Slashdot Mirror
Hot New Silicon Graphics Workstations
Jonathan C. Patschke writes: "SGI have finally unveiled their newest-generation visual workstation, the Silicon Graphics Fuel. Features include a MIPS R14k CPU, Vpro graphics, and a PCI bus (finally)." As you would expect from SGI, it looks good,
and the specs are impressive. I only see IRIX listed, but with the
specs on this thing, it may not be slow :)
These boxes look great but everything I read about the new render farms show that people like dreamworks are all switching to large linux render farms and SGI just for front end or no SGI at all and using the new big G4's from apple.
faster Mac OS X machine? No way dude. If you are looking purely at mhz, yeah, the Mac is faster, but the architecture is vastly different. The address bus and memory bus are larger, and even if most people think IRIX is a pile, it was designed for graphics i/o.
:)
:)
I wouldn't mind having one of these, but I wish they would bring back their old logo
Want a really fun machine? Get the Origin 2800 w/ 250 CPU's
--Jon
Or a G4 1GHZ DP, with 1.5GB RAM, 2x2MB DDR Cache, 3x72GB Ultra160 SCSI and a 22" Flat panel, oh, and another $2500 to buy a real Video Card. The Apple will conveniently run those wonderfull Alias Apps too. I suspect a G4 1GHz DP will out perform a Blade 1000, let alone that MIPS R14KA@500MHz.
The Crazy Finn
"You've got an invalid haircut" -Warren Zevon - Life'll Kill Ya
Fuel isn't based on UMA, it's based on the same exact set of ASICs that powers the Origin 3000. This is basicly a 1 CPU version of O3K. Compared to Octane2... Fuel has 3.2x RAM, 2x faster CPU bus, 1/2 the interconnect latency, plus a faster SCSI bus. Neat stuff. Not to mention that VPro graphics are 48bit, compared to the 32bit you find elsewhere. Film people like that.
I only see IRIX listed
That's becuase this is their latest MIPS system, not some x86 box. Despite some progress, Linux does not really run on SGI MIPS boxes. And some of us like IRIX just fine, thank you :-)
Do you even know anything about perl? -- AC Replying to Tom Christiansen post.
Here are some numbers from Spec.org
This is for Spec CFP 2000 (i.e. floating point). I picked the SGI Origin 3200, which has a similar processor (although I'm not sure if its identical or not).
Advanced Micro Devic Epox 8KHA+ Motherboard, AMD Athlon (TM) XP 2000+ 1 596 642
Intel Corporation Intel D850MD motherboard (2.2 GHz, Pentium 4 pro 1 766 777
SGI SGI Origin 3200 1X 500MHz R14k 1 436 463
The Spec CINT 2000 numbers look similar, I just didn't feel like cutting and pasting.
So, sure your average P4 or Athlon is faster, but its not as simple as a matter of Megahertz.
My concern about SGI is that these machines have the same graphics V10 and V12 that they've been using for years now. I heard that these were designed as the last hurrah of designers who have since gone on to Nvidia or ATI.
I wonder if SGI has the manpower left to design new, innovative graphics architectures, or will they be just slapping more texture and cranking the clock on old designs.
dave
A faster Mac? Please.
This thing looks to have the same terrifying memory bandwidth as its
big brother, the Octane2. 3.2GBps. On a dedicated port crossbar.
The Mac is STILL struggling along with PC133 SDRAM. And the Mac has a
"Geforce4MX", which is basically a faster GF2MX, not a fourth
generation part. Compare that to the SGI graphics subsystem for a
laugh.
For processor bound tasks, yes, the 7455 G4 will be faster than the
R14k, but for overall system performance, ESPECIALLY when pushing big
models around, you'd be goofy stupid to try and use a Mac if you could
afford one of these babies (to say nothing of the Octane2).
Peace,
(jfb)
To spur "enterprise Linux," Big Bang, the distributed two-phase commit.
Am I the only one who thinks a 600mhz CPU w/4mb of L2 cache can smoke a 1+ghz x86/POW whatever w/32/128k of cache? And it really does 64bits, not just considers it, yippy! If any body read the article these are for front-end workstations, NOT render farms. Yes the big guns in hollywood use Linux for render farms, but the workstations are still SGI. Why, becuase PC's still suck at COMPLEX 3D. They are getting better, but they are still ghugging on stuff an SGI whips through (higher data throughput and bus speed mean A LOT!) Besides, why would you spend 10 grand on a high-end graphics station and use it for a render farm? You'd never use the graphics capabilities in a render farm, that's 100% CPU crunching. You gotta love those HUGE mips CPU's. You could make a really cool toaster(for bread not video) with a couple R5000's stuck side by side! Thank god SGI stopped making those stupid NT boxes with the reversed PCI slots and custom memory! Let's hope they didn't integrate the graphics on the Mboard this time. A company I worked for actually bought a bunch of SGI 320's in '99. And when the company went bust 10 months later, we all got $5,000 unupgradable paper weights as a consolation prize.
