xena:/root% cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 6
model : 6
model name : AMD Athlon(tm) Processor
stepping : 1
cpu MHz : 995.558
cache size : 256 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep
mtrr pge mca cmov pat pse36 mmx fxsr syscall mmxext 3dnowext 3dnow
bogomips : 1985.74
processor : 1
vendor_id : AuthenticAMD
cpu family : 6
model : 6
model name : AMD Athlon(tm) Processor
stepping : 1
cpu MHz : 995.558
cache size : 256 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep
mtrr pge mca cmov pat pse36 mmx fxsr syscall mmxext 3dnowext 3dnow
bogomips : 1985.74
xena:/root% df
Filesystem 1k-blocks Used Available Use% Mounted on
/dev/hda2 3882016 919408 2765412 25% /
/dev/hda1 69973 9829 56531 15% /boot
/dev/md0 107308824 37595628 64336896 37% /mov
The answer is yes. There is such a thing as a dual Athlon. It has two ethernet ports which we used to develop FIREHOSE. It uses AGP PRO which is backwards compatible with AGP 2x. It uses 266Mhz DDR.
As for performance our experiences are biased because this system is almost exclusively used for video software development not games like most. It needs a reliable operating system like Linux and very fast media storage drives.
The inverse telecine, a grueling memory excercise which takes 3 hours on a dual PIII 933 and 2 hours on a dual Alpha, takes about 2 hours on the dual Athlon.
Our 100 Gig SCSI raid, consisting of 6 15,000 rpm drives on the motherboard's two SCSI 160 channels gives a full 110MB/sec read and write with RAID 0. With RAID chunks set to 1MB the write accesses go to 160MB/sec and read accesses go to 90MB/sec sustained. This system would make a good motion capture tool. Previous Intel attempts at onboard disk I/O would give 50MB/sec.
As for compiling, we compiled the Cinelerra main executable, a brutal C++ compiler test resulting in a video editing program, and observed the following:
Dual PIII 933:
time make -j 3
real 3m10.386s
user 5m52.000s
sys 0m15.260s
Dual Athlon 1Ghz:
time make -j 3
real 2m44.692s
user 4m54.150s
sys 0m12.950s
What about parallel MPEG2 compression? AMD benefits here from #1 3DNOW assembly language, #2 faster hard drives, and #3 really fast memory. The dual PIII 933 only has IDE drives but this is a tiny fraction of the delay in MPEG2 encoding.
Dual Athlon 1Ghz:
time mpeg2enc -q 15 -n 45 -d -f 15 anorexic1.mov
real 4m24.988s
user 1m14.570s
sys 0m3.030s
Dual PIII 933:
time mpeg2enc -q 15 -n 45 -d -f 15 anorexic1.mov
real 5m46.673s
user 1m52.860s
sys 0m3.620s
For MPEG2 decoding, the dual Coppermine can decompress a 1280x720 digital TV signal at 25fps. The dual Athlon can decompress it at 35fps. The dual Athlon is using a TNT2 and the dual Coppermine is using a Geforce II.
This system is as reliable as any other Tyan server board even though the 760MP chip gets hot enough to cook popcorn. The DDR stick is quite cool compared to our RAMBUS grill irons of yesteryear. Although it requires a 460W power supply with an 8 pin subconnector, the system doesn't use nearly that much power running Linux. The real power users are the 15,000 rpm drives.
While all this high performance feels good, whether the dual Athlon becomes the next big thing or a distant reminder of our long lost obsession with desktops is pretty clear. When Intel introduced the dual pentium Pro in the early 90's it was on every magazine front page. Today new desktop innovations are hardly a blip on the radar. The last 3 years have shown increasing distaste for big, powerful, hard to use desktops and increasing devotion to lower powered handhelds.
x ora amd roxxa d+
Ciao
Rispondi quotando
