Summary: the cores compared

Below, you find a comparison of the Intel Xeon/Pentium 4, the Opteron/Athlon 4, the G5 and the previous CPU in the Apple Power: the G4 of Motorola.

CPU feature

Motorola G4+

G5 (IBM PowerPC 970)

Intel Xeon P4 Irwindale

AMD Opteron Troy

Process technology

0.18 µ CU SOI

0.09 µ CU SOI

0.09 µ CU

0.09 µ CU SOI

GP Register Width (bit)

32

64

64

64

Number of transistors (Million)

33

58

169

106

Die Size (mm²)

106

66

+/-130 (112   for 1 MB L2)

115

Maximum Clockspeed (MHz)

1400

2700 (liquid cooled)

3800

2600

Pipeline Stages ( fp)

7

16 (21)

31 - 39*

12 (17)

issue rate (Instruction per clockcycle)

3 + 1 Branch

4 + 1 branch

4 ports, max. 6 (3 sustained)

6 (3 sustained)

Integer issue rate (IPC)

3 + 1 Branch

2

4 (3 sustained)

3

Floating point issue rate (IPC)

1

2

1

3

Vector  issue rate (IPC)

2-4 ( Altivec)

2-4 ( Altivec, velocity)

4  Single(SSE-2/3)

4  Single(SSE-2/3)

2 Double (SSE-2/3)

2 Double (SSE-2/3)

Load & Store units

1

2

2

2

"instructions in flight" (OOO Window)

16

215 (100)

126

72

Branch History Table size (entries)

2048

16384

4096

16384

L1-cache (Instruction/Data)

32 KB/32 KB

64 KB/32 KB

12k µops (+/- 8-16 KB)/16 KB

64 KB/64KB

L2-cache

256 KB

512 KB

2048 KB

1024 KB

L3-cache

2 MB DDR SRAM 64 bit at 1/4 th of core clock

none

None

none

Front Side Bus (MHz)

166

1350 (675 DDR)

800 (200 Quad)

N/A

Front Side Bus (GB/s)

1.3 Half Duplex

10,8 Full Duplex

6.4 Half Duplex

N/A

Memory Bandwidth (GB/s)

2.7

6.4

6.4

6.4

Core Voltage

1.6V

1,1V ?

1.38V

1.4V

Power Dissipation

30W at 1 GHz

+/- 59 (Typical) -80 Watt (max)

110 W (Typical)

92,6W (Max)


*31 is branch misprediction pipeline length, 39 is the length of the total pipeline including decoding stages before the trace cache.

Let us summarize: in theory, the PowerPc 970FX is a very wide, deeply pipelined superscalar monster chip, with excellent Branch prediction and fantastic features for streaming applications. And let us not forget the two parallel FPUs and the SIMD Altivec unit, which can process up to 4 calculations per clock cycle.

The disadvantages are the rather coarse way that the 970FX handles the instruction flow and the high latency to the RAM.

Enough theory. Let us see how the G5 2.5 GHz and 2.7 GHz compares to the 3.6 GHz Xeon Irwindale and Opteron 250 (2.4 GHz). The Opteron 852 arrived just a day before my deadline, but I think that you will know how the 252 performs compared to the 250. Before we tackle performance, here are a few quick notes about power dissipation.

Power to the PowerPC

How power thirsty is this PowerPC 970FX? His predecessor, the 0.13µ SOI PowerPC 970 was a pretty cool chip. It consumed about 42W at 1.8 GHz (1.3v). The newer 0.09µ SOI PowerPC 970FX CPU is reported to dissipate about 55-59W at 2.5 GHz. However, a few annotations must be made.

First of all, IBM and Apple tend to increase the core voltage when running at higher clock speed. This makes the needed power increase more than linearly. For example, the 1.8 GHz PowerPC 970 consumed 42 Watt, but the 2 GHz version (both 0.13µ CPUs) needed 66 Watt.

Secondly, the TDP IBM talks about is typical , not maximum like AMD's.

Let us clarify this by checking IBM's and Apple's numbers. For the 90 nm, IBM's own documents tell us that the PowerPC 970FX only consumes 24.5 Watt at 2 GHz (1V). However, the same 0.09µ SOI PowerPC970FX is reported to consume about 55W at 2.3 GHz (1.1V?) in the Xserve, according to Apple's own website. Typically, you would expect the G5 to consume about 28 Watt (24.5 * 2.3 / 2) at 2.3 GHz, when using the 24.5 Watt at 2 GHz as a reference. Apple talks about "at most" (maximum), and IBM about "typical".

