TigerDirect lists Bulldozer, 990FX bundle a little early

No, AMD still hasn’t pulled the trigger on the desktop version of Bulldozer—at least, no announcement has hit the wires quite yet. Tiger Direct must have jumped the gun, then, because it’s already listing a bundle made up of one Asus Sabertooth 990FX motherboard and one AMD FX-6100 processor of the Bulldozer variety.

The bundle will set you back $379.99, a purported $9.99 discount over the cost of the two components on their own: $199.99 for the motherboard and $189.99 for the processor. According to the listing, the FX-6100 has six cores, a 3.3GHz clock speed, 14MB of cache, a 95W thermal envelope, and a Socket AM3+ package. The Sabertooth 990FX is somewhat less new and exciting, because we reviewed it back in May.

Premature listings for FX-series chips, including the FX-6100, actually appeared almost a month ago. This time, though, we’re talking about a major e-tailer with a listing that has yet to be taken down. TigerDirect quotes a lead time of "7 – 21 Days," which means Bulldozer must be getting close. (Thanks to TR forum gerbil NeronetFi for the link.)

Comments closed
    • burntham77
    • 8 years ago

    Wow, 95 watts on a 3.3 ghz 6-core chip? That’s really good! Christmas time is here again.

    • sschaem
    • 8 years ago

    A bit more clarity ?
    [url<]http://www.cpubenchmark.net/high_end_cpus.html[/url<] From this, and a price of $245, the FX-8150 will deliver the most bang for the $ based on the pass mark type of workload. [url<]http://www.passmark.com/products/pt.htm[/url<] Unknown: with a turbo of 4.2ghz, can lightly threaded games still be better then on a i7-2500k?

      • ermo
      • 8 years ago

      Yeah, the Xeon E3 1245 runs at 2500K speed (3.3 GHz, 3.7 GHz boost) with hyperthreading enabled and 8M cache and the FX-8150 is hot on its heels.

      Not bad at all, though it does look like the 2600K / E3 1275 (3.4 GHz, 3.8 GHz boost) is just out of reach.

      Of note, the Ph II X4 965 @ 3.4GHz scores 4289, which is around half of the FX-8150.

      • shaq_mobile
      • 8 years ago

      there have been a few “benchmarks” that were leaked. are these results verified? not that i doubt the results, but after all this article is about jumping the gun…

    • HisDivineOrder
    • 8 years ago

    “at least, no announcement has hit the wires quite yet” and “This time, though, we’re talking about a major e-tailer with a listing that has yet to be taken down. TigerDirect quotes a lead time of “7 – 21 Days,” which means Bulldozer must be getting close.”

    Soooo… trying to send a message on the sly or just being wistful? 😉

    • ronch
    • 8 years ago

    Good thing Bulldozer isn’t a cure for cancer. If it is, I can’t imagine how many would have died just because it’s so delayed.

      • Bensam123
      • 8 years ago

      …and yet there still is no cure for cancer.

    • Chrispy_
    • 8 years ago

    Benchmarks or shut up.

    The Bulldozer bandwagon is now, [i<]officially[/i<], more hot air than the Infineon Labs Phantom Gaming Console or the late Duke Nukem Forever. The opterons based on Bulldozer shipped over a month ago, but we have heard nothing about how the architecture is doing. NOT EVEN A SQUEAK.

      • DavidC1
      • 8 years ago

      The real meaning of “shipped”. That’s when AMD(it can be anyone) tells the system builders, “Here, you can start gathering the components and plans so you can sell it when we are REALLY finished with the product.”

    • ronch
    • 8 years ago

    Powerpoint slides, road maps, performance rumors, YouTube videos, price lists, die shots, a spunky FX website (“FX – Coming Soon!”), AM3+ motherboards, box shots, DIRT demos…

    But can you buy it already? No. Do you know how good it is, exactly? No.

    But so what?! It’s now listed on TigerDirect!!!

      • thermistor
      • 8 years ago

      Every time I read “Coming Soon!” I think of the joke in History of the World.

      The Romans have a god for everything, they even have a god for premature ejaculation. He’s not here, but I hear he’s Coming Soon!

