anything directly to do with its creation.

Correct, but i'm not saying Athlon 64 or P4 Extreme. I'm saying they have crap thats 13 years old, you'd think by 97 they would have said hay technology is moving really fast, lets figure out a way to use our low power technology with incorporating a dual G3 design....

You still haven't answered my question -- why? What difference does it make? What science could they do with a dual G3 that they can't do with a RAD6000?

A 20MHz chip is perfectly sufficient to do everything the rover needs to do. Why do you give a crap what kind of processor it runs?

In case you hadn't noticed, almost everything nowadays has computer chips in them. Cars, microwaves, washing machines, Mars rovers. Very few of these devices have a need for anything approaching the computational power of a modern desktop. I don't care if my washing machine runs a dual G5 any more than I care if the Mars rovers do. I just care that they can do the job they were designed to do, and do it well.

You know what? When you're designing a Martian lander, feel free to put in a dual PowerPC G4. Until then, maybe you should assume that the people who managed to design and build this machine and send it to a different fucking planet safely, might just possibly have some idea of what they're doing. The phrase "rocket science" didn't come to mean "really insanely complicated stuff" by accident. These folks are very, very smart.

Mars Rover. didn't they forget something?

[Anonymous Coward]

I guess they didn't include any aerogel on the rovers, save the Mars Pathfinder. Here's a better link about aerogel, which I believe was an earlier post on /.

http://stardust.jpl.nasa.gov/tech/aerogel.html

As a former Windriver employee...

[PinkyGigglebrain]

I can give you some answers.

The fire wall is easy to set up, but the cost of the single VxWorks license would put you in hock for a year or so.

The OS itself needs less memory than you would think. For the rover I would guess the OS image is about 500K to 1M, but that doesn't say anything about how much memory its really using for the OS since it can run out of FLASH and leave the 128M of RAM for the stacks used by the tasks (individual RTOS proccess). I would imagine a lot of that RAM is being used for holding the images while they are being sent back to Earth or evaluated for navigation info.

The PowerPC its using is just a radiation hardend version a common CPU, I don't know which, and you can do a lot with a 20Mhz PPC. The training department used 860 PPC/20Mhz in the classes and never had any issues with task execution times, at least not until the system clock was cranked up to absurededly high speeds.

Unfortunatly for hobbiests and hackers everywhere Windriver has never came out with a "Personel" or non-comercial version of their software/license packages, despite one of the instructors pushing for it every chance he got. Managment at Windriver didn't go for the idea, in fact when they started shuting down the training department that instructore was the first to go. So much for Wind River encouraging inovation by the private developer.

Bottom line is that VxWorks is a great OS, versital to the point that someone even ported Doom onto it and ran it on a Kodak digital camera. However, in my opinion, the company is a bit lacking in the area of inspirering creative inovation and development by anyone who is not representing a MAJOR company/organisation like JPL or Boeing.

If you really want to try out a firewall on a real-time OS give QNX a shot, they have (last time I checked) a free development tools package that included a non-comercial "hobbiest" license for x86 CPUs and it supports USB, ethernet, openGL, and just about everything else VxWorks does.

That thing...

[vile7707]

got a hemi in it?

Yep...

[Anonymous Coward]

Another good reason to look down on WindRiver.

haha

[Anonymous Coward]

Slim Jims and black stuff

Re:I hope the flash memory was not commodity hardw

[fshalor]

The day we have reconfigrable hardware will be a good one. (And I don't mean just redoing the control system's code and the operations.) But that day is yet to come. For now, from an industrial engineering standpoint having two birds with the same configuration is the same as multiplying the probabilities of them both failing together: it just makes sense.

Now, if they were different...that would be additive, and thus the overall chance of having a catastrophic failure would be higher.

Okay, I'm sick now, I've just realized it's actually painful to use an OS with the ctrl-a -e -b -f -n -y -v things disabled... THis is like the 10th time I've brought up the darn composer window in mozilla. Darn windows... Back to the ibook for more slashdotting!

The quest for other planets, stars and black holes

[fredrikv]

When VxWorks today was reported to be the premier operating system on Mars, Microsoft officials immediately announced that the next version of Windows, codenamed "Longhorn", is developed specifically for Uranus.

the environmental concerns over RTGs

[linoleo]

If obsessed environmentalists don't like NASA sending up probes with any radioactive material ('it might blow up, ohh..'), then how did this little tidbit get by them?

