Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Sunday October 15 2017, @07:07PM   Printer-friendly
from the watch-for-bugs-at-the-chip-buffet dept.

High Performance Computing (HPC) Chips – A Veritable Smorgasbord?

No this isn't about the song from Charlotte's Web or the Scandinavian predilection for open sandwiches; it's about the apparent newfound choice in the HPC CPU market.

For the first time since AMD's ill-fated launch of Bulldozer the answer to the question, 'Which CPU will be in my next HPC system?' doesn't have to be 'Whichever variety of Intel Xeon E5 they are selling when we procure'.

In fact, it's not just in the x86 market where there is now a genuine choice. Soon we will have at least two credible ARM v8 ISA CPUs (from Cavium and Qualcomm respectively) and IBM have gone all in on the Power architecture (having at one point in the last ten years had four competing HPC CPU lines – x86, Blue Gene, Power and Cell).

In fact, it may even be Intel that is left wondering which horse to back in the HPC CPU race with both Xeon lines looking insufficiently differentiated going forward. A symptom of this dilemma is the recent restructuring of the Xeon line along with associated pricing and feature segmentation.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 1, Interesting) by Anonymous Coward on Sunday October 15 2017, @10:07PM (5 children)

    by Anonymous Coward on Sunday October 15 2017, @10:07PM (#582775)

    Oracle's approach to Sun's technology seem to be "Fuck it".
    Their willingness to allow those to go out of support (surreptitiously or otherwise) seems very high.

    ...and Oracle is the epitome of closed technology.
    This is a company stuck in the late 20th Century.
    Why people would continue to give them money confounds me.

    Don't developers write their code in a high level language today, making the hardware brand immaterial as long as there is a compatible compiler available? [googleusercontent.com] (orig) [gnu.org]

    ...and on the Oracle software side, has anyone ever had a database that was so complex that FOSS couldn't handle it?

    -- OriginalOwner_ [soylentnews.org]

    Starting Score:    0  points
    Moderation   +1  
       Interesting=1, Total=1
    Extra 'Interesting' Modifier   0  

    Total Score:   1  
  • (Score: 2, Informative) by Ethanol-fueled on Sunday October 15 2017, @10:26PM (4 children)

    by Ethanol-fueled (2792) on Sunday October 15 2017, @10:26PM (#582780) Homepage

    "Fuck it" regarding Sun's physical technology, sure, but muh Java.

    Regarding your "high-level language" comment, there are plenty of situations where people use inline assembly within their C and C++ code. And that assembly is dependent on the machine.

    • (Score: 1, Informative) by Anonymous Coward on Monday October 16 2017, @02:55AM (3 children)

      by Anonymous Coward on Monday October 16 2017, @02:55AM (#582893)

      For over a decade, I've heard people intimate something like "If you can do assember that is better-optimized than what the compiler can deliver, you're in the 1 Percent of coders and have way too much time on your hands."

      Comments?

      -- OriginalOwner_ [soylentnews.org]

      • (Score: 2) by takyon on Monday October 16 2017, @05:23AM

        by takyon (881) <reversethis-{gro ... s} {ta} {noykat}> on Monday October 16 2017, @05:23AM (#582913) Journal

        Yeah, here's a comment. We should ditch all programming languages except JavaScript and Rust.

        --
        [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
      • (Score: 3, Informative) by coolgopher on Monday October 16 2017, @07:09AM

        by coolgopher (1157) on Monday October 16 2017, @07:09AM (#582921)

        Or you're working with more obscure hardware with lower quality compilers (that might not even be aware of the full instruction set).

        Or you're working on seriously performance critical code.

        Or you're working on sufficiently low-level stuff where you need register-level control anyway.

        Or you're in the demo scene and write seriously tight stuff for the joy of taking the hardware beyond what anyone thought possible.

        To mention a few.

      • (Score: 1) by Ethanol-fueled on Monday October 16 2017, @10:24PM

        by Ethanol-fueled (2792) on Monday October 16 2017, @10:24PM (#583192) Homepage

        Even if you don't directly use assembly within your C/C++ code when doing embedded stuff you still have the embedded mindset, such as using the keywords volatile and register.

        Volatile marks the portion of code as something which could change outside the code, such as by the hardware, so you don't want the compiler to optimize the code away if it is going to assume that the code is redundant and that there is no code which could change the value of whatever's in the register.

        Register throws the variable into a register to keep it on-hand (close to the processor, you can think of a register as being a kind of RAM that's directly on the processor metal) for super-fast operations. But optimizing compilers "know" when to throw certain variables into registers anyway, and knowing the optimization options available and how to use them is an art in itself.

        Embedded code is fucking everywhere -- dildos, realdolls, IoT devices, weapons and other robotics (3-D printers are a good example), calculators, wristwatches, etc. You don't see much embedded work written in Javascript or Ruby.