Subject: Editorial, Storage | June 17, 2014 - 09:56 AM | Allyn Malventano
Tagged: sandisk, fusion-io, buyout
Fusion-io was once a behemoth of flash memory storage. Back when SSDs were having a hard time saturating SATA 3Gb/sec, Fusion-io was making fire breathing PCIe SSDs full of SLC flash and pushing relatively insane IOPS and throughput figures. Their innovations were a good formula at the time. They made the controller a very simple device, basically just a simple bridge from the PCIe bus to the flash memory. This meant that most of the actual work was done in the driver. This meant that Fusion-io SSDs were able to leverage the CPU and memory of the host system to achieve very high performance.
Fusion-io ioDrive 160 creams the competition back in 2010.
Being the king of IOPS back in the early days of flash memory storage, Fusion-io was able to charge a premium for their products. In a 2010 review, I priced their 160GB SSD at about $40/GB. In the years since, while flash memory prices (and therefore SSD products) have steadily dropped in price while achieving higher and higher performance figures, Fusion-io products have mostly remained static in price. All of this time, the various iterations of the ioDrive continued to bank on the original model of a simple controller and the bulk of the work taking place in the driver. This actually carries a few distinct disadvantages, in that the host system has to spent a relatively large amount of CPU and memory resources towards handling the Fusion-io devices. While this enables higher performance, it leaves less resources available to actually do stuff with the data. This ends up adding to the build cost of a system, as more CPU cores and memory must be thrown at the chassis handling the storage. In more demanding cases, additional systems would need to be added to the rack space in order to handle the additional storage overhead in addition to the other required workloads. Lastly, the hefty driver means Fusion-io devices are not bootable, despite early promises to the contrary. This isn't necessarily a deal breaker for enterprise use, but it does require system builders to add an additional storage device (from a different vendor) to handle OS duties.
In 2014, the other guys are making faster stuff. Note this chart is 4x the scale of the 2010 chart.
Lets fast forward to present times. Just over a week ago, Fusion-io announced their new 'Atomic' line of SSDs. The announcement seemed to fall flat, and did little to save the continuous decline of their stock price. I suspect this was because despite new leadership, these new products are just another iteration of the same resource consuming formula. Another reason for the luke warm reception might have been the fact that Intel launched their P3700 series a few days prior. The P3700 is a native PCIe SSD that employs the new NVM Express communication standard. This open standard was developed specifically for flash memory communication, and it allows more direct access to flash in a manner that significantly reduces the overhead required to perform high data throughputs and very high IO's per second. NVMe is a very small driver stack with native support built into modern operating systems, and is basically the polar opposite of the model Fusion-io has relied on for years now.
Intel's use of NVMe enables very efficient access to flash memory with minimal CPU overhead.
Fusion-io's announcement claimed "The Atomic Series of ioMemory delivers the highest transaction rate per gigabyte for everything from read intensive workflows to mixed workloads.". Let's see how this stacks up against the Intel P3700 - an SSD that launched the same week:
|Model||Fusion-io PX600||Intel P3700|
|Interface / Flash type||PCIe 2.0 x8 / 20nm MLC||PCIe 3.0 x4 / 20nm MLC|
|Read BW (GB/sec)||2.7||2.7||2.7||2.7||2.7||2.8||2.8||2.8|
|Write BW (GB/sec)||1.5||1.7||2.2||2.1||1.2||1.9||1.9||1.9|
|4k random read IOPS||196,000||235,000||330,000||276,000||450,000||460,000||450,000||450,000|
|4k random write IOPS||320,000||370,000||375,000||375,000||75,000||90,000||150,000||175,000|
|4k 70/30 R/W IOPS||Unlisted||150,000||200,000||240,000||250,000|
|Endurance / TB||12.0||12.3||12.3||12.3||18.3||18.3||18.3||18.3|
|Warranty||5 years||5 years|
We are comparing flagship to flagship (in a given form factor) here. Starting from the top, the Intel P3700 is available in generally smaller capacities than the Fusion-io PX600. Both use 20nm flash, but the P3700 uses half the data lanes at twice the throughput. Regarding Fusion-io's 'transaction rate per GB' point, well, it's mostly debunked by the Intel P3700, which has excellent random read performance all the way down to its smallest 400GB capacity point. The seemingly unreal write specs seen from the PX600 are, well, actually unreal. Flash memory writes take longer than reads, so the only logical explanation for the inversion we see here is that Fusion-io's driver is passing those random writes through RAM first. Writing to RAM might be quicker, but you can't sustain it indefinitely, and it consumes more host system resources in the process. Moving further down the chart, we see Intel coming in with a ~50% higher endurance rating when compared to the Fusion-io. The warranties may be of equal duration, but the Intel drive is (on paper / stated warranty) guaranteed to outlast the Fusion-io part when used in a heavy write environment.
For pricing, Intel launched the P3700 at a competitive $3/GB. Pricing data for Fusion-io is not available, as they are behind a bit of a 'quote wall', and no pricing at all was included with the Atomic product launch press materials. Let's take a conservative guess and assume the new line is half the cost/GB of their previous long-standing flagship, the Octal. One vendor lists pricing directly at $124,995 for 10.24TB ($12.21/GB) and $99,995 for 5.12TB ($19.53/GB), both of which require minumum support contracts as an additional cost. Half of $12/GB is still more than twice the $3/GB figure from Intel.
My theory as to why SanDisk is going for Fusion-io?
- A poor track record since the Fusion-io IPO have driven the stock price way down, making it prime for a buyout.
- SanDisk is one of the few remaining flash memory companies that does not own their own high end controller tech.
- Recent Fusion-io product launch overshadowed by much larger (Intel) company launching a competing superior product at a lower cost/GB.
So yeah, the buyout seemed inevitable. The question that remains is what will SanDisk do with them once they've bought them? Merging the two will mean that Fusion-io can include 'in house' flash and (hopefully) offer their products at a lower cost/GB, but that can only succeed if the SanDisk flash performs adequately. Assuming it does, there's still the issue of relatively high costs when compared to freshly competing products from Intel and others. Last but not least is the ioDrive driver model, which grows incresingly dated while the rest of the industry adopts NVMe.
Subject: Editorial, General Tech, Graphics Cards, Processors, Chipsets | June 13, 2014 - 06:45 PM | Scott Michaud
Tagged: x86, restructure, gpu, arm, APU, amd
According to VR-Zone, AMD has reworked their business, last Thursday, sorting each of their projects into two divisions and moving some executives around. The company is now segmented into the "Enterprise, Embedded, and Semi-Custom Business Group", and the "Computing and Graphics Business Group". The company used to be divided between "Computing Solutions", which handled CPUs, APUs, chipsets, and so forth, "Graphics and Visual Solutions", which is best known for GPUs but also contains console royalties, and "All Other", which was... everything else.
Lisa Su, former general manger of global business, has moved up to Chief Operating Officer (COO), along with other changes.
This restructure makes sense for a couple of reasons. First, it pairs some unprofitable ventures with other, highly profitable ones. AMD's graphics division has been steadily adding profitability to the company while its CPU division has been mostly losing money. Secondly, "All Other" is about a nebulous as a name can get. Instead of having three unbalanced divisions, one of which makes no sense to someone glancing at AMD's quarterly earnings reports, they should now have two, roughly equal segments.
At the very least, it should look better to an uninformed investor. Someone who does not know the company might look at the sheet and assume that, if AMD divested from everything except graphics, that the company would be profitable. If, you know, they did not know that console contracts came into their graphics division because their compute division had x86 APUs, and so forth. This setup is now more aligned to customers, not products.
Subject: Editorial | June 12, 2014 - 02:28 PM | Ken Addison
Tagged: Z97X-SOC Force, video, titan z, radeon, project tango, podcast, plextor, nvidia, Lightning, gtx titan z, gigabyte, geforce, E3 14, amd, 4790k, 290x
PC Perspective Podcast #304 - 06/12/2014
We have lots of reviews to talk about this week including the GeForce GTX TITAN Z, Core i7-4790K, Gigabyte Z97X-SOC Force, E3 News and more!
The URL for the podcast is: http://pcper.com/podcast - Share with your friends!
- iTunes - Subscribe to the podcast directly through the Store
- RSS - Subscribe through your regular RSS reader
- MP3 - Direct download link to the MP3 file
Hosts: Ryan Shrout, Josh Walrath, Jeremy Hellstrom and Allyn Maleventano
Pimp Next Week Events
0:03:45 Podcast #305 with David Hewlett!
Week in Review:
News items of interest:
Hardware/Software Picks of the Week:
0:57:40 Ryan: Ken's Switching Hardware
1:02:15 Josh: This saved me last weekend.
1:04:00 Allyn: Sony DSC-RX10
Subject: Editorial | June 4, 2014 - 07:42 PM | Ryan Shrout
Tagged: video, pcper, live, austin evans
Tonight's live edition of the PC Perspective Podcast is going to have a special guest, the Internet's Austin Evans. You likely know of Austin through his wildly popular YouTube channel or maybe his dance moves.
But seriously, Austin Evans is a great guy with a lot of interