AMD CHEATS... THEY MANIPULATED THE ATHLON TEST RESULTS...
(excuse me if already posted. I've been gone all week)
Update: Benchmarking firm
claims AMD manipulated Athlon
test results
By Mark Hachman
Electronic Buyers' News
(08/19/99, 09:33:13 AM EDT)
A Canadian benchmarking operation claims Advanced Micro
Devices Inc. modified tests to give its Athlon microprocessor
an unfair advantage over Intel Corp.'s Pentium III.
FutureMark Software Corp. Ltd. said AMD is violating the
trademark of the Toronto company's 3DMark 99 MAX test
suite and terms of a related licensing agreement. While
FutureMark has not taken any legal action against AMD, the
company said AMD's actions were "inappropriate."
According to FutureMark, AMD modified the 3D Mark 99
MAX test by optimizing the DLL test code for its recently
released Athlon. AMD's own test results demonstrated that,
in one of the two 3Dmark tests, the 600-MHz Athlon
outperformed the 600-MHz Pentium III by more than 30%. Of
the 13 test results accompanying AMD's Athlon
performance assertions, the 3DMark figures were among the
highest in favor of the chip.
Both parties have agreed that AMD has the right to optimize
the FutureMark code to demonstrate the maximum possible
performance on the Athlon microprocessor. AMD erred,
according to FutureMark, by comparing the optimized Athlon
results to the generic test suite used with the Pentium III.
FutureMark said that official results should have been based
on the same 3DMark test being applied across the board to
all processors.
"No other CPU manufacturers have been given a chance to
make optimizations," FutureMark said in a statement. "As
such, it is inappropriate to compare the AMD special DLL
version results with the official 3DMark 99 MAX version
results."
FutureMark's claim apparently has uncovered the industry's
latest example of "specmanship," a complicated practice
whereby a company manipulates a convoluted range of test
applications and benchmarks. Benchmarks are a series of
tests performed to reduce a chip's performance in a variety
of applications to a single, simplified number for marketing
purposes. The process is further complicated because the
choice of tests itself is somewhat subjective.
A spokesman for AMD, Sunnyvale, Calif., noted that
mention of the customized 3DMark test suite was fully
disclosed on page 39 of a 42-page software performance
guide that accompanied the test results.
"The reference should have been included in some other
places," the spokesman said. "But in the excitement of the
[Athlon] launch, that reference was overlooked." He added
that the two companies were in talks to resolve the matter.
In the guide, AMD makes no mention of optimizing its code
for the other test suites. Calls to those test-software
manufacturers have not been not returned. Several other
companies that test microprocessors have independently
concluded that the Athlon generally outperforms the Pentium
III in a variety of applications.
But that's not unusual. In 1997, for example, Jon Peddie
Associates (JPA), a Tiburon, Calif., research company,
found that NEC Corp.'s PCX2 chip performed better than
3Dfx Interactive Inc.'s Voodoo Graphics chipset. After the
results were made public, JPA, in a statement published on
its home page, said it was informed of a driver optimization
installed by NEC that could be disabled by renaming the
binary executable test. When that driver was turned off, the
chip didn't perform as well as the first time. NEC and its
partner, VideoLogic Ltd., however, stood behind the original
set of benchmarks.
ebnews.com |