        • ronch
        • 8 years ago

        [quote<]The Romans have a god for everything[/quote<] FX will be the NEW god of processors!

      • chuckula
      • 8 years ago

      Hey there ronch… you forgot about the comic book — er — “graphic novel” that AMD marketing made! Remember, a 216 mm^2 Sandy bridge is a huge fat bloated monster compared to AMD’s 8 core super-efficient 315 mm^2 die.. the marketing told me so!

        • ronch
        • 8 years ago

        Oh yeah. That. Thanks, chuckula..

    • tbone8ty
    • 8 years ago

    Can’t wait!

    I’ll be playing in my sandbox next week!

    Bulldozer for the win!

      • MadManOriginal
      • 8 years ago

      I can literally see the AMD flag you’re waving when I read your post.

    • Dposcorp
    • 8 years ago

    “Item Number: B69-5413

    Model: SABERTOOTH990FX Bundle

    Shipping: Usually Ships in 7 – 21 Days

    Guess we know what the launch window is.

    • NeronetFi
    • 8 years ago

    TD actually has a couple bundles listed. Some say they ship in 24 hrs 😛 I think these are just place holders getting ready for possible launch next week 🙂

    PS: Cyril Thank you for the props in the article 🙂

      • sschaem
      • 8 years ago

      Not sure if those are place holder. if you navigate to MOTHERBOARDS/SOCKET AM3+ those bundles are listed on the first page.

    • flip-mode
    • 8 years ago

    In single threaded tasks this setup would probably be a little faster than my 3.2 GHz Phenom II X4.

      • sschaem
      • 8 years ago

      Yes, Turbo will make it so the worse case is equal performance.
      But for code that use SSE4, AVX, FMA, the FX-6100 could have a 30% lead in single threaded situation.

      Unknown.. iddle power consumption. But since this is very important to the server space, the noise, heat and power consumption might be much slower then the X4.

      If you have a Radeon GPU, I can bet that AMD optimized the driver to death to take advantage of all the FX features giving the FX an edge over the SSE2 only capable Phenom II.

        • Goty
        • 8 years ago

        Idle power consumption should be much improved over Phenom II, mainly due to the fact that Bulldozer should implement power gating, not to mention any improvements you’ll see when moving on from 45nm.

        • ermo
        • 8 years ago

        For the record, Phenom I and II both support SSE3 and SSE4a (though not the full SSE 4.x instruction set extensions). So you’d think that AMD would have at least gone for SSE3 optimizations already?

        The last AMD CPUs to only support up to SSE2 were the earlier K8 steppings, and SSE3 support was added in the later 90nm and 65nm steppings, though they didn’t support single cycle SSE3 operations like intel and had roughly half the performance in benchmarks that targeted SSE3 for that very reason. One of the improvements in Barcelona/K10 was single cycle SSE3 support.

        Point in case (look for ‘*’ in the ‘flags:’ line):

        [code<] ermo@dante ~ $ cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 16 model : 6 model name : AMD Athlon(tm) II X2 240e Processor stepping : 2 cpu MHz : 800.000 cache size : 1024 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 apicid : 0 initial apicid : 0 fpu : yes fpu_exception : yes cpuid level : 5 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm 3dnowext 3dnow constant_tsc rep_good nopl nonstop_tsc extd_apicid *pni monitor cx16 popcnt lahf_lm cmp_legacy svm extapic cr8_legacy abm *sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt npt lbrv svm_lock nrip_save bogomips : 5614.24 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 48 bits physical, 48 bits virtual power management: ts ttp tm stc 100mhzsteps hwpstate [/code<] pni = Pentium New Instructions = SSE3 Incidentally, yours truly has it that the dual core energy efficient 45nm Athlon II cpus with 2x1 MiB cache were the most efficient and most elegant realizations of the K10 (they repartitioned the 4x512 KiB caches to 2x1MiB), closely followed by Thuban. To wit: the 32nm shrink of the K10 core used in Llano uses 1MB L2 cache per core too for efficiency reasons. Flame away...

          • MadManOriginal
          • 8 years ago

          And aside from that AMD doesn’t improve their drivers in an AMD-CPU-selling vacuum. They have to compete with nVidia so they would still optimize video drivers to give the best performance possible, not just make AMD CPUs look good.