Give us "obsessed" environmentalists some credit, 99% of us can differentiate quite well, thank you:

1) Each Mars rover's RTGs contain less than an ounce (8 LWRHUs with 2.67g each) of Pu-238; Pathfinder's a quarter-ounce (3 LWRHUs). nothing you'd like to swallow, indeed: lethal dose for Pu-238 is 30 micrograms inhaled. It's one of the most toxic, mutagenic, and carcinogenic substances known to man. While an ounce of such stuff is not trivial, it pales next to Cassini's load: 10.8 kg (24 pounds) of Pu-238. This should suffice to explain the different level of concern generated by these missions even to cyclopedian dunderheads.

2) As for blowing up, I have some news for you: Challenger. Columbia. NASA's own estimate for the probability of "catastrophic launch failure" of the rocket used to send up Cassini was 1 in 20. Sending things into orbit on top of the world's largest firecrackers is *inherently* unsafe. Might blow up, indeed. Even so, catastrophic launch failure was *not* a major concern: the Pu-238 is embedded in a ceramic matrix, and its container designed to withstand explosion of the launch vehicle. NASA worst-case scenarios were a contaminated launch pad.

3) What *was* cause for concern were the Earth gravity-assist flybys. A navigational error (metric/english conversion, anyone?) could have sent Cassini to burn up in the atmosphere. In the intervening years in space, its own alpha-rays weaken the ceramic matrix to the point where such an accident could release powdered Pu-238. Worst-case scenario for *that* was as much release of Pu-238 as all above-ground nuclear tests, ever, *combined*. Not Armageddon, but a Chernobyl-scale disaster in terms of the number of statistically attributable additional cancer deaths.

4) The concentration of so much lethal potential (300 million doses) in *any* one device of whatever nature raises certain concerns. Pu-238 is *the* material of choice for dirty bombs - 250 times as carcinogenic as the Pu-239 used in nuclear weapons. The RTG requirements of NASA and DoD are the only reason why Pu-238 is produced and stockpiled in the US. A single stolen 150g pellet of Pu-238 could be used to give up to a million people in a chosen location certain lung cancer. Nice terrorist potential there - 9/11 in slow mo. Cassini carried 72 such pellets. I sure hope security around these things is 100%. Wait a minute, there is no such thing as 100% security - doh!

Having said all that, I (environmentalist and all) still think that, for lack of better alternatives, the outer planet missions and the small Martian RTG heaters are worth the risk. IMO NASA actually does quite a good job evaluating these risks for each mission (thus no juiced-up Mars buggy, ever - sorry!), it's the DoD putting RTGs in LEO that worries me.

Finally, to the smartass trolls spewing baseless comments such as "more radioactivity in my backyard", "radiation is radiation", "we tested nukes in the atmosphere in the 50s with no ill effect": 1) you're utterly wrong, 2) think before you post. If thinking is too hard for you, at least google before you post. If googling is too hard for you, shut up and learn something.

- nic

Re:NASA is run by idiots...

[bala_dei]

why dont you sheild total motherboard or total electronics involved insted shield the processor alone?

They do malloc() at startup

[dvd_tude]

See this slideset, in particular the one titled "Memory Allocation and Usage" (slide 18 I believe.) They do all their allocation at init time.

So far I see lots of flamage about how VxWorks malloc() sucks, but no concrete explantion why (if?) the number of files caused the crash. Also, there isn't any indication (so far) that the FAT was hosed.

So... what was the real problem? Was it indeed a malloc() failure that led to the reset loop? If so, what was being malloc()'d?

I'm really curious about this, since a project I worked on uses FAT16 on flash; in my case it was a multimedia terminal running Thread-X in 32MB RAM and 64MB Flash, with third-party DOS FAT and TCP/IP add-ons. We had lots of files on this thing, and it didn't seem to flake out like this.

Re:NASA is run by idiots...

[JGski]

Fair question.

The problem is that space has multiple types of radiation, and shielding that protects from one type can actually increase the radiation exposure due to another. Simply putting everything in "thicker box" will make overall radiation effects worse and add more weight to boot.

This radiation problem (for both electronics and humans) is one of the major reasons a mars mission wasn't jumped at after the moon landings in the 70s (aside from the bad economy at the time). It's one of the likely factors that could doom the Bush mars project. The astronauts will be getting a significant radiation exposure and there's probably nothing that can prevent it - it's possible the exposure could be enough to be fatal, especially if there is a solar flare at any time during the mission. See Mitchner's book Space - it's depiction of the radiation risks in human space travel is very accurate.

JGski

Re:But The Question is..

[McPierce]

Didn't you read the article? It says right there that it runs VxWorks...

Re:space shuttle uses 1969-vintage ibm 360 compute

[Anonymous Coward]

My computer is faster than the space shuttle =)