Supercomputer Vs. 2nd windtunnel

Here are our CFD links and discussions about aerodynamics, suspension, driver safety and tyres. Please stick to F1 on this forum.
CHT
CHT
-6
Joined: 14 Apr 2008, 05:24

Supercomputer Vs. 2nd windtunnel

Post

Some questions concerning the subject.

1) Will a 2nd windtunnel or a supercomputer like BMW's Albert2 bring more benefits to teams currently operating with only 1 wind tunnel?

2) And what are the pros and cons of supercomputer vs. 2nd wind tunnel?

nae
nae
0
Joined: 29 Mar 2006, 00:56

Re: Supercomputer Vs. 2nd windtunnel

Post

my understanding is that a wind tunnel is used to confirm the efforts of
designers and super computers, so once you have a decent calibration and an understanding what your super computers designs translate to in a wind tunnel
you could probably use the tunnel less

you need both at first for speed then , in theory at least, just the super computer
..?

pnagy
pnagy
0
Joined: 04 Mar 2006, 03:05
Location: Budapest,Hungary

Re: Supercomputer Vs. 2nd windtunnel

Post

A supercomputer will allow teams to test designs quicker than a wind tunnel because they do not need to make the parts to be tested. Its also cost less to make a computer than a wind tunnel

SoundMan
SoundMan
0
Joined: 08 Jan 2008, 19:41
Location: UK

Re: Supercomputer Vs. 2nd windtunnel

Post

ppl, keep bangin on about BMWs computer, Renult has the biggest computer by far!

ING Renault F1 Team: Appro Xtreme-X2, 1024 sockets, 4096 cores (AMD QC Opteron)

Scuderia Ferrari Marlboro: Acer/IBM/Racksaver, 1000+ processor sockets (upgrading to QC Opteron)

BMW Sauber F1 Team: Dell, Albert2, 512 sockets, 2048 cores (Intel Xeon)
Vodafone McLaren Mercedes: Silicon Graphics Altix, 512 Sockets, 1024 cores (Intel Itanium 2)

Red Bull Renault: IBM, 512 sockets, 1024 core (upgrading to AMD QC Opteron)

Panasonic Toyota F1 Team
: Fujitsu, 320 Sockets, 640 cores (Intel Itanium 2)

AT&T WilliamsF1 Toyota
: Lenovo Unnamed, 332 Sockets, 664 cores (Intel Xeon 5100)

Honda F1 Racing: SGI Altix ICE, unknown number of socket/cores, water-cooled Quad-Core Intel Xeon

Scuderia Torro Rosso Ferrari: N/A (uses Red Bull infrastructure)

Super Aguri Honda
: N/A (uses Honda F1 infrastructure)

Force India Ferrari
: Rental system (unknown specifications)

see article: http://www.tomshardware.co.uk/AMD-Formu ... 27951.html

User avatar
GTO
0
Joined: 09 Jun 2005, 01:16
Location: Oil Country

Re: Supercomputer Vs. 2nd windtunnel

Post

I'm not a comp. tekkie...pls. clarify, what defines the power of these supercomputers? Is it sockets, cores, or calculations per second capability?

User avatar
sac
0
Joined: 25 Jan 2008, 02:51

Re: Supercomputer Vs. 2nd windtunnel

Post

They are typically ranked by max LINPACK performance (in GFlops). The Tom's list is about right in terms of order but a bit suspect in terms of equipment detail.

Conceptual
Conceptual
0
Joined: 15 Nov 2007, 03:33

Re: Supercomputer Vs. 2nd windtunnel

Post

sac wrote:They are typically ranked by max LINPACK performance (in GFlops). The Tom's list is about right in terms of order but a bit suspect in terms of equipment detail.
I'm telling you, the emails that I received from my questions about using the new video cards to do double-precision CFD work say ALOT about the possible potential of moving all CFD from CPU to the GPU. And since you can build a rack that has a single quad-core CPU 64-bit, and 4 dual processor, dual core (16 cores) 128-bit video cards for about $4k USD each, and attain similar results for about $30M less, I think you may see some movement to a CFD system like that.

My REAL question about all of this is: If the F1 regs for 2009+ require a 50% decrease in downforce, are they really necessary at all? Haven't the teams already destroyed that barrier? Will the CFD upgrades and extra wind tunnels really benefit when those rules go into effect?

...................just rambling on, and on tonight! Gotta lay off the caffiene!

Chris

User avatar
WhiteBlue
92
Joined: 14 Apr 2008, 20:58
Location: WhiteBlue Country

Re: Supercomputer Vs. 2nd windtunnel

Post

actually the Renault computer isn't available before July. they are going to miss the boat for much of the 2009 work. but perhaps they have an agreement to use time on a demo machine somewhere.
Formula One's fundamental ethos is about success coming to those with the most ingenious engineering and best .............................. organization, not to those with the biggest budget. (Dave Richards)

AeroGT3
AeroGT3
0
Joined: 29 Mar 2006, 23:22

Re: Supercomputer Vs. 2nd windtunnel

Post

SoundMan wrote:ppl, keep bangin on about BMWs computer, Renult has the biggest computer by far!

ING Renault F1 Team: Appro Xtreme-X2, 1024 sockets, 4096 cores (AMD QC Opteron)

Scuderia Ferrari Marlboro: Acer/IBM/Racksaver, 1000+ processor sockets (upgrading to QC Opteron)

BMW Sauber F1 Team: Dell, Albert2, 512 sockets, 2048 cores (Intel Xeon)
Vodafone McLaren Mercedes: Silicon Graphics Altix, 512 Sockets, 1024 cores (Intel Itanium 2)

Red Bull Renault: IBM, 512 sockets, 1024 core (upgrading to AMD QC Opteron)

Panasonic Toyota F1 Team
: Fujitsu, 320 Sockets, 640 cores (Intel Itanium 2)

AT&T WilliamsF1 Toyota
: Lenovo Unnamed, 332 Sockets, 664 cores (Intel Xeon 5100)

Honda F1 Racing: SGI Altix ICE, unknown number of socket/cores, water-cooled Quad-Core Intel Xeon

Scuderia Torro Rosso Ferrari: N/A (uses Red Bull infrastructure)

Super Aguri Honda
: N/A (uses Honda F1 infrastructure)

Force India Ferrari
: Rental system (unknown specifications)

see article: http://www.tomshardware.co.uk/AMD-Formu ... 27951.html
Not anymore! :-)

alexbarwell
alexbarwell
0
Joined: 20 Mar 2008, 14:19
Location: London

Re: Supercomputer Vs. 2nd windtunnel

Post

Remember guys, size isn't important, it's what you do with it :oops:
I am guessing a bit, but doesn't CFD have a lot to do with 3D modelling? Same for CGI (TV and film background for me)that use a shedload of processor cores to crunch it all out. It is probably easier in CFD to simulate something close to real-world turbulence that could be difficult to implement physically. Can the CFD modelling factor in dynamic positioning of the car as well ie. yaw, pitch, roll changing aspects of airflow?
I am an engineer, not a conceptualist :)