            • Waco
            • 8 years ago

            If they had some magic optimization for Zambezi you can be damn sure they’d put that in there even if it only helped those who have one.

          • sschaem
          • 8 years ago

          SSE3 is not where the beef is at, but ok I should have labeled PhenomII SSE3 capable (even so they cant run SSSE3 code).
          But their is nothing of real interest in SSE4a.. it sound fancy but its just

          EXTRQ/INSERTQ Combined mask-shift instructions.
          MOVNTSD/MOVNTSS Scalar streaming store instructions.

          SSE4.1 and SSE4.2 does offer decent benefit for rendering software that SSE2/3 is lacking.

      • albundy
      • 8 years ago

      there’s still single threaded tasks? update your software dude!

        • BestJinjo
        • 8 years ago

        If you have a program that scales only up to 4 threads on a quad core CPU, guess what, single threaded performance is a factor since each core is running 1 thread at a time. Since 90% of non-content creation programs scale to 2-4 threads at most, a faster quad core processor is going to be superior in the near future (i.e., 2-3 years) than a slow 6-8 core processor. Compare a 2500k @ 4.5ghz vs. the FX-6100 @ 4.5ghz. For most programs, it’s going to come down to the difference in single threaded performance.

        Considering leaked benchmarks put FX-8150 between a 2500k and 2600k in Cinebench (one of THE best multi-threaded apps), this tells us that each core in the module design is actually very weak compared to SB. Otherwise, an FX-8150 would be nearly 2x faster than 2500k in Cinebench.

        Currently, Sandy Bridge has a 40% advantage over Phenom II at the same clocks. Bulldozer has a mountain to climb. Maybe that’s why they are offering users more cores to cater to the crowd that renders, encodes video, does encryption etc. But don’t kid yourself, Photoshop, games, tons of office apps, do not scale beyond 4 cores.

          • MadManOriginal
          • 8 years ago

          The ‘more cores’ is a marketing smokescreen as far as I’m concerned so it bothers me that they are marketing the CPUs with ‘moar cores’ to the less well-informed. Intel could market based on total cores presented to the OS (including hyperthreaded cores) and be doing the same thing, but then all the nerds would throw a tizzy. Why so few people are bothered by AMDs marketing core count is beyond me.

      • ronch
      • 8 years ago

      Oh boy. Not another speculation on performance. Let’s just wait for an official, independent review, ok?

        • flip-mode
        • 8 years ago

        I’ve been waiting for what feels like too long now.

          • ronch
          • 8 years ago

          You, me, and the rest of the frickin’ AMD-lovin’ universe, dude.

      • ermo
      • 8 years ago

      In single threaded tasks, one module can use the resources of both ‘cores’, which should give you an extra execution unit over the Phenom II. Despite the deeper pipeline of the BD design, my guess is that one module at 3.2 GHz (no turbo) will be anywhere from 10-33% faster than one Ph II core at 3.2 GHz (no turbo) in normal integer tasks (no vectorization), depending on how much AMD improved the prefetch logic.

      But [i<]We shall see[/i<] as the Zen Master says...

        • ronch
        • 8 years ago

        Each integer scheduler knows what the other integer scheduler is doing? As far as we know a thread only gets assigned to one scheduler.

          • ermo
          • 8 years ago

          Admittedly, this is my interpretation of the statement
          [quote<]When only one thread is active, it has full access to all shared resources[/quote<] from the BD architecture slides here at TR. Since each module has 4 fetch and decode units, I took that to imply that a module could schedule instructions from a single thread to each core as it saw fit from the shared module front end. But speculation is indeed all it is -- maybe I can't have my cake and eat it too...

    • dpaus
    • 8 years ago

    I can hardly wait to see real benchmarks, because at $380, that could be a very compelling value proposition….

      • chuckula
      • 8 years ago

      [quote<]because at $380, that could be a very compelling value proposition....[/quote<] You know, the next time I hear about how Intel rips you off because the motherboards cost $200, I'll remember this post about what a wonderful value proposition it is to have a $200 AMD motherboard and a chip that is the moral equivalent of a 3 core CPU that costs the same as a 2400.....

