AMD FX-4100 Bulldozer

As mentioned over the weekend, a Phoronix reader that was excited about AMD's Bulldozer products had went out and immediately purchased an FX-4100 processor. This user graciously let me SSH into the system as soon as Ubuntu Linux was installed so that benchmarks from the AMD FX-4100 could be conducted. Here is a look at the AMD FX-4100 Bulldozer on Linux compared to Llano Fusion hardware and Intel Sandy Bridge processors.

The FX-4100 is currently AMD's lowest-end FX Bulldozer processor with a 3.6GHz base frequency and 3.8GHz in its higher operating state, 4MB of L2 cache, 8MB of L3 cache, and 95 Watt TDP. Like the other Bulldozers, it is an AM3+ socket, 32nm SOI processor, DDR3-1866MHz memory support, and comes out of the Global Foundries facility in Dresden, Germany. The AMD FX-4100 is a quad-core offering unlike the six-core FX-6100 or the octal-core FX-8100 series.

Fortunately, AMD has agreed to send out a Bulldozer kit so in the coming days there will be more FX CPU benchmarks. Ideally, it will be an octal-core FX part. There are already some FX-8150 Linux benchmarks submitted by a Phoronix reader using the Phoronix Test Suite and OpenBenchmarking.org. In the testing today the FX-4100 is just being tested in a stock configuration. There are some GCC 4.5 vs. GCC 4.6 compiler benchmarks I also did from this system, which will be out in the next few days on Phoronix. There are not any tests yet of the Linux kernel patch by AMD that has reported to optimize the Bulldozer performance under Linux, but once I have physical access to a Bulldozer setup, I will be conducting such tests. Overclocking of the FX-4100 also could not be handled remotely.

With this FX-4100 testing also being done remotely for now, I wasn't able to mirror the rest of the system hardware in an identical manner to this user's system, but did the best possible job and also limited the Linux benchmarks strictly to those that will stress the CPU without touching the disk, graphics, or extensive memory use.

For comparison what was also benchmarked was an AMD A8-3850, Intel Core i3 2120, Intel Core i5 2400S, and Intel Core i5 2500K. These other systems were also running Ubuntu 11.04 from the Linux 2.6.38 64-bit kernel. Testing across all of the systems was facilitated by the Phoronix Test Suite and OpenBenchmarking.org.

Below is a sample of the /proc/cpuinfo output for the FX-4100 CPU. The full cpuinfo output can be found on the OpenBenchmarking.org log page.

processor : 3
vendor_id : AuthenticAMD
cpu family : 21
model : 1
model name : AMD FX(tm)-4100 Quad-Core Processor
stepping : 2
cpu MHz : 1400.000
cache size : 2048 KB
physical id : 0
siblings : 4
core id : 3
cpu cores : 2
apicid : 19
initial apicid : 3
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc extd_apicid aperfmperf pni pclmulqdq monitor ssse3 cx16 sse4_1 sse4_2 popcnt aes xsave avx lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt lwp fma4 nodeid_msr topoext arat cpb npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold
bogomips : 7224.27
TLB size : 1536 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 48 bits physical, 48 bits virtual
power management: ts ttp tm 100mhzsteps hwpstate [9]