5 ms·

So my question is - how much extra power does the IBM Power8 cpu need in order to get performance that is on par with Intel's Haswell?

kev009 11y ago
Look at the memory bandwidth numbers. These completely outclass the Xeon in certain types of workloads, apples to oranges.
- ksec 11y ago
  I thought after all these years and improvement Intel is invisible in performance. Turns out i was wrong. Although I would still buy Intel Chips for future prove. IBM is simply too incompetent to compete.
  - dorfsmay 11y ago
    Until about 5 years ago the Power architecture dominated in the CPU performance domain. Saying "IBM is too incompetetent" makes you sound like you don't really know this market.
    - Tinyyy 11y ago
      Hey, I wanna know more, can you elaborate?
      - wtallis 11y ago
        Intel's only got the ability to develop one microarchitecture family. When they try for more than one, things fall through the cracks and you get the Pentium 4 dead-end or the Atom's lackluster performance. Intel's primary microarchitecture is aimed at laptops. It scales reasonably down to large tablet power levels and up to workstation power levels. For the high-performance server market, they can only throw cores and cache at the problem, with some enterprise features bolted on. IBM's always targeted the high-performance server market. For a while, their Power cores were also being used for workstations and even a desktop by Apple, but that's never been the focus. They include things like decimal arithmetic and SMT and hardware transactional memory and they've been selling the high-end parts at 4-5GHz speeds for a long time.
        cbd1984 11y ago
        OK, what's Intel missing for high-performance servers? They've got respectable performance, they've got VT-x and all the other virtualization hardware, what's missing?
        dijit 11y ago
        their underlying architecture wasn't built for it. it's like trying to build a skyscraper on the foundations of a log cabin.. you can get so far up and things will start to sink back into the mud.
        the_why_of_y 11y ago
        The x86 architecture was not amenable to a high performance pipelined implementation, so what Intel did since Pentium Pro is to JIT the x86 instructions into an internal RISCy instruction set that can be executed out-of-order with competitive performance. The x86 architecture was limited to 32-bit, severely limiting the virtual address space as well as hampering OS implementation with hacks like PAE, so what AMD did is to extend it to 64-bit. The x86 architecture was not designed for SMP scalability due to the rather strict memory ordering requirements, but most of the architectures with laxer memory models (in particular, Alpha, which was the most lax of all) are out of business today (in fact, of the commercially relevant server architectures today, only POWER has lax memory ordering; SPARC can in theory but is usually configured to run with TSO which is similar to x86). The x86 architecture was not designed for OS virtualization because various instructions did not trap when executed in user mode, so what Intel and AMD did is define a new protection level ("ring -1") to run the hypervisor so this works efficiently now. What actual problem do you see with x86 that cannot be solved or worked around by some creative engineering at Intel/AMD?
        gsnedders 11y ago
        > The x86 architecture was not amenable to a high performance pipelined implementation, so what Intel did since Pentium Pro is to JIT the x86 instructions into an internal RISCy instruction set that can be executed out-of-order with competitive performance. POWER does the exact same (I can't remember which revision).
    - ksec 11y ago
      Because I really dont. Yes Power used to dominated performance. I have always thought with every tick and tock Intel has closed the gap or leap forward already. But where is it used? Is it shrinking? I have no idea, at least everything have shown http://www.theplatform.net/2015/06/04/x86-servers-dominate-the-datacenter-for-now/ http://www.theplatform.net/2015/06/04/x86-servers-dominate-t... Intel is still winning / dominating, there were talks of Google using POWER8, haven't seen anything more. And knowing Intel with a clear Roadmap coming up. Do Power8 have any place in Cloud Computing? I too, wish Intel would have more competition, but i dont see the choice of using it. Most of the distributed and In Memory Processing aren't using it. But then I guess i am not the target audience.
      - gsnedders 11y ago
        In general, what's kept POWER where it is is that you need to be willing to pay far higher prices than Intel charge, and have far higher power consumption than high-end Intel parts, and have the cooling to cope with far higher exhaust temperatures.
- StillBored 11y ago
  Yes if you need more memory bandwidth, POWER is where its at. OTHO, as I've been saying the base cores don't do as well on single threaded loopy or problems with a decent L1/L2 cache hit rate. The specint_rate numbers look good because of the x8 threading, the single core results probably don't look that good. So for many workloads it won't be that great. For the one I was working on out of the box, the power was 1/2 as fast. But that wasn't fair because we had a couple highly optimized x86 code paths. Doing some basic POWER optimizations brought the performance in line with the x86. But on some benchmarks it would win, and lose on others. So while we utilized a _LOT_ of memory/IO bandwidth the fact that there was nearly 2.5x available in the power system over our E5 didn't give us enough of a boost to make it worth the higher price tag (nearly 3x in our case, cause we were comparing with a supermicro machine). Maybe this newer power machine changes that a little.
  - vardump 11y ago
    Supermicro is where it's at. 10k buys you 4x dual socket 6-core Xeon E5-v3 servers (2U Twin^2). Totally 48 cores, 256 GB DDR4 RAM (minimum config).
Sanddancer 11y ago
They use less power from what I can tell. The IBM's server has a TDP of 190 watts, and the equivalent Xeons are 135 watts a piece, or 270 watts total.