Still, that is a huge gap between "typical" and "maximum" power dissipation. The 55 Watt number seems to indicate that the core voltage must have been increased significantly at 2.3 GHz. The maximum power dissipation of the 2.5/2.7 GHz G5 inside the liquid-cooled PowerMacs might thus be quite a bit higher than in the 1U Xserve, probably around 80 Watt for the 2.7 GHz. That is a lot of power for a 66 mm² CPU, and it probably explains why Apple introduced liquid cooling. The liquid cooling system inside our PowerMac wouldn't get warm and wouldn't be necessary at all if the two 2.5 GHz CPUs were only dissipating a 59 Watt maximum.

IBM PowerPC 970FX: Superscalar monster Benchmark configuration
Comments Locked

116 Comments

View All Comments

  • seanp789 - Saturday, June 4, 2005 - link

    well thats great and all but yours news says apple is switchign to intel so i dont think much will be changing in the power pc lineup
  • Brazilian Joe - Saturday, June 4, 2005 - link

    I would like to see this Article re-done, with more benches to give a clearer picture. I think MACOS X should be pitched against Darwin in the PPC platform, since there may be hidden differences. Darwin works on x86 too (and x86_64?), it would be very interesting to see the SAME OS under the Mac Platform running on different hardware. And having the software compiled with the same compiler present on Darwin, we should get a more consistent result. Linux and BSD should not be ditched, however. The perfornance difference Of linux/FreeBSD/OpenBSD in PPC vs PC is also a very interesting subject to investigate.
    I think this article, along with all the complaints of inconsistency in the results, sohuld fuel a new series of articles: One, Just comparing Darwin/MacOS X on Both platforms. Another For Linux, using a GCC version as close as possible to that used on Darwin. Another for FreeBSD, and yet another for OpenBSD. The last article Would get everything and summarize. I think this would be much more complete and satisfying/informative for the reader crowd.
  • iljitsch - Saturday, June 4, 2005 - link

    There seems to be considerable confusion between threads and processes in the review. I have no trouble believing that MacOS doesn't do so well with process gymnastics, but considering the way Apple itself leverages threads, I would assume those perform much better.

    I don't understand why Apache 1.3 was used here, Apache 2.0 has much better multiprocessor capabilities and would have allowed to test the difference between the request-are-handled-in-processes and requests-are-handled-in-threads ways of doing things.
  • Phil - Saturday, June 4, 2005 - link

    #79: Wow. I had no idea that they were actually going to do it, I had assumed that it was typical industry nonsense!
    If this is true, then IMHO Apple won't be in much of a better position (with regards to this article) as they'll still need to work on the OS, regardless.

    Can anyone speculate as to why they *really* want to switch to x86/Intel? I wonder if they'll consider AMD too...
  • rorsten - Saturday, June 4, 2005 - link

    Uhm, the estimation for power consumption is completely wrong. The only significant CMOS power consumption - especially for an SOI chip - is the current required to charge or discharge the gates of the FETs, which only happens when a value changes (the clock accounts for most of the power consumption on a modern synchronous chip). Since we're talking about current only, this is purely resistive power, I^2R style, and since the current is related to the number of transitions per second, increasing the clock rate linearly increases the current which quadratically increases the power consumption.
  • kamper - Saturday, June 4, 2005 - link

    Here's another story about Apple and Intel from cnet:
    http://news.com.com/Apple+to+ditch+IBM%2C+switch+t...

    Interesting in the context of this article but I won't believe it without much more substantial proof :)

    +1 on getting a db test using the same os on all architectures whether it be linux or bsd

    +1 on fixing the table so that it renders in firefox
  • shanep - Saturday, June 4, 2005 - link

    Re: NetBSD.

    Sorry, I just noticed it is not supported yet by NetBSD.

    Forget I mentioned it.
  • shanep - Saturday, June 4, 2005 - link

    "Wessonality: Our next project if we can keep the G5 long enough in the labs."

    How about testing these machines with NetBSD 2.0.2 to keep the hardware comparison on as close an equal footing as possible.

    This should mostly remove many red herrings associated with multiple differences in software across different hardware.
  • michaelok - Saturday, June 4, 2005 - link

    "i've had for awhile about OS X server handling large numbers of thread. My OS X servers ALWAYS tank hard with lots of open sessions, so i keep them around only for emergencies. T"

    Moshe Bar (openMosix) has been an avid Mac follower for years, I see he has a few suggestions for OSX, including ditching the Mach so you can run FreeBSD natively, which has much better peformance. In fact, thread performance is one of FreeBSDs strong points, although Linux has largely caught up.

    Also research his Byte articles, you can see how a proper comparison can be done, although he does not claim to be a benchmarking expert.

    http://www.moshebar.com/tech/?q=node/5
    http://www.byte.com/documents/s=7865/byt1064782374...
  • johannesrexx - Saturday, June 4, 2005 - link

    Everybody should use Firefox by default because it's far more secure. Use IE only when you must.

Log in

Don't have an account? Sign up now