POWER7 Spec

By: Michael S (already5chosen.delete@this.yahoo.com), February 10, 2010 12:30 pm
Room: Moderated Discussions
Linus Torvalds (torvalds@linux-foundation.org) on 2/10/10 wrote:
---------------------------
>someone (someone@somewhere.com) on 2/10/10 wrote:
>>
>>[bunch of same old Linus shit deleted]
>
>I note that the deleted part was where all the numbers
>that you didn't actually reply to were.
>
>It's a fact that Nehalem (and Core 2) gets more
>work-per-cycle than Itanium has, despite Itanium running
>at much lower clock frequencies.
>

Well, the fact that your numbers demonstrate is that fast-clocked Core2-6M with fast FSB and fast main memory (i.e. non-FBD) gets the same work-per-cycle as Itanium. Only when you go down in frequency to about 2GHz you get a little more work-per-cycle, but only on well-tuned system.
Core2-4M gets somewhat lower work-per-cycle than Itanium.

>You can delete the numbers all you want, but you just
>look stupid for doing so. Your arguments about memory
>models are all totally pointless - exactly because of the
>whole issue of the x86 chips running at a much higher
>frequency
.
>
>Don't you get it? I've compared performance cycle-to-cycle.
>Itanium got an almost 2x advantage because it's running at
>roughly half the speed, so the Itanium memory subsystem
>is not at all inferior to the Core 2 one I compared to.
>
>Yes, in absolute frequency, those Core 2 chips had higher
>FSB frequencies, but when normalizing to CPU cycles the
>frequencies are pretty much the same.

When normalized to CPU cycles the Core2 FSB frequencies are lower, much lower.
Don't forget that Core2 has 64-bit quad-pumped FSB vs Itanium's 128-bit double-pumped FSB.
So C2D-1333=Itanium-666 both in latency and in throughput. In fact, latency-wise Itanium-666 could be even few ns faster, because it uses simpler clock-synchronous bus protocol.

>So the Itanium box
>basically got memory that was twice as fast, had bigger
>caches, was the absolute top-of-the-line best gcc result
>you can find.

That doesn't follow. Itanium FSB is faster but MCH is somewhat slower and memory itself is significantly slower. I'd guess random access on rx2660 is in 85ns range vs 55-to-60 ns on desktop C2D. So, per clock, Itanium is 25-30% faster rather than twice faster.

Now, when you compare vs FBD-based Xeons then, indeed, Itanium is about twice faster per clock, but in this case its per clock 403.gcc score is higher as well.

The best Xeon-3333/FBD 403.gcc score is 21.9/16.9
403.gcc most definitely doesn't like FBDIMMs.

>
>And Itanium still got less work done per cycle.
>
>Comprende?
>
>Yes, you can have your wet dreams about trying to skew the
>thing even more in Itaniums favor, but if you were honest
>you'd just admit that by doing the clock-for-clock
>comparison, I already gave Itanium a huge head-start. And
>it still finished last.
>
>Linus

I disagree. If that was the case low-clocked FBD-based Xeons would have higher scores than Itanium. It doesn't happen.

The best Xeon-1867/FBD 403.gcc score =13.7/11.2 i.e. ~5% lower per clock than the best of Itanium.

http://www.spec.org/cpu2006/results/res2009q1/cpu2006-20081219-06310.html
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
POWER7 SpecParadox2010/02/08 11:05 AM
  POWER7 SpecThu Nguyen2010/02/08 12:58 PM
    POWER7 SpecIan Ameline2010/02/08 09:22 PM
      POWER7 SpecThu Nguyen2010/02/08 11:54 PM
        POWER7 SpecIan Ameline2010/02/09 06:46 AM
          POWER7 SpecLinus Torvalds2010/02/09 07:57 AM
            industry-standard single-threaded performance benchmarks absentMichael S2010/02/09 08:09 AM
              industry-standard single-threaded performance benchmarks absentParadox2010/02/09 08:33 AM
                industry-standard single-threaded performance benchmarks absentMichael S2010/02/09 09:30 AM
                  industry-standard single-threaded performance benchmarks absentParadox2010/02/09 10:52 AM
                    industry-standard single-threaded performance benchmarks absentJesper Frimann2010/02/09 11:33 AM
              industry-standard single-threaded performance benchmarks absentLinus Torvalds2010/02/09 08:48 AM
                industry-standard single-threaded performance benchmarks absentMichael S2010/02/09 09:26 AM
                  industry-standard single-threaded performance benchmarks absentLinus Torvalds2010/02/09 09:58 AM
                    industry-standard single-threaded performance benchmarks absentJesper Frimann2010/02/09 12:17 PM
                      industry-standard single-threaded performance benchmarks absentLinus Torvalds2010/02/09 12:54 PM
                        industry-standard single-threaded performance benchmarks absentJesper Frimann2010/02/09 02:10 PM
                    industry-standard single-threaded performance benchmarks absentParadox2010/02/09 01:22 PM
                industry-standard single-threaded performance benchmarks absentanon2010/02/09 10:21 AM
            POWER7 SpecJesper Frimann2010/02/09 12:30 PM
            POWER7 SpecDoug Siebert2010/02/09 05:38 PM
              POWER7 SpecLinus Torvalds2010/02/09 07:28 PM
                POWER7 SpecLinus Torvalds2010/02/09 07:28 PM
                  POWER7 SpecIan Ameline2010/02/09 08:02 PM
                    POWER7 SpecDoug Siebert2010/02/09 10:18 PM
                POWER7 Specsomeone2010/02/09 08:20 PM
                  POWER7 SpecLinus Torvalds2010/02/10 09:17 AM
                    POWER7 SpecLinus Torvalds2010/02/10 09:46 AM
                      POWER7 SpecEmil2010/02/10 11:06 AM
                    POWER7 SpecIan Ameline2010/02/10 10:13 AM
                    POWER7 Specsomeone2010/02/10 11:01 AM
                      POWER7 SpecLinus Torvalds2010/02/10 11:32 AM
                        POWER7 SpecMichael S2010/02/10 12:30 PM
                          POWER7 SpecLinus Torvalds2010/02/10 01:25 PM
                            POWER7 Specmpx2010/02/10 02:58 PM
                        POWER7 Specnemlis2010/02/11 12:24 AM
                          POWER7 Specnone2010/02/11 12:52 AM
                            POWER7 Specnemlis2010/02/11 01:52 PM
          POWER7 Specmpx2010/02/09 08:18 AM
          POWER7 SpecMichael S2010/02/09 09:08 AM
        POWER7 SpecJesper Frimann2010/02/16 02:29 AM
          POWER7 SpecJamie Lucier2010/02/16 06:35 AM
            POWER7 Specanon2010/02/16 07:16 AM
              POWER7 SpecJamie Lucier2010/02/16 10:46 AM
                POWER7 Specanon2010/02/16 03:42 PM
                  POWER7 SpecJamie Lucier2010/02/18 06:07 AM
              POWER7 SpecThu Nguyen2010/02/16 01:23 PM
            POWER7 SpecParadox2010/02/16 09:57 AM
              POWER7 SpecJamie Lucier2010/02/16 11:22 AM
                POWER7 SpecParadox2010/02/16 02:00 PM
            POWER7 SpecJesper Frimann2010/02/16 02:14 PM
              POWER7 SpecJamie Lucier2010/02/17 10:41 AM
  POWER7 SpecJesper Frimann2010/02/09 04:43 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell tangerine? 🍊