The collected works of R-ten-K - Page 2

porter wrote:

Nai wrote: Has anyone emulated irix-mips

The PlayStation One emulators do R3000 well.

I seem to remember SimOS at Stanford did MIPS pretty well and had enough SGI HW emulation in to to be able to boot old versions of Irix (I think it was used in the Splash project, what eventually led to cc-numa so Stanford got a lot of info on SGI HW internals, the early prototypes of splash were in fact a bunch of 4D machines glued together).

I was told that the simos people eventually founded vmware based on the technology.

Don't know if it is active (I doubt it).

For mips emulation, here is a repository I have found helpful.

http://www.linux-mips.org/wiki/Emulators

LOL we submitted at the same time about SimOS.

From what I read from you about HP-PA it sounds like the architecture is fairly ugly... it would be interesting to find out why the #$#@ they decided to make those privileged ops such a PITA to deal with. The overhead, in instructions, that introduces in those handlers must have offset the simpler HW (I assume that was their goal). The OS people must have shitted some bricks, but then again maybe that is the reason why HP-UX was so "unique." HP sure did have some odd "isms."

Wow. I did not know they had commercialized simics. I knew one of the students involved when it was an academic project (still is?) back in Norway (or Sweden?). As the previous poster said under most configs, simics used to be slower than molasses. I assume it still is, because the folks at CMU are currently trying to speed up simics (and possibly other full system emulators) by offloading some of the processor emulation and timing models to their version of the ramp HW accelerator (a board with a bunch of FPGAs interconnected). They used to have some public systems running a frontend executing simics, and 16 SPARCS accelerated via FPGAs. It was quite interactive, pretty neat since it is a large multicore system fully emulated (normally you would be lucky to get thousands of instructions emulated per second in an architectural simulation of so many cores, basically a 2 second benchmark would take a day to run).

If you are interested, you can get QEMU to model a full sparcstation (maybe they have included ultras by now) and can boot solaris off it.

If you can access some back issues (as in 80s vintage) of IEEE Computer magazines, you will read some very candid articles which include the design decisions/trade offs behind some of the major architectures. Attending some of the panels at ISCA/MICRO/ASPLOS, ISSCC, or hotchips is always a hoot when interacting with industry people.

From what I have read, It seems that the HP-PA was truly made by committee: it was based on a previous CISC effort, had to support their legacy minis, and someone decided to bolt on the latest "RISC" fad to make it buzzword compliant. But it always seemed like an odd RISC. I think the feature I remember was their odd cache strategy: usually they implemented a single very large cache level. Which from what I was told, was probably because they had some serious issues with register spills. And now that you brought the handler issues, it makes a lot of sense to have a big honking cache when so much overhead is expected when servicing interrupts, exceptions, etc.

I also remember that at some point, the next generation Amigas were supposed to be based around HP-PA? It must have been weird to be motorola in the 80s, everyone leaving you for yet another RISC vendor.

Itanic technically it is a dog, however it seems that in hindsight it must have been a marketing tour de force since it did annihilate the competition: AXP, MIPS, and to a lesser extent SPARC/POWER. Which I guess, under Intel's POV, it must have been a worthy investment.

I can't believe the IA64 people decided to implement register windows. When I was in grad school, one of the projects I had to develop an out-of-order scheduler for a machine with register windows (SPARC). And it was a massive pain in the ass... I was impressed when I read that the itanium people had managed to get an out-of-order implementation internally, but I assume it is so complex as to not be worth it.

I liked the concept of predication, but now it is clear that it is a bad idea from a power consumption standpoint.

Power may not be around for long, I think most of Power7 is subsidised by DARPA. Power6 is interesting, in the sense that IBM decided to ditch out of order and go for sheer speed. Still, they do some really strange things to that architecture, like the decimal ALUs (ugh).

If anything, the only architecture I am looking forward to see release soon is ROCK from SUN. It is mighty late, but I am interested in seeing an actual implementation of transactional memory. Although, after playing with nehalem... I hope the sun people get it right.

LOL April fools early this year?

hmmmm... 8Wx5000= 40Kw




Never mind the 40Kw in a 3U factor... I wanna know how SGI thinks they are going to package 5000 atoms in the volume of a 3U


Sad that they are coming down to this though... come on...

