Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
I'm sorry if this sounds a little nooby, but why the hell would you need all that processing power? what could you actually do with it that would help you in "everyday" life?

The "everyday" needs of computational scientists might include: molecular modeling, numerical algorithms and software for mathematics, quantum chemistry, molecular static and dynamics, computational geophysics, biophysics, computational fluid dynamics, air quality simulations, computational biology, plasma and gas physics, and weather prediction. :)
 
They should run SETI@home on that... We'd have new life found by tomorrow:rolleyes:

Oh, and they should see how long it takes it to compute pi to the trillionth decimal.
 
Intel's Tera-scale project aims for > 1 teraFLOPS sometime in the 2010s (I think 2015) for regular computers. It will most likely take another decade or so to get 29 teraFLOPS, and maybe another decade to get a handheld device to 29 teraFLOPS. So I'm thinking about 30 years.

That's about the point in time it will take a set of 5 blu-ray DVDs to load MicroSoft Word into your system.

P.S. I still have version 1.0 of Word kicking around my office that shipped on a single 400k floppy. (Free to a good home)
 
Good to see the project get an upgrade, but to "only" be in the top 10 whereas before they were third is a little disappointing. If I recall, the bigger news about this was the cost/benefit ratio since, at the time, the number 1 and 2 spots cost considerably more.

The computer seems to fit the design purpose quite well. There's no more sense in strapping an 800 HP engine in your Focus just so that it'll have the most horsepower on the planet than there is in building a 1000 Tflop computer just so you can say you did. Plus the whole advantage of this kind of design is that it's scalable. There's really no point in building it vastly beyond design needs.

It does amaze me, though, where the top ten supercomputers in 2003 place on the June 2008 list. The progress really has been quite stunning in terms of the ability to build computers that can do very complex computations. The downstream benefit is that it means that these mid-grade supercomputers become more and more accessible.
 
They should run SETI@home on that... We'd have new life found by tomorrow:rolleyes:

Oh, and they should see how long it takes it to compute pi to the trillionth decimal.

(about pi) i believe thats how they measure computer speeds.. at least one of the ways.
 
http://gallery.me.com/acarpenetti#100028

Sorry, I don't have any from when they were the Power Macs. I believe when they were Power Macs, it took up about about six rows of racks. This took it down to 1.5 rows. Huge space savings.

Edit:
I knew there were more pics taken.
http://www.arc.vt.edu/arc/SystemX/initial_gallery.php
http://www.arc.vt.edu/arc/SystemX/upgrade_gallery.php

Nice pics. It'll be interesting to see how everything looks/speeds/technology is another few years down the road.
 
Intel's Tera-scale project aims for > 1 teraFLOPS sometime in the 2010s (I think 2015) for regular computers. It will most likely take another decade or so to get 29 teraFLOPS, and maybe another decade to get a handheld device to 29 teraFLOPS. So I'm thinking about 30 years.

whats the average FLOPS that a computer these days produces?? (im talking an 8 core MP, not a lappy or anything lol).

That's about the point in time it will take a set of 5 blu-ray DVDs to load MicroSoft Word into your system.

P.S. I still have version 1.0 of Word kicking around my office that shipped on a single 400k floppy. (Free to a good home)

very true. imagine all the free (crappy) artwork youd get with that!!!!! :eek:
 
Wow, all that and still can't play or author blu-ray...

Guess those sales make up for the 5,000 wedding videographers and 10,000 video editors who picked far cheaper PC's so they could deliver blu-ray to their clients.

Apple sure showed us...

:apple:
 
Guess those sales make up for the 5,000 wedding videographers and 10,000 video editors who picked far cheaper PC's so they could deliver blu-ray to their clients.

Apple sure showed us...

:apple:

Yep, wasting time fooling around with building supercomputers that might help develop cures for heart disease, cancer, Alzheimer’s or other ills of humankind when they could be offering blu-ray support for "videographers"...

Just what the heck were they thinking?!?!?! :confused:
 
So I'm amazed by the article...more amazed that I have been a member for awhile, and haven't found an article to comment on that would make me feel like I knew nothing of what was going on......Found it....Intriguing insight into the lead balloon over my head!!!
 
whats the average FLOPS that a computer these days produces?? (im talking an 8 core MP, not a lappy or anything lol).

Theoretical double-precision FLOPS using SSE = 4*(Clock speed)*(Number of cores)

So a 3.2 GHz 8-core Mac Pro has a theoretical FLOPS output of 102.4 GFLOPS.

Now sustained FLOPS is somewhat less.
 
ok thank you very much for that, best reply ive ever had :)

The "Top500" list, mentioned in the original MacRumors article, uses the Linpack Benchmarks to rank the "supercomputers".

Run that same test on a given MacPro and you have a meaningful comparison, otherwise it's more theory than real world... Not that there's anything wrong with that sort of discussion... Just saying... :D
 
GPUs don't offer the same wide range of calculations a CPU can perform, but they definitely have a LOT of power. Just look at folding@home and how much GPU calculations have helped.

As for right now, in terms of GPU FLOP power:

ATI 4870X2 : 2.4 TFlops
ATI 4870: 1.2 TFlops
ATI 4850: 1.0 TFlops
Nvidia GTX 280: 933 GFlops

Forgot what the GTX260 was but its ~700-800 or so GFlops IIRC.

However, the API for using Shaders to compute is very much in the infant stages. Nvidia has been banking on its own developed CUDA platform whereas ATI and Apple are supporting OpenCL (Open Compute Language).

Microsoft recently announced DirectX11 features which will include ComputeShade which will likely bring another standard for GPGPU purposes.

The FLOPS wars are only going to get better soon. I actually kind of echo the GPU makers in that CPUs are growing irrelevant beyond being necessary to setup the OS and so on. Actual calculations ported to the GPU such as folding, encoding, decoding, etc. have shown huge improvements.
 
GPUs don't offer the same wide range of calculations a CPU can perform, but they definitely have a LOT of power. Just look at folding@home and how much GPU calculations have helped.

As for right now, in terms of GPU FLOP power:

ATI 4870X2 : 2.4 TFlops
ATI 4870: 1.2 TFlops
ATI 4850: 1.0 TFlops
Nvidia GTX 280: 933 GFlops

Forgot what the GTX260 was but its ~700-800 or so GFlops IIRC.

However, the API for using Shaders to compute is very much in the infant stages. Nvidia has been banking on its own developed CUDA platform whereas ATI and Apple are supporting OpenCL (Open Compute Language).

Microsoft recently announced DirectX11 features which will include ComputeShade which will likely bring another standard for GPGPU purposes.

The FLOPS wars are only going to get better soon. I actually kind of echo the GPU makers in that CPUs are growing irrelevant beyond being necessary to setup the OS and so on. Actual calculations ported to the GPU such as folding, encoding, decoding, etc. have shown huge improvements.

got any idea how powerful the HD2600 in the imac or the X1600 (crappy i know) in my MBP is?? i wonder if the new snow leopard can utilise any power from these cards and increase performance (any ideas on % increases??)
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.