No, the Octane2 and Fuel are 12bit per RGBA, and the PC world 8bit (48/4 = 12, 32/4 = 8)
This is obviously important when you can buy 36-bit and 48-bit RGB photo scanners.
Hands in my pocket
wonder if SGI has the manpower left to design new, innovative graphics architectures, or will they be just slapping
more texture and cranking the clock on old designs.
I've been told that a speed boost along the lines of a "V14" and "V16" will be available in May, with a totally new gfx line (compatible with existing machines as just a new gfx card) becoming available this fall.
Then, of course, there are neat new SGI gfx offerings such as Onyx InfinitePerformance...
http://www.sgi.com/visualization/onyx/3000/ip/
First of all, the Fuel workstation is sort of a cool new evolution... it uses the existing V10 and V12 graphics from Octane2, and the chipset from a single Origin 3000 node. This means instant software compatibility and one hell of an awesome base to run future graphics and CPU offerings. Compared to a single CPU Octane2... Fuel has *half* the latency, *3.2x* the RAM thruput, and *twice* the CPU interconnect thruput. And it run the same OS and the same apps. All for about 1/3 to about 1/2 the price. Sounds like a pretty resonable update to me. And an Octane2 ain't too shabby for real-time interactive apps, either. If you haven't already, find one to play with. A VPro-based Octane running IRIX 6.5.12 or newer is a 3D beast, and yet rock solid stable. Even makes for one hell of an uncompressed, realtime HD video solution, if you can afford the RAID and HD interface. I've never seen a PC or Mac HD solution come even close to Octane2. And Fuel is that much better...
/ 3000/ip/tech_info.html.
Folks run IRIX for HD video editing, effects compositing, and 3D modeling for a reason -- it works and it doesn't have the "crap out" effect when working under a huge load. Sure the CPUs in an SGI aren't extremely powerful, but that doesn't matter much -- it's the crossbar switch architecture (Octane/Octane2 is based on Origin 2000, Fuel is based on Origin 3000) and wide busses that make the difference. Batch jobs and long haul rendering is all done on a farm of cheap PC's anyway (unless you're ILM, which owns six Origin 2000s, each with 128 CPUs).
Secondly, SGI is coming up with some way cool graphics offerings. In my opinion, the new Onyx InfinitePerformance graphics is bigger news than the new workstation:http://www.sgi.com/visualization/onyx
SGI screwed up big time in the past, but they're working on fixing the situation. They can't do everything at once, but they're working as hard as they can. They're a pretty wide spread company. Hell, they even own Alias-Wavefront (ever heard of Maya?). They're doing some other cool things, too. Their developer program is now free to commercial developers, but hobbyists with a real project are invited as well in a case-by-case basis. They're even giving away a Fuel workstation at the SGI Global Developer Conference next month. And it's not just a drawing, either. The winner of the machine will be a hobbyist with an attendee-voted best project. Very, very cool stuff.
http://www.sgi.com/developers
http://www.sgievents.com/developer2002/
See any Dual P4 "Xeon" motherboard, like this one:
8 60 /P4DCE.htm
http://www.supermicro.com/product/motherboards/
Tyan is offering a Pentium 4 Xeon board currently. So technically it does have the P4 moniker but it is also a Xeon so I guess it isn't a full-fledged P4.
Looks like SGI should consider joining Apple in the PowerPC world...that Power4 looks pretty awesome!
That Power4 also costs like $100,000 for each (4-way CMP) processor module alone, so, gee, it'd better be pretty awesome. The 1 GHz G4+ that powers the current generation of Macs would probably score about the same as the R14k on SPEC, or a bit lower...but we don't know because Apple is too cowardly to submit themselves to legitimate benchmarks when they have a bunch of fools running around believing that a G4 is faster than a P4 or Athlon, and Motorola doesn't bother because they know the G4+ is actually designed for the embedded signal processing market, where SPEC scores are not too relevant. Just because the G4 and Power4 are both "in the PowerPC world" doesn't mean they have similar performance characteristics.
In any case, where the R1x000 really shines is in scalability to very high processor count NUMA configurations (not at issue in this case of course). It'd still be a world-class processor line if SGI hadn't given up 5 years ago by essentially stopping R1x000 development and committing to Itanium instead. They've finally realized their mistake and apparently have some extra tweaks on the way (R16k and R18k), but it's probably too little too late.
Were I SGI at the moment, I'd drop IRIX for Linux, port everything that made IRIX special, and run it all on proprietary P4 or Xeon boards with all the special SGI graphics goodies. Although that was the idea behind their NT line and that didn't do so well, did it...
SGI had some amazing tech back in the day, but having more or less rolled over and died the past few years it might be difficult for them to stay ahead of the commidity hardware crowd. (Re: 48-bit color, if johnc has his way--and he usually does--commidity graphics cards will have 48 or 64-bit internal color soon enough.) But they appear to be finally waking up and making a go at it, so best of luck to them.
