Cache and Memory Performance

I mentioned earlier that cache latencies are higher in order to accommodate the larger caches (8MB L2 + 8MB L3) as well as the high frequency design. We turned to our old friend cachemem to measure these latencies in clocks:

Cache/Memory Latency Comparison
  L1 L2 L3 Main Memory
AMD FX-8150 (3.6GHz) 4 21 65 195
AMD Phenom II X4 975 BE (3.6GHz) 3 15 59 182
AMD Phenom II X6 1100T (3.3GHz) 3 14 55 157
Intel Core i5 2500K (3.3GHz) 4 11 25 148

Cache latencies are up significantly across the board, which is to be expected given the increase in pipeline depth as well as cache size. But is Bulldozer able to overcome the increase through higher clocks? To find out we have to convert latency in clocks to latency in nanoseconds:

Memory Latency

We disable turbo in order to get predictable clock speeds, which lets us accurately calculate memory latency in ns. The FX-8150 at 3.6GHz has a longer trip down memory lane than its predecessor, also at 3.6GHz. The higher latency caches play a role in this as they are necessary to help drive AMD's frequency up. What happens if we turn turbo on and peg the FX-8150 at 3.9GHz? Memory latency goes down. Bulldozer still isn't able to get to main memory as quickly as Sandy Bridge, but thanks to Turbo Core it's able to do so better than the outgoing Phenom II.

L3 Cache Latency

L3 access latency is effectively a wash compared to the Phenom II thanks to the higher clock speeds enabled by Turbo Core. Latencies haven't really improved though, and Bulldozer has a long way to go before it reaches Sandy Bridge access latencies.

The Impact of Bulldozer's Pipeline Windows 7 Application Performance
Comments Locked

430 Comments

View All Comments

  • actionjksn - Thursday, October 13, 2011 - link

    AMD Nailpuller? That was some funny shit right there HA HA HA
  • Spam not Spam - Wednesday, October 12, 2011 - link

    Just skimmed the review; not as awesome as I had hoped for, sadly. That being said, I'm thinking it might well be a nice improvement for the stock, C2D Q6600 in my Dell. I could go Intel, obviously, but... I dunno. I've got an odd fascination with novel things, even if they are rough to begin with. Hell, I've even got a WP7 phone :p
  • wolfman3k5 - Wednesday, October 12, 2011 - link

    Then make sure to get a quality power supply and motherboard to go with it. Also, your power bill will increase, but not directly from the Bulldozer CPU, nope, but from all the heat that it will make... you will need to run your air conditioner which is a power hog.

    /* Patiently waiting for AMD's next gen architecture codenamed "Bendover" */
  • ckryan - Wednesday, October 12, 2011 - link

    Is 'ScrewdOver' next on the roadmap after 'Bendover'? I'll have to look in the official AMD leaked slide repository.

    I still think some intrepid AMD faithful will try BD out just because they're wired that way, and many of the are going to like it. I bet it compares better to Lynnfield than Sandy Bridge... Except Ivy Bridge is closer in the future than SB's launch is in the past. This could be an interesting and relevant product after a few years, but the need is dire now. AMD is going to kill off the Phenom II as fast as possible.
  • themossie - Wednesday, October 12, 2011 - link

    Bendover -> ScrewdOver -> Screwdriver (I'll bring the OJ) -> Piledriver.
    Courtesy of numerous internal leaks at AMD.
  • themossie - Wednesday, October 12, 2011 - link

    My apologies, didn't realize Piledriver was real.
  • bill4 - Wednesday, October 12, 2011 - link

    AKA, the reason all of us who are commenting are reading this review. Gaming performance. And AMD chose not to even compete there. Bunch of monkey overs at AMD CPU engineering?

    It's now a non starter in the enthusiast market.

    I've often though recently that AMD (or any manufacturer really, but AMD as a niche filler would be a more obvious choice given their market position) would do well to try to position itself as the gamers choice, and even design it's CPU's to excel in gaming at the expense of some other things at times. I really suspect this strategy would lead to a sales bonanza. Because really the one area consumers crave high performance is pretty much, only gaming. It's the one reason you actually want a really high performance CPU (provided you dont do some sort of specialized audio/video work), instead of just "good enough" which is fine for general purpose desktoping.

    Instead they do the exact opposite with Bulldozer, facepalm. Bulldozer is objectively awful in gaming. Single handedly nobody who posts at any type of gaming or gaming related forum will ever buy one of these. Unbelievable.

    Perhaps making it even more stinging is there was some pre-NDA lift supposed reviewer quote floating around at about how "Bulldozer will be the choice for gamers" or something like that. And naturally everybody got excited because, that's all most people care about.

    Combine that with the fact it's much bigger and hotter than Intel's, it's almost a unmitigated disaster.

    This throws AMD's whole future into question since apparently their future is based on this dog of a chip, and even makes me wonder how long before AMD's engineers corrupt the ATI wing and bring the GPU side to disaster? The ONLY positive thing to come out of it is that at least AMD is promising yearly improvements, key word promising. Even then absolute best case scenario is that they slowly fix this dog in stages, since it's clearly a broken architecture. And that's best case, and assumes they will even meet their schedule.

    Anand lays so much of the blame at clockspeed, hinting AMD wanted much more. But even say, 4.3 ghz Bulldozer, would STILL be a dog in all important gaming, so there's little hope.
  • shompa - Wednesday, October 12, 2011 - link

    I have used many AMD systems. Have deployed 1000 of AMD CPU inside Unix workstations at my old work. I cheer for AMD.
    But.
    AMD is going to have a hard time ahead. Selling its fabs to Global foundries was the biggest mistake of them all.

    We are in the post PC world. If Tablets are computers: 2012 20% of PCs will use ARM. This is many lost CPU sales for AMD/Intel.

    I predict that AMD will be gone within 3 years. Maybe someone buys them? After the settlement with Intel, AMD now can transfer its X86 license to the next buyer. (pending Intels approval)

    Maybe Google could buy AMD and build complete computers ?
  • wolfman3k5 - Wednesday, October 12, 2011 - link

    I see two things that might happen to AMD
    1) They will transform in to a GPU manufacturer completely (and of course they will make those silly APUs)
    2) If that damn x86 license is transferable, they could merge with NVidia. Neither of these two companies looks to hot these days, so they might as well work together.
  • philosofool - Thursday, October 13, 2011 - link

    We may be in the "post PC" era, but don't count x86 out. Recent studies indicate there's a corollary to Moore's law that applies to compute power per watt; the study goes back to 1961. This suggests that x86 is only a few years away from running on mobile devices, which is what MS and Intel are betting on. And frankly, it makes sense. Ultimately, I don't want two different things (a mobile device and a PC), I want a PC in my pocket and one on my desk.

Log in

Don't have an account? Sign up now