I like to think of it as the new Alpha, but with a fancy bytecode compressor (called x86-64) which doubles as a compatibility layer.:)
True. An Athlon-64 with a 1 MB cache does well
alongside an Itanium with a far larger cache.
Costs and price will always be less for
the AMD64 than the Itanium. There is some
goodness in getting small code size.
Because of its NUMA ability, the Opteron is
almost as lovable as the Alpha.
With time I think it will win people's
hearts the way Alpha did.
You can get a very powerful brand new portable
for $600 from walmart.com. This is a real price
without any rebates troubles. It is full featured
with DVD, wireless, etc. I have one and recommend
it to anyone with limited funds.
A good review would have pitched the 3.6Ghz nacoma vs an Opteron 150, would have tested both in 32 and 64 bit and tried to use some application benchmarks.
There are benchmarks from
anadtech.com and
xbitlabs.com
that show AMD64 chips have higher performance on 64-bit code.
Since there are more registers in 64-bit mode, it seems very reasonable
for it to run 64-bit code faster.
However, both
theinquirer.net
and
infoworld.com
claim that the 64-bit performance of Xeon-Nocona is
no higher than its 32-bit performance.
At first this seems unreasonable, since it will also have
the additional registers that helped AMD.
However, some of the
64-bit instructions can be longer,
so relying on a big cache may not work as well and
high memory bandwidth may be more important. So it could well be
that AMD's chips are better suited for 64-bit code.
Though Xeon-Nocona has been
available for more than a month it seems there
there are no substantial reports on 64-bit performance of Nocona.
Is there anyone here who can report anything about the
64-bit performance of Nocona?
Can you run it again with JDK1.5.0 32bit on
Xeon to see how different it is with same
JDK version? Also, I assume you were running 32-bit Linux
on the Xeon of the same FC2 release used on AMD?
I tried to post on your blog but could not
see how to get a username/password.
You should have a link to "make new user" on
your login page.
There is another interesting
unofficial Intel roadmap.
Since
Tejas and Jayhawk have been canceled
there is just not much new Intel x86 in the server, workstation, and desktop
in the next 11 months. The other interesting thing is that Intel's first
duel CPUs will first be in Itanium and notebooks. Since heat is
something of a problem in current single core server, workstation,
and desktop chips, it makes sense that these will not be the first
duel-core chips. AMD plans to have high end
duel core chips next year.
Also, people are right that in the SPEC results
the 3.4 Ghz chip has a 2 MB level-3 cache and
the 3.6 Ghz does not.
All of this blather is about "Windows XP 64-Bit Edition for 64-Bit Extended Systems" and is, of course, ignoring the other six versions including a released and supported Windows XP 64-bit Edition for the Itanium family.
My guess is that AMD made more AMD64 chips this
month than Intel has made Itaniums chips since the
beginning of time. So shortening that long
name to "Win64" does not really
seem to leave out anything important.:-)
I've got this foreboding sense of dread that we'll spend thousands upon thousands of dollars on hardware, and thousands of man hours getting everything up and running, only to discover, a year from now, that Win64 won't be stable on that platform.
Try the Win64-Beta now (it is free for a year)
on one machine.
If it installs on your hardware then the final
release should also. Also, if the Beta can
run on your hardware and run the applications
that you need, you might be fine to
just stick with it.
My, but does anyone else think the submitter live in a rather sheltered world?
True, the submitter lives on a tropical island
in the
Caribbean.:-)
I should have said "mainstream choices
for 64-bit OS on AMD64". And I did not mean to slight BSD, in my mind "Linux" maps
nearly to "free Unix type OS".
2. A human being has trouble surviving a re-entry inside a spaceship covered with heat-resistant tiles, do you really think a bacterium sitting on a rock that is heated up to a few thousand degrees has a chance in hell of surviving the trip?
Many meteors ablate like an Apollo heatshield
as they enter the atmosphere. The heat is
used up turning the surface into a gas and little
heat is conducted inside the meteor. Meteors start out
very cold, so meteorites are often very cold to the
touch when found right after impact.
I remember reading about the Shuttle experiment with the tether, the idea, if I recall, was to see if usable energy could be generated in this manner. Everybody was surprised at how quickly a charge built up and burned out the cable. This doesn't sound good for space elevators!
Any takers on this item?
This voltage/current comes from moving a wire
relative to the Earth's magnetic field.
In a space elevator the cable is mostly
moving with
the Earth's magnetic field. So it won't
be much like that test case, which was
moving like 17,000 mph relative to the
Earth's magnetic field.
The main thing this is good for is for
propulsion. A rotating tether can
pickup and toss payloads but it looses
some momentum unless there is other traffic
going the other way. But with an
electrodynamic tether pushing on the Earth's
magnetic field you can get momentum without
using rocket fuel. This is way cool.