kev009 11y ago

Look at the memory bandwidth numbers. These completely outclass the Xeon in certain types of workloads, apples to oranges.

ksec 11y ago
I thought after all these years and improvement Intel is invisible in performance. Turns out i was wrong. Although I would still buy Intel Chips for future prove. IBM is simply too incompetent to compete.
- dorfsmay 11y ago
  Until about 5 years ago the Power architecture dominated in the CPU performance domain. Saying "IBM is too incompetetent" makes you sound like you don't really know this market.
  - Tinyyy 11y ago
    Hey, I wanna know more, can you elaborate?
    - wtallis 11y ago
      Intel's only got the ability to develop one microarchitecture family. When they try for more than one, things fall through the cracks and you get the Pentium 4 dead-end or the Atom's lackluster performance. Intel's primary microarchitecture is aimed at laptops. It scales reasonably down to large tablet power levels and up to workstation power levels. For the high-performance server market, they can only throw cores and cache at the problem, with some enterprise features bolted on. IBM's always targeted the high-performance server market. For a while, their Power cores were also being used for workstations and even a desktop by Apple, but that's never been the focus. They include things like decimal arithmetic and SMT and hardware transactional memory and they've been selling the high-end parts at 4-5GHz speeds for a long time.
      - cbd1984 11y ago
        OK, what's Intel missing for high-performance servers? They've got respectable performance, they've got VT-x and all the other virtualization hardware, what's missing?
        dijit 11y ago
        their underlying architecture wasn't built for it. it's like trying to build a skyscraper on the foundations of a log cabin.. you can get so far up and things will start to sink back into the mud.
        the_why_of_y 11y ago
        The x86 architecture was not amenable to a high performance pipelined implementation, so what Intel did since Pentium Pro is to JIT the x86 instructions into an internal RISCy instruction set that can be executed out-of-order with competitive performance. The x86 architecture was limited to 32-bit, severely limiting the virtual address space as well as hampering OS implementation with hacks like PAE, so what AMD did is to extend it to 64-bit. The x86 architecture was not designed for SMP scalability due to the rather strict memory ordering requirements, but most of the architectures with laxer memory models (in particular, Alpha, which was the most lax of all) are out of business today (in fact, of the commercially relevant server architectures today, only POWER has lax memory ordering; SPARC can in theory but is usually configured to run with TSO which is similar to x86). The x86 architecture was not designed for OS virtualization because various instructions did not trap when executed in user mode, so what Intel and AMD did is define a new protection level ("ring -1") to run the hypervisor so this works efficiently now. What actual problem do you see with x86 that cannot be solved or worked around by some creative engineering at Intel/AMD?
        gsnedders 11y ago
        > The x86 architecture was not amenable to a high performance pipelined implementation, so what Intel did since Pentium Pro is to JIT the x86 instructions into an internal RISCy instruction set that can be executed out-of-order with competitive performance. POWER does the exact same (I can't remember which revision).
  - ksec 11y ago
    Because I really dont. Yes Power used to dominated performance. I have always thought with every tick and tock Intel has closed the gap or leap forward already. But where is it used? Is it shrinking? I have no idea, at least everything have shown http://www.theplatform.net/2015/06/04/x86-servers-dominate-the-datacenter-for-now/ http://www.theplatform.net/2015/06/04/x86-servers-dominate-t... Intel is still winning / dominating, there were talks of Google using POWER8, haven't seen anything more. And knowing Intel with a clear Roadmap coming up. Do Power8 have any place in Cloud Computing? I too, wish Intel would have more competition, but i dont see the choice of using it. Most of the distributed and In Memory Processing aren't using it. But then I guess i am not the target audience.
    - gsnedders 11y ago
      In general, what's kept POWER where it is is that you need to be willing to pay far higher prices than Intel charge, and have far higher power consumption than high-end Intel parts, and have the cooling to cope with far higher exhaust temperatures.
StillBored 11y ago
Yes if you need more memory bandwidth, POWER is where its at. OTHO, as I've been saying the base cores don't do as well on single threaded loopy or problems with a decent L1/L2 cache hit rate. The specint_rate numbers look good because of the x8 threading, the single core results probably don't look that good. So for many workloads it won't be that great. For the one I was working on out of the box, the power was 1/2 as fast. But that wasn't fair because we had a couple highly optimized x86 code paths. Doing some basic POWER optimizations brought the performance in line with the x86. But on some benchmarks it would win, and lose on others. So while we utilized a _LOT_ of memory/IO bandwidth the fact that there was nearly 2.5x available in the power system over our E5 didn't give us enough of a boost to make it worth the higher price tag (nearly 3x in our case, cause we were comparing with a supermicro machine). Maybe this newer power machine changes that a little.
- vardump 11y ago
  Supermicro is where it's at. 10k buys you 4x dual socket 6-core Xeon E5-v3 servers (2U Twin^2). Totally 48 cores, 256 GB DDR4 RAM (minimum config).

Sanddancer 11y ago

They use less power from what I can tell. The IBM's server has a TDP of 190 watts, and the equivalent Xeons are 135 watts a piece, or 270 watts total.