MaximumPC 2007 112

(Dariusz) #1

B


y now you’ve probably heard about Intel’s new
QuickPath Interconnect, briefly known as the Common
System Interface (CSI). QuickPath is Intel’s answer to
HyperTransport, the high-speed point-to-point serial inter-
face that AMD adopted years ago. HyperTransport con-
nects the processor core to the on-chip memory controller.
On AMD’s multicore chips, HyperTransport also connects
the processor cores together.
Thanks partly to HyperTransport, AMD’s processors
have enjoyed advantages in memory performance,
system integration, and power consumption. Now Intel
is overcoming those advantages. New Intel micro-
architectures like Nehalem are faster and more effi-
cient, and QuickPath will match or exceed the per-
formance of HyperTransport. QuickPath will appear in
future Intel CPUs based on Nehalem.
However, QuickPath serves another purpose: It gives
Intel an additional way to differentiate its x86 micropro-
cessors from each other. These differences will be subtle
but could measurably affect performance.
The oldest way to differentiate microprocessors
within a product line is to offer them at various clock
frequencies. All else being equal, higher speeds are
better. Another differentiation that became common in
the 1990s is to offer different-size caches. All else being
equal, bigger caches are better.
More recently, yet another differentiation is to offer
multiple processor cores. All else being equal—and
assuming that multicore software is available—the more
cores, the better. Ideally, those cores are integrated on
a single die. Or multiple dies can be united in a single
package.
Nehalem-based CPUs will differentiate in all those
ways, but also in another: the configuration of their
QuickPath connections. Consider the possibilities for
a quad-core CPU. A lower-cost, lower-performance
version could link the four cores together in a simple
square. Each core could communicate with its two
neighbors in one hop, but cores at opposite corners
would require two hops.
Now picture a square configuration crossed with an
“X” in the middle. These extra QuickPath connections
would give each core a one-hop connection to every
other core. This design is more expensive but delivers
greater performance.
Of course, AMD can do the same with
HyperTransport. The point is that future multicore chips
will differentiate themselves by their interconnects, as
well as by the usual factors. Two multicore CPUs with
identical clock speeds, caches, cores, and integrated
features may perform quite differently, depending on the
arrangement of their internal pathways.

Tom Halfhill was formerly a senior editor for Byte magazine
and is now an analyst for Microprocessor Report.

Intel Thinks


Different


FAST FORWARD


TOM
HALFHILL

DECEMBER 2007 MAXIMUMPC 09


Google’s purchase of the popular video-sharing
site is starting to make sense—AdSense, that is.
Google’s highly profi table ad-serving program will
now offer video units to its website affi liates. So
an AdSense ad can contain site-related YouTube
footage—framed by targeted ad text, of course.
Revenue generated by clicks will be split among
Google, the videomaker, and the website publisher.

Google Cashes


YouTube Check


Stiff Fine


Levied for


Copyright


Infringement


The RIAA has fi nally caught a
break. The fi rst of its 20,
peer-to-peer copyright
infringement lawsuits to go to
trial has resulted in a $222,
judgment against Jammie
Thomas, a Minnesota woman
accused of distributing 24
copyrighted songs on the P2P
service Kazaa.
The judgment is a dra-
matic turnaround for the RIAA.
In recent months, its war on
fi le-sharing has been dogged
by universities and ISPs that
refused to release individuals’
account information and judg-
es that quashed subpoenas
and dismissed the recording
industry’s lawsuits.
The defense’s argument—
essentially, “you can’t prove it
was her”—was no match for
the recording industry’s strong
case. Its lawyers noted that
the IP address associated with
the infringement was appar-
ently used only by Thomas’s
password-protected computer
and cable modem, and that the
infringing Kazaa account name
was identical to the name
Thomas used for her other
accounts.
Thomas plans to appeal
the decision, citing a specifi c
jury instruction which stated, in
part, that making copyrighted
fi les available on peer-to-peer
networks violates copyright
“regardless of whether actual
distribution has been shown.”
Her defense argues that the
Copyright Act doesn’t say that
and judges have handed down
contradictory rulings on that
topic in the past. Giving such an
instruction to the jury obviated
the need for the RIAA to prove
that anyone but its investigators
downloaded the fi les.
The win certainly adds bite
to the RIAA’s threats against
fi le-sharers, but it remains to
be seen how many of its other
lawsuits will be as clear cut.


P2P Paranoia Pays Off
Blocklists are essential if you don’t want your peer-
to-peer activity spied on by the Man. That’s what
researchers at the University of California, Riverside
discovered after combing through 100GB of TCP head-
er information from P2P networks. A blocklist contains
the IP ranges associated with snoops for the RIAA,
MPAA, and others interested in tracking fi le-swapping.
The researchers found that users who failed to employ
this useful device eventually connected to one of the
suspect IP addresses, without fail.

Vonage’s Fate in the Balance
If VoIP provider Vonage can weather a spate of
patent lawsuits, it just might survive. The com-
pany recently settled a federal court case with
Sprint, agreeing to pay the telco $80 million for
past and future use of its IP. Now it must resume a
patent-infringement court battle with Verizon. The
Federal Court of Appeals has already remanded
a jury’s $58 million-plus judgment against Vonage
in that case; now the com-
pany is hoping a retrial in U.S.
District Court will end the liti-
gation once and for all.
Free download pdf