A fully loaded blue gene takes significantly more space than a 3U rack.

The packaging tech needed to get 10K cores in there ain't trivial. Besides, where are they fitting the memory, what about interconnection, etc? Is this coming directly from SGI?

Sweet system! Lucky b*stard...

The 900Mhz R16K must be rarer than hen's teeth... how did you manage the score?

recondas wrote:

jan-jaap wrote: What makes it special are the looks of course. It also has a couple of connectors on the mainboard you won't find in a regular Fuel .

That's a beast that's rarely <if ever> been seen in the wild. If you're willing you might see if Pete would allow some detailed photos in the Gallery - for the historical record.

I may be having a brain fart here... I have seen the exact same case skins internally inside SGI, but it was a generic x86 ATX machine.

recondas wrote:

R-ten-K wrote: I have seen the exact same case skins internally inside SGI, but it was a generic x86 ATX machine.

You aren't the only nekochan reader to have had the opportunity to have been "internally inside SGI", but you're certainly in the minority.

I never claimed I was. The prototypes I was referring to: having identical bezel shape to the fuel, greyish skinned, but x86 feecee internals. I don't know enough to know if they were released in the wild as actual products or not. That is why I was referring about seeing them only inside SGI.

That being said, following hamei's... it seems the chassis and the side skins of the x86 330/550 are pretty much the same as the SGI fuel. So maybe those are the machines I remember seeing... who knows. My recollection is that the bezel was definitively solid like the fuel though, and looking at the blue fuel that originated this thread was that reminded me of those feecees as looking identical to jj's system. But now, I am all confused in my twilight years...

Also, it seems that mechanically the case of the fuel and the SUN Blade 1500 are almost the same chassis. Does anyone know if this was some OEM design popular at the time? I am asking because I have some fuel misc boards laying around... but no chassis. I can have access to a dead blade 1500... so maybe the old swaperoo may work out, with some help from Mr. Dremmel.

How similar are the Origin 300s to the 350s?

I may have access to some 300s.

I don't consider $20 to be lots of money

Those boxes w/o host interface and software are pretty much useless....

japes wrote: The damage from a leak would be the mainboard though, not the powersupply. Though it sounds like the powersupply might have failed. I think there should be a LED somewhere on the board to indicate standby power.

From the picture it doesn't look like there's residue on the bottom of the case though. If there is a leak it probably can be cleaned up and repaired.

I had a dual G5 fail due to a leak in the cooling system. The power supply in the G5s sits underneath of the processors. The leak in my system affected the Powersupply and not the mainboard, probably due to gravity being what it is... Alas AppleCare replaced the whole system.

Also, the liquid cooling system and the processors are one monolithic package. Meaning, you have to replace/exchange the whole processor/cooling module. You could service it, if you are brave, but it did not seem (when I inspected it) that Delphi had in mind small repairs to the system when they designed the cooling That was for the dual 2.5Ghz G5 model, I have no clue if that is the same case with the quad G5 module.

Edit: looking at the pictures, it seems some of the system fan trays are missing. Did you remove them to take the pictures? If so, disregard this comment. But if you did not, the G5 will not boot or even turn on sans all the system fans present.

I have been using the new aluminum BT keyboard from apple, and I am pretty pleased with it. It is not everybody's cup of tea, but I prefer the ultra low profile to the split keyboards. It is very small so I can place almost anywhere on the desk...

Input devices (and monitors) are things that I don't mind spend good monies on, since I have to spend a significant portion of the time using them. Ironically, computers themselves are becoming commodity, to the point that once I find a monitor/keyboard/mouse combination they tend to outlive multiple CPUs (I still have my kinesis keyboard in the closet, that thing lasted me for years... alas I don't think I have a PS2-ported computer anymore).

If you can get one of the classic kinesis keyboard and pedal sets, they are built like a tank... and they do make a difference. They are not too noisy, but they are quite pricey.

I used one for years, and I highly recommend them. The layout is a bit awkward at the beginning, but within less than a week one gets used to it.