The notes mention a 3.5 GPa nanotube/PBO rope.
It should be pointed out that PBO without
nanotubes is 5.8 GPa. So the strength is
reduced by having the nanotubes.
The strongest real carbon-nanotube rope is
1 GPa. Edwards elevator design needs 100 GPa
ropes.
In Edwards book, "Space Elevators",
page 26 he predicted that we would have
100 GPa ropes by 2005 or earlier.
We just are not close to this and not
moving fast. Last year the record was
also 1 GPa.
Now Edwards is
predicting another 2 years. He will
be wrong again. We had "graphite whiskers"
48 years ago that had 20 GPa and we
can not make strong ropes of these yet.
These are easier to bind to than
nanotubes. So 2 years is just much too
optimistic for 100 GPa.
The combination of a reusable-suborbital rocket
and a rotating tether could be built today.
The tether can get energy back from tourists
returning to Earth. So if your main traffic is
tourists going up and down, the tether energy
is easy. Another fun trick is that if you
had a series of tether in LEO, GEO, and Lunar
orbit you could send stuff to the moon and
send moon rocks (or other stuff) back to the Earth without
needed to add energy. You just keep the
total mass going each way balanced.
Because of this, orbital or lunar tourism
will not take much more energy than suborbital
rocket rides. So we should see it within
the next 20 years.
I have a site,
spacetethers.com
that
has info and a Java applet tether simulator.
There is also lots of info at
tethers.com
Re:It amazes me how expensive these things are
on
70 Megapixel Webcam
·
· Score: 1
Scanning the moon by having the Earth turn is a really cool idea. I found the guy at:
> "Could inexpensive cruises to the moon happen within our lifetimes?"
If you use 3 rotating space tethers you
could setup something that makes it very cheap
to go to the moon and back.
Imagine a tether around the moon. As it drops
off someone it picks up a bucket of regolith
of the same weight.
So it does not have any change in energy.
It can catch something tossed from GEO and
use that energy to toss the bucket of
regolith to GEO. The GEO tether uses the
energy of catching this regolith to pickup
something tossed from a LEO tether, where
it sends the regolith. The LEO tether uses
the energy of this regolith to pickup
a payload from a reusable suborbital rocket.
So in a nutshell you use the energy of some
regolith coming from the moon to Earth to
balance the energy of some payload going
from Earth to the moon. If you have tourists
coming back you can use them instead of
regolith.
The SpaceX Falcon-1 has a reusable 1st stage
that could work for this. Spectra-2000 is
a very strong material used for fishing line
and would make a good tether. So materials
are not a problem.
It could be done in the next 10 years really.
So if by "inexpensive" you mean "under $100,000",
and by "in our lifetimes" you mean "20 to 50 years" then I think the answer to your question
is yes.
The space elevator idea needs materials that we
may not have for 50 years or more.
Rotating space tethers can be built today.
They do not need anything stronger than Spectra-2000, which is used for fishing line. Even though we could really build these,
space tethers don't get nearly the press that space elevators get. I think this is because the space elevator is simpler to understand.
A rotating space tether is long (like maybe 500 km long) with a weight at one end. It
is rotating around its center of mass, which
is close to the weighted end. The whole thing
is in orbit around the Earth. It is spinning fast, like maybe 2.5 km/sec at the tip.
The tip is going backwards relative to the
orbital speed when it is down and forwards
when it is up. So when it is down something
can grab on even though it is not going at
orbital speed.
Rotating space tethers can catch a payload
from a reusable suborbital rocket and toss
it into orbit. The reusable 1st stage of the
SpaceX Falcon-1 could lift a payload high
enough and fast enough for a practical tether
to catch it. This could make for very cheap
access to space.
Rotating tethers can recycle energy. If one
tourist is coming down and another is going
up there is no need to use an ion-drive or
other thruster. This is a fantastic win.
But even if you have to use a thruster it
can be a high ISP thruster or you can even
push on the Earth's magnetic field with an
EDT.
Rotating space tethers can toss a payload
every 100 minutes. A space elevator could
easily take days to climb.
I have a Java applet that you can use to
simulate space tethers at my site:
http://spacetethers.com/
With this you can see how they work. In my
unbiased opinion, this is the best way to
get into space with current materials.
Slashdot Mirror
Because of its NUMA ability, the Opteron is almost as lovable as the Alpha. With time I think it will win people's hearts the way Alpha did.
You can get a very powerful brand new portable for $600 from walmart.com. This is a real price without any rebates troubles. It is full featured with DVD, wireless, etc. I have one and recommend it to anyone with limited funds.
But rumor has it that Intel's 64-bit performance is no faster than its 32-bit performance, and slower than AMD's 64-bit performance.
Though Xeon-Nocona has been available for more than a month it seems there there are no substantial reports on 64-bit performance of Nocona. Is there anyone here who can report anything about the 64-bit performance of Nocona?
I tried to post on your blog but could not see how to get a username/password. You should have a link to "make new user" on your login page.
Also, people are right that in the SPEC results the 3.4 Ghz chip has a 2 MB level-3 cache and the 3.6 Ghz does not.
It seems like you might as well just run the free Beta.
I should have said "mainstream choices for 64-bit OS on AMD64". And I did not mean to slight BSD, in my mind "Linux" maps nearly to "free Unix type OS".
Many meteors ablate like an Apollo heatshield as they enter the atmosphere. The heat is used up turning the surface into a gas and little heat is conducted inside the meteor. Meteors start out very cold, so meteorites are often very cold to the touch when found right after impact.
Yes, I really think they could survive the trip.
The main thing this is good for is for propulsion. A rotating tether can pickup and toss payloads but it looses some momentum unless there is other traffic going the other way. But with an electrodynamic tether pushing on the Earth's magnetic field you can get momentum without using rocket fuel. This is way cool.
The strongest real carbon-nanotube rope is 1 GPa. Edwards elevator design needs 100 GPa ropes. In Edwards book, "Space Elevators", page 26 he predicted that we would have 100 GPa ropes by 2005 or earlier. We just are not close to this and not moving fast. Last year the record was also 1 GPa.
Now Edwards is predicting another 2 years. He will be wrong again. We had "graphite whiskers" 48 years ago that had 20 GPa and we can not make strong ropes of these yet. These are easier to bind to than nanotubes. So 2 years is just much too optimistic for 100 GPa.
The tether can get energy back from tourists returning to Earth. So if your main traffic is tourists going up and down, the tether energy is easy. Another fun trick is that if you had a series of tether in LEO, GEO, and Lunar orbit you could send stuff to the moon and send moon rocks (or other stuff) back to the Earth without needed to add energy. You just keep the total mass going each way balanced.
Because of this, orbital or lunar tourism will not take much more energy than suborbital rocket rides. So we should see it within the next 20 years.
I have a site, spacetethers.com that has info and a Java applet tether simulator. There is also lots of info at tethers.com
Scanning the moon by having the Earth turn
is a really cool idea. I found the guy at:
http://www.k3pgp.org/astropix.htm
The image is very good.
If you use 3 rotating space tethers you could setup something that makes it very cheap to go to the moon and back.
Imagine a tether around the moon. As it drops off someone it picks up a bucket of regolith of the same weight. So it does not have any change in energy. It can catch something tossed from GEO and use that energy to toss the bucket of regolith to GEO. The GEO tether uses the energy of catching this regolith to pickup something tossed from a LEO tether, where it sends the regolith. The LEO tether uses the energy of this regolith to pickup a payload from a reusable suborbital rocket.
So in a nutshell you use the energy of some regolith coming from the moon to Earth to balance the energy of some payload going from Earth to the moon. If you have tourists coming back you can use them instead of regolith.
The SpaceX Falcon-1 has a reusable 1st stage that could work for this. Spectra-2000 is a very strong material used for fishing line and would make a good tether. So materials are not a problem.
It could be done in the next 10 years really. So if by "inexpensive" you mean "under $100,000", and by "in our lifetimes" you mean "20 to 50 years" then I think the answer to your question is yes.
See http://spacetethers.com/ for more info.
-- Vince
Rotating space tethers can be built today. They do not need anything stronger than Spectra-2000, which is used for fishing line. Even though we could really build these, space tethers don't get nearly the press that space elevators get. I think this is because the space elevator is simpler to understand.
A rotating space tether is long (like maybe 500 km long) with a weight at one end. It is rotating around its center of mass, which is close to the weighted end. The whole thing is in orbit around the Earth. It is spinning fast, like maybe 2.5 km/sec at the tip. The tip is going backwards relative to the orbital speed when it is down and forwards when it is up. So when it is down something can grab on even though it is not going at orbital speed.
Rotating space tethers can catch a payload from a reusable suborbital rocket and toss it into orbit. The reusable 1st stage of the SpaceX Falcon-1 could lift a payload high enough and fast enough for a practical tether to catch it. This could make for very cheap access to space.
Rotating tethers can recycle energy. If one tourist is coming down and another is going up there is no need to use an ion-drive or other thruster. This is a fantastic win. But even if you have to use a thruster it can be a high ISP thruster or you can even push on the Earth's magnetic field with an EDT.
Rotating space tethers can toss a payload every 100 minutes. A space elevator could easily take days to climb.
I have a Java applet that you can use to simulate space tethers at my site:
http://spacetethers.com/
With this you can see how they work. In my unbiased opinion, this is the best way to get into space with current materials.
-- Vince