Yes it is: http://www.sgi.com/fun/freeware/games.html
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
That $100,000 cost is fairly meaningless, since there is an extreme markup on server hardware, and the chip isn't in mass production...I'd venture to say that it can be mass-produced cheaper than P4, as I'll bet it has a lower gate count.
Yes and no. Sure the HPC market where the Power4 currently plays has huge markups and very low production volumes...but that also means designs which could not possibly be cost effective in the desktop market. A single Power4 multi-chip module contains 4 2-way CMP dies, 256-bit interconnect between each pair of dies, and, oh yeah, a measely 128MB of eDRAM.
Each one of the 4 dies takes up 400mm^2 on a
G5 will essentially be this architecture.
The G5 is an upcoming 32-bit embedded chip made by Motorola (like the G4 and G4+), and does not resemble the (64-bit) Power4's internal architecture in the slightest. Whether this chip will be the basis of the next generation of Macs is of course not yet known.
Please cite some reference to support this (wild in my opinion) claim.
Because Apple does not have the integrity (nor, according to the oft-repeated excuse, the FORTRAN compiler) to submit SPEC runs for a G4-based computer, there are no official SPEC scores for the G4. However, we do have Motorola's *estimated* *SPEC95* scores for the 7450 (a.k.a. G4+) at 733MHz. (Here, second page, on the left.)
They are 32.1/23.9, SPEC95 int/fp. By comparison, a 400MHz R12k (best I could find for SPEC95; it is an old benchmark after all) scores 24.2/43.5 SPEC95 int/fp; 25% worse on int, and 82% better on fp.
That same 400MHz R12k scores 347/343 on SPEC2k int/fp. (Sorry, but no more links; the scores are all available at www.spec.org) Assuming equivalent SPEC95-to-SPEC2k ratios (a faulty assumption, but then again we're using estimated scores in the first place), we get our 733MHz G4+ scoring 460/188(!!) on SPEC2k int/fp.
For a scaling factor we'll use the Coppermine PIII, since it has SPEC2k scores available for both 733MHz and 1GHz configs. 1GHz is 22%/16% faster than 733MHz at SPEC2k int/fp. (If you repeat my calcs, be sure to use the 1 GHz PIII scores using the same compiler version as the 733MHz scores.) So applying that to our "estimated" SPEC2k scores for 733MHz G4+, we get even-more-estimated SPEC2k scores of 563/219 for a 1GHz G4+.
So, a decent spot (32%) better than the 500MHz R14k at int, and a significant bit (53%) worse at fp. Plus the CPU in the new SGI Graphics Fuel can be up to 600MHz and uses DDR and not SDRAM like the one I got the scores from.
So...hope that helped.
Re: the Power4 SPEC scores(Also this was a single-CPU system, so I don't think it was a multi-CPU module.)
SPEC2k is single-threaded. The score was obtained using a 4-way Power4 "Turbo" module with 3 of the cores "turned off". The rather sneaky thing is this gave the remaining core access to all 128MB L3, which means the SPEC score probably overstates single-threaded performance a bit.
What makes you think that Power4 technology won't make it's way into desktop chips? IBM manufactures desktop PowerPC chips as well, and certainly shows no sign of giving up on PowerPC in general. There have recently been rumors of Apple switching from Motorola to IBM for it's chips...we'll see what happens.
Power4 is simply not a desktop chip design. Even using one of the 4 dies in the MCM as the basis for a desktop CPU is a shakey proposition, since they're too big (again, 400mm^2 on
Of course, it may be quite likely that Apple turns to IBM instead of Motorola for the next generation of Mac CPUs (especially as it looks somewhat likely that Moto will exit the semi business in the coming year). But it will not look anything like a Power4.
Well, it's not the first time Slashdot misses the point. :-( SGI didn't released just the Fuel workstation today. In fact, that the smallest and most insignificant part of their announcement.
The actual announcement reffers to the so-called Visual Area Networking - a concept that, basically, boils down to distributed visualising and data processing over a network.
With VAN, a user can interact with an InfinitePerformance supercomputer (usually an Onyx 3000 with several hundred processors), let the big iron do the data processing, and receive the resulting images over a network to a thin client. That "thin" client may be a Fuel workstation, a PDA, some device used by US troops to get realtime maps of the enemy positions, whatever.
The point is, many people, working from many different locations, can work together using their thin clients, but manipulating data on the same supercomputer. I've seen some impressive demos, where two people were immersed into the same VR environment, and were manipulating objects on the same scene, at the same time, over the network. Given the fact that the scene was not just a pure graphical computer-games scene, but an actual simulation with real physical laws and everything, that was pretty damn cool.
I tried to submit the actual story, but it was rejected. Instead, Slashdot caught this ridiculous story about "yet another workstation from SGI". Come on people, get real...