Awesome job guys! I can't wait to try out the 64bit version to get some of that old 4DWM feeling back...

Honestly, unless you get it very cheap (or a song) it is not worth it. 10.5 will be the last OSX to run on them, and the batteries tend to show their age. I still use my PowerBook 15" every now and then, and for most basic tasks it does OK.

However, if small form factor is your thing... you can get a cheapo netbook and put OSX in it, or you can get one of the new 13" unibodys for around $1K (less with academic discount). It is slightly larger but will run circles in terms of CPU, resolution, and battery life.

I always advice against making any significant investment on a product that has been EOL'd. That been said, if the price is cheap... go for it.

ajerimez wrote: I wish some enterprising company would come up with an alternative to Flash player that's actually optimized for the underlying hardware. It can't be that hard!

A G4 has plenty of horsepower to do video w/o any significant issue (probably not HD, but normal 480p stuff should be no problem). The issue is that to get the required performance implies using altivec (or any other SIMD extensions)... and it is nearly impossible to find programmers who understand data parallelism correctly. Media-oriented ISA extensions have been around for over a decade in full force, yet there are still few tools/people who can exploit them properly.

People mock me when I stress the importance of proper technical education, esp. with the increasing levels of complexity in modern computing systems. So it should not be a surprise that a lot of products out there pretty much suck, because a lot of people out there don't know what the heck they are doing... literally.

I assume whatever it was, it involved one blue screen of death too many, a six pack of beer, and a loaded gun nearby. Not necessarily in that order.

If XP does everything you need it to do, then there is no reason to upgrade.

For other users, probably is more a function of newer HW support. Especially things like DirectX11, DirectCompute, some of the new VM stuff, etc. However, I doubt there will be much software out there to take advantage of that at least 1 year from now at the very earliest.

I have a Windows 7 media center machine, pretty neat. My only gripe with it is that netflix support has not been added yet, and I would like Amazon to actually make an effort to integrate their video on demand service with the windows media center. At work, we're still using XP for those having to use windows. Vista was almost skipped completely, I assume a sh*tload of heads must have rolled at microsoft for the Vista fiasco.

pierocks wrote:

iKitsune wrote: Ford Crown Victora Police Interceptor. Fun, fast, safe, and affordable by mortals. Besides, it's like driving a land-boat, the last of the great body-on-frame rear-wheel-drive American gashog monsters.

This is exactly what I drive Fantastic car...it's an absolute tank and it hauls ass!

A close friend of mine in college got one, in black no less, as a hand me down from his grandpa. It was awesome to drive it in LA and see the traffic slow down around it, since a lot of drivers would get disoriented for a few seconds and think it was a cop car. I have never seen so many taillights flashing at once for no apparent reason.

We had an awesome halloween when we drove around in it dressed like 70s cops en route to a few parties that day. We even jury rigged one of them detachable "detective" flashing light on the sucker.

It drove like sh*t though.

As far as supercars are concerned, I'd take an Aston Martin One-77 any day.

Huh?

I thought Sony already announced they were going to make their development kits available for free for the PSP? I thought the PSP Go was selling quite well. Not that I particularly care, since I am not much of a gamer.

porter wrote:

R-ten-K wrote: I thought Sony already announced they were going to make their development kits available for free for the PSP?.

They said they'd drop the price to 10% of what it was.

I think the problem they will have is you can write as much code as you want, you just can't load it or run it on the PSP.

I was a tad confused because there are already some homebrew kits for the PSP, I had no clue that it was impossible to load stuff onto them.

http://www.develop-online.net/news/3277 ... P-dev-kits

I hope the days of certain companies thinking that developing for their platforms should be treated as a privilege are numbered.

Do they support ZFS boot volumes yet?

Good, I have been following opensolaris mostly. So I had little clue where Solaris 10 was at.

(SUN's message wrt Solaris is rather unfocused me thinks).

Funny you say that, since I think that is how I initially ended up with OpenSolaris.

wow! Pretty cool stuff I must say.

That looks like a serial or FDDI concentrator board.

Follow the instructions in this page:

http://dev.maxxdesktop.com/trac/maxxdes ... tion_Guide


It was easy as pie last time I tried (on a fedora system at least).