A 1400 MB/s SSD: ASRock's Z97 Extreme6 And Samsung's XP941

Features
By
published

Z97 ushers in new and exciting ways to attach and use storage devices. With support for M.2 PCIe and SATA Express, two sides of the same SSD coin, Z97 improves on Z87. But not everywhere. AsRock add to Z97 with some new tricks, and so we take a look.

Results: A PCIe SSD's Random Performance

Random performance is another issue entirely. We already know that throughput ceilings aren't as much of a concern when it comes to moving around lots of small chunks of data. In fact SATA 6Gb/s is typically sufficient for heavy random workloads. 

The Samsung XP941 employs AHCI, which has some inherent overhead that chokes the potential of solid-state storage. NVMe was designed to address this. However, Intel's NVMe driver isn't expected until the end of 2014. As a result, we have to accept that a PCIe-based SSD utilizing AHCI is probably going to demonstrate modest advantages, at best. How does the XP941 stack up in the two different connectors exposed by ASRock's Z97 Extreme6?

Random Read Performance

Given what we saw on the previous page, it'd be easy to assume that Samsung's XP941 is capable of massive transactional performance working with small random transfers. The truth is a matter of relativity.

Yes, 120,000 IOPS is an impressive result. But that number doesn't reflect the potential of Samsung's hardware the same way sequential transfers do. And yes, the four-lane Ultra M.2 slot does yield better results. However, the scaling isn't there to indicate that a two-lane interface attached to the PCH was really hamstrung, either.

Even more telling, the two- and four-lane interfaces track alongside the SATA 6Gb/s-based 840 Pro up until a queue depth of 16. Desktop workloads typically don't see that much concurrency, so the XP941 wouldn't confer much benefit.

Random Write Performance

This is mostly what I would have expected based on our previous work with PCIe-based SSDs utilizing AHCI. The SATA-attached 840 Pro takes top honors, even if it isn't the fastest SSD around. Samsung's XP941 falls flat connected to the two-lane M.2 slot. It fares better when we hook up with the four-lane Ultra M.2 interface, though not in any way that'd lead us to favor such a configuration over familiar SATA. 

Bottom line: the random performance of a PCIe-based SSD is more pedestrian than the impressive sequential scores, largely due to AHCI. Still, if you stopped here and didn't look at any other benchmark, you might conclude that Samsung's XP941 is the greatest desktop-oriented SSD ever. But our testing in Iometer isn't necessarily indicative of how the drive behaves in the real world. We need to go into more depth.

Christopher Ryan
20 Comments Comment from the forums
  • aminebouhafs
    Once an SSD in plugged into the Ultra M.2 slot, the bandwidth between central processing unit and graphics processing unit is cut-down by half. Therefore, while the end-user gets additional SSD performance, the end-user may lose some GPU performance because of insufficient bandwidth between it and the CPU.
    Reply
  • JoeArchitect
    Very interesting article and a great read. Thanks, Chris - I hope to see more like this soon!
    Reply
  • wussupi83
    great article! - although z97 still seems boring
    Reply
  • Eggz
    This makes me excited for X99! With 40 (or more) lanes, of PCI-e (probably more), there will be no need to compromise. We have to remember that the Z97 Chipset is a consumer-grade product, so there almost has to be tradoffs in order to justify stepping up to a high-end platform.

    That said, I feel like X99, NVMe, and and M.2 products will coincide nicely with their respective releases dates. Another interesting piece to the puzzle will be DDR4. Will the new storage technology and next-generation CPUs utilize it's speed, or like DD3, will it take several generations for other technologies to catch up to RAM speeds? This is quite an interesting time :)
    Reply
  • Amdlova
    Chris test the asrock z97 itx... and another thing... my last 3 motherboard from asrock and i want to say Asrock Rock's!
    Reply
  • Damn_Rookie
    While storage isn't the most important area of computer hardware for me, I always enjoy reading Christopher's articles. Very well written, detail orientated, and above all else, interesting. Thanks!
    Reply
  • hotwire_downunder
    ASRock has come along way, I used them a long time back with disappointing results, but I have started to use them again and have not been disappointed this time around.

    Way to turn things around ASRock! Cheap as chips and rock steady!
    Reply
  • alidan
    @aminebouhafs if i remember right, didn't toms show how much performance loss there is when you tape gpu cards to emulate having half or even a quarter of the bandwidth? if i remember right back than the difference was only about 12% from 16 lanes down to either 4 or 8
    Reply
  • Eggz
    13445787 said:
    @aminebouhafs if i remember right, didn't toms show how much performance loss there is when you tape gpu cards to emulate having half or even a quarter of the bandwidth? if i remember right back than the difference was only about 12% from 16 lanes down to either 4 or 8

    PCI-e 3.0 x8 has enough bandwidth for any single card. The only downside to using PCI-e lanes on the SSD applies only to people who want to use multiple GPUs.

    Still, though, this is just the mid-range platform anyway. People looking for lots of expansion end up buying the X chipsets rather than the Z chipsets because of the greater expandability. I feel like the complaint is really misplaced for Z chipsets, since they only have 16 PCI-e lanes to begin with.
    Reply
  • cryan
    Once an SSD in plugged into the Ultra M.2 slot, the bandwidth between central processing unit and graphics processing unit is cut-down by half. Therefore, while the end-user gets additional SSD performance, the end-user may lose some GPU performance because of insufficient bandwidth between it and the CPU.

    Well, it'll definitely negate some GPU configurations, same as any PCIe add-in over the CPU's lanes. With so few lanes to work with on Intel's mainstream platforms, butting heads is inevitable.

    Regards,
    Christopher Ryan


    Reply