        • TheEmrys
        • 8 years ago

        Because dpaus definately speaks for all complainers everywhere….

          • dpaus
          • 8 years ago

          And as part of the agreement, I get all their downvotes! Take [i<]that[/i<], NeelyCam!

        • dpaus
        • 8 years ago

        As to your first point, I agree (although it hasn’t been me saying that), but as to your second, well, let’s wait to see those ‘real benchmarks’ I mentioned, shall we?

        • albundy
        • 8 years ago

        what was intel’s preorder costs where when it’s six core was released?

          • chuckula
          • 8 years ago

          [quote<]what was intel's preorder costs where when it's six core was released?[/quote<] A whole lot higher because, as we'll be seeing in more detail next Wednesday, a 6 core Intel CPU from early 2010 is in a completely different league than a 6 core AMD CPU from late 2011.

        • destroy.all.monsters
        • 8 years ago

        Since when do cpus have a “moral equivalent”? Inquiring minds want to know.

          • chuckula
          • 8 years ago

          [quote<]Since when do cpus have a "moral equivalent"? Inquiring minds want to know.[/quote<] Since Bulldozer took the term "core" and twisted it. The 6 core chips will basically be like a hyperthreaded 3-core Sandy Bridge chip running at a somewhat lower clockspeed. Since such a chip doesn't actually exist, they are the "moral equivalent" of what one would look like.

      • sschaem
      • 8 years ago

      I dont see how 380$ is a deal for a Fx-6100 ?

      $245 for an FX-8150 + $90 for an AM3+ motherboard = $335.. Now thats a deal ,even so its just list price.

        • FuturePastNow
        • 8 years ago

        Indeed. I paid $110 for my last motherboard and that was still too much. $90 sounds about right. Maximize the processor (and RAM and graphics) budget instead.

        I don’t know why someone would buy a $200+ “gamer” motherboard.

          • dpaus
          • 8 years ago

          Well, in my case, this is for ‘white box’ systems for clients – clients who run [i<]lots[/i<] of monitors off each workstation, hence my excitment over all those x16 PCIe slots.

          • flip-mode
          • 8 years ago

          Motherboards have market segmentation like just about any other product. There are millions of products that I think are overpriced that others do not. Every TechReport staff member thinks $100 is a great value for a mechanical keyboard and they all think that discrete sounds cards are worth the money – but I have a distinctly different opinion. It goes the same with motherboards. Why is everyone getting a wedgie over a $200 AMD motherboard? There are $300 and higher Intel motherboards.

            • MadManOriginal
            • 8 years ago

            People are getting a wedgie over calling it a ‘value proposition’. And although at least dpaus explains it later for his particular use, there are surely less epxensive motherobards that have lots of PCIe slots too (or just get video cards that can drive a lot of monitors.)

          • BestJinjo
          • 8 years ago

          1. E-peen
          2. Very high discretionary income, where 2% more performance is worth 100% price increase
          3. Noob PC builders (most noob PC builders think a $200-350 motherboard is better, but they are just paying for features and “marketing” that it overclockers better, when it fact it doesn’t, not on air cooling).
          4. Extreme overclocking (i.e., LN2).
          5. If you want to run 16GB of DDR3-2133 or faster in all 4 DIMMs, then a $350 will do that vs. a $130 one that will struggle. But memory bandwidth doesn’t matter for modern CPUs, so refer to #1.

          For 99% of users, a $130 board such as the Asroch Z68 Extreme3 gives you everything you need (and yes, PCIe 8x/8x is only a 2% performance hit over 16x/16x).

        • travbrad
        • 8 years ago

        This just shows how easy it is to fool people with bundle “deals”.

        It’s like the Comcast triple-play package where you end up paying exactly the same as you would for all 3 services separately (after the “introductory” price).

          • NeelyCam
          • 8 years ago

          And “digital voice” is next to useless. Every single friggin’ bundle I’ve seen from Comcast or Verizon includes digital voice.

          I don’t know [i<]anybody[/i<] who doesn't have a cell phone.

Pin It on Pinterest

Share This