AMD Athlon 64 & Athlon 64 FX - It's Judgment Day
by Anand Lal Shimpi on September 23, 2003 1:25 PM EST- Posted in
- CPUs
32-bit vs. 64-bit Performance
Our entire benchmark suite to this point has been on 32-bit applications under a 32-bit OS, mostly because there are no good desktop 64-bit applications at this point in a popular 64-bit OS (not to mention the issues with 64-bit Windows XP we described earlier).
Under Linux however we don't have to wait for applications to be released in a 64-bit version, we can simply recompile them. Linux would thus provide us with an excellent venue to see the tangible performance increases from exposing the additional general purpose registers in 64-bit mode.
We ran all benchmarks on Red Hat Enterprise 2.9.5WS (Taroon), a beta release, booted in single user mode to avoid system services interfering with benchmark results. Neither Red Hat 9 nor 9.0.93 Beta (Severn) supply a 64-bit compiler or libraries, which is why we used Taroon.
The Taroon kernel initially had issues on startup requiring us to disable APIC and ACPI support to get it to install. Once actually running the OS was quite stable however DMA disk access was disabled for some reason.
We used the following compiler that came with Taroon:
gcc 3.2.3 20030502 (Red Hat Linux 3.2.3-16)
And the following kernel:
2.4.21-1.1931.2.393.ent
With this compiler and kernel we ran the following tests:
Whetstone
A simple C loop measuring floating point performance, configured to do double precision calculations.
Compiled with:
-O3 -msse2 -mfpmath=sse (and -m32 for 32bit, -m64 for 64bit)
The performance improvements due to 64-bit are in the 10 - 20% range we mentioned earlier.
Bytemark
An old integer CPU benchmark (FP results were discarded) - for more information on the tests visit this site.
Compiled with:
-O3 -msse2 -mfpmath=sse (and -m32 for 32bit, -m64 for 64bit)
Here we do see a small 2% drop in performance when moving to 64-bit in one test, however the rest of the tests show a 0 - 15% improvement across the board.
Lame 3.93
A MP3 encoder; encoded a 40minute .wav file (403MB).
Lame args: -b 192 -m s -h --quiet <file> - >/dev/null
(192kbps, simple stereo, high quality, output to nothing to avoid disk hits)
Compiled with:
-O3 -fomit-frame-pointer -fno-strength-reduce -malign-functions=4 -funroll-loops
-ffast-math -msse2 -mfpmath=sse (again, -m32 for 32bit, -m64 for 64bit)
The performance improvement here is astounding - in 64-bit mode the Athlon 64 FX managed to finish the encode 34% quicker than in 32-bit mode, if these results are any hint of what could be in store for Windows users, there's a lot of promise behind the Athlon 64...assuming we get software support in time.
We wanted to do a transcode benchmark but that didn't work out - one library found a bug in gcc and transcode refused to compile. It actually forced a compile error because a structure came out padded, meaning they didn't expect anyone to run it on a 64bit machine just yet.
Facebook
Twitter
RSS
122 Comments
View All Comments
Anonymous User - Tuesday, September 23, 2003 - link
Anyone know how the new AMD CPU compares to the Apple G5? I am not an Mac-Apple guy, but my in-laws are, and I'd like to be in the know in case we get into a friendly "discussion" about the Windows and Mac platforms.Anonymous User - Tuesday, September 23, 2003 - link
#58 Fanbois? lolAnonymous User - Tuesday, September 23, 2003 - link
This review appears to be in the same general lines as the rest of the Opteron/Pentium comparisons; I'm pleased that AMD has managed to shore up their shortcomings, but the price point is what's keeping me away from going directly from a pre-XP AMD Athlon to Athlon64. If I spend $400+ on a processor, it better be the king of the hill for the next year at least, or at least the mobo should be upgradeable to compensate for CPU obsolesence.And I'm surprised no one's figured out how to unlock Opteron multipliers yet, since that's basically the heart of the early-day AXP overclocking scene... Bridge blowing, soldering, "wire mods", etc. Shame, shame on you overclocking enthusiasts for not throwing everything into unlocking the hottest new processor (figuratively, not literally; Prescott and P4EE take that award at 103W and 150W, respectively). :P Talk about good wholesome fun, take an Opteron at 3.4GHz (using multipliers) and slap that Zalman Cu-7000 thing on it; a Pen-what?
#58: No, there are dumber fanboys than Intel fanboys, trust me. Just visit Something Awful. :/
Anonymous User - Tuesday, September 23, 2003 - link
intel fanbois rank among the top percentile of dumbest fanboi's on the internet.Anonymous User - Tuesday, September 23, 2003 - link
Is AMD actually planning on selling these versions of the 64? They and the hardware will be obsolete the day they are purchased. THe two biggest advantages the chip has can't even be used yet. The new mobos can't handle any more Ram than the current Pentium boards, I thought being able to use more ram was one of the selling points of the 64? Although that point seems to be moot anyway until a new 64 bit os is out.Anonymous User - Tuesday, September 23, 2003 - link
#36 You're right dude. Intel indeed said that prescott 3.2 GHz can't touch the performance of the 3.2 GHz P4EE. Logical actually, since prescott has no extra L3 cache, and a longer pipeline. The only benefits are: larger L1 cache, larger L2 cache and SSE-3 (only needed for sysmark-2004 LOL!, and other intel benchmarketing partners)Anonymous User - Tuesday, September 23, 2003 - link
Anagram for Intel Fanboy - INANE BOTFLYAnonymous User - Tuesday, September 23, 2003 - link
THG review: triple-guaranteed bullshit. Anandtech review: Infidel profane pagan loutish review. Ace's Hardware review: For great justice!11Anonymous User - Tuesday, September 23, 2003 - link
original pentium 66 was pants got beat by a 486original pentium 4 was just as bad
give it 6 months for the chip to mature. hopefully the athlon64 is a success cause if amd go bust we all pay double for cpus
Anonymous User - Tuesday, September 23, 2003 - link
There's some confusion on using the term 32bit and x86 here. I believe what was mean in response to what #32 said, is that A64 runs x86 natively the same way a XP does with no emulation, (as was outlined in previous Anandtech articles) just by disabling half of the 64-bit registers. So it had better run at least as well as the Athlon XP/P4 or there is something seriously wrong... not something to brag about.#50, For an Intel fanboy you sure don't know your history. Using 386 would be more appropriate as that was the change from 16-bit to 32-bit... and things have not fundamentally changed in the instruction set since then.