Comparing Sockets: 939 vs. 940 vs. 754

To get a clearer idea of exactly what Socket 939 brings to the performance table, we brought some real test results to some of the speculation that has been brewing on the web as to whether 939 is really faster than 754. We suspect our objective tests of Socket 939/940/754 will fly in the face of some of the absurd speculation and sloppy test results that are being posted about the new Socket, but the truth is rarely as exciting as controversies created to stand out from the crowd.

The comparison is simple - there are the 3 sockets that all have processors that can run at 2.2Ghz. To keep the comparison as fair as possible we tested the 3 sockets -754, 940, and 939 with 3 processors with 1MB of on-chip cache all running the same 2.2GHz speed . This gave a head-to-head comparison of the single-channel memory controller of Socket 754 to the Dual-Channel Registered Memory of 940 to the latest Dual-Channel Unbuffered Socket 939.

Sockets were compared using the standard motherboard test suite to give a broad comparison of performance. General Performance was compared using Veritest Multimedia Content Creation 2004 and Business Winstone 2004. results were also compared in PCMark 2004.

General Performance - 2.2GHz & 1MB Cache

General Performance - 2.2GHz & 1MB Cache

Winstones are usually very static at a given CPU speed on a processor. Even wide variations in memory bandwidth and graphics performance rarely have much impact on the Winstone scores. The increases in Winstone scores were only 2.6 to 3% from 754 to 939, but the pattern was very consistent with 939 the fastest, 940 close to 939 and 754 slower than either socket for dual-channel memory. PCMark 2004 was an even more modest spread of 1.4% from slowest to fastest 2.2GHz.

High End Workstation Performance - SPECViewperf 7.1.1

Workstation performance is more sensitive to memory bandwidth, and we do see a wider range in variation among the 2.2GHz processors in SPECviewperf. 754 to 939 ranged from +6.5% in UGS to an 11.4% increase in DX. Considering the CPU's are all the same core at 2.2GHz this is a wide variation just from different memory controllers. The pattern was the generally the same fastest to slowest of 939-940-754, except 940 did outperform 939 in a couple of the SPECviewperf benchmarks.

Development Workstation Performance Comparing Sockets: Gaming Performance
POST A COMMENT

38 Comments

View All Comments

  • Viditor - Saturday, June 5, 2004 - link

    #37 - "Doesn't seem to me to make any sense spending 5 Grand on a 64 bit system until the OS is available?"

    5 GRAND?!?!?!
    What are you PUTTING in there?
    Even the most expensive (FX53) chip is ~$840, and the ASUS mobo is ~$190...
    Reply
  • tmhartsr - Friday, June 4, 2004 - link

    Hey guys - where is the 64 bit OS? Doesn't seem to me to make any sense spending 5 Grand on a 64 bit system until the OS is available? Also really need PCI Express MOBO. But especially - how about an update on the OS. Perfect opportunity for an Apple OS-64 written especially for the AMD 64?!#* Reply
  • Falco. - Thursday, June 3, 2004 - link

    will the S754/940 heat sink and fans work with 939 ?? Reply
  • XRaider - Wednesday, June 2, 2004 - link

    Thanks Viditor, I appreciate it! :) I will have to keep my eyes open on this heat stuff and see what other people will state about this. Reply
  • TrogdorJW - Tuesday, June 1, 2004 - link

    Pretty interesting, but I'm not super surprised by the results. I would imagine that the 3700+ CPU would be about the same distance from the 3800+ as the 3400+ is from the 3500+. Given the prices, I really don't see much reason for celebrating the release (finally!) of socket 939.

    I did some speculation on some other forums about some related issues AMD and Intel seem to be encountering, for any who want to read a longer post:
    http://forums.firingsquad.com/firingsquad/board/me...
    Reply
  • Viditor - Tuesday, June 1, 2004 - link

    XRaider - To be clear, AMD and Intel actually report 2 different numbers when they talk about heat dissapation.
    AMD reports the MAXIMUM TDP FOR THE WHOLE CPU LINE (both now and in the future...) This means that 104watts is the max thermal design power (the absolute worst it could get) for all 939 cpus at 13nm.

    Intel reports ONLY THE "TYPICAL" TDP FOR THE SPECIFIED CHIP. This means that Intel runs a series of software (they won't release WHICH software they use) and measure the power at that time. They don't report the actual maximum theoretical thermal levels. Intel have a different TDP listed for each processor, and often a different TDP for different stepping of processors.

    So, to answer your question (sorry about the length), while the 939 line of parts have an absolute theoretical max of 104 watts, you will probably see it somewhere in the 80's...
    Reply
  • mikeymasta - Tuesday, June 1, 2004 - link

    These benchmarks look great, but I REALLY would appreciate it if we could see some OpenSSL benchmarks?
    Because I would love to get my boss to get an AMD based server but hes one of those know all "Intel' is always better type people
    In the server enviroment just an openssl benchmark could turn things around.
    Just install linux or FreeBSD, you could even use a live linux CD so you dont have to install and then just type "openssl speed"
    most linux dists would have openssl in
    /usr/bin/openssl
    /usr/bin/openssl speed > speed_log 2>&1
    to log it to a file
    We are thinking about getting the Sun based AMD
    http://www.sun.com/servers/entry/v20z/ but it costs a lot!

    If you could do that for all new CPU benchmarks I would be very happy.
    Reply
  • Pumpkinierre - Tuesday, June 1, 2004 - link

    #23 The french site said the spec is 104W and found the heatsink temp (applied above) to be 50C under load and 60C@1.8V overclocking (216MHz) on a 3800+ using a MSI K8Nneo2 (nf3-250) and Asus A8V (K8T800pro). Overall they stated that the cpu ran hot compared to S754 a64s:

    http://www.x86-secret.com/articles/cpu/s939/s939-7...

    Both mobos BIOSs defaulted to DDR333 with 4 double sided DIMMS.

    Xbit say that 104W is the overall spec for S939 but present 130nm cpus stay at 89W TDP (Prescott anyone?). However the temps (cpu diode-Asus A8V)are mildly frightening: 41-64 (idle-load) for the 3800+ but the 3500+ is the coolest 38-58. They even get 40-60C with the 3400+ (but dont quote ambient) which still makes me wonder about the coolness of these cpus. They got the 3500+ to 232MHz FSB and the 3800+ to 215 but didnt quote the voltage. They quoted the same AMD info for the DIMMS:
    http://www.xbitlabs.com/articles/cpu/display/athlo...

    All in all it looks like the heat issues arent confined to Intel biut Cool and Quiet might help out.
    Reply
  • Pollock - Tuesday, June 1, 2004 - link

    "In the final analysis, we aren't talking about the be all end all of platforms and performance, but, certainly, anyone who wants an Athlon 64 system should look no further than socket 939 for its flexibility, overclockability, and performance."
    I don't remember reading much about overclockability anywhere? Yeah, maybe a statement about the higher HT speed possibly helping, but I want to know more than that!
    Reply
  • SHO235V8 - Tuesday, June 1, 2004 - link

    Derek, what about the compatability issues and heat issues? Any news on these fronts and when will these parts be available? I have been waiting for the 939 for some time and my desktop gets slower everyday! Thanks ;) Reply

Log in

Don't have an account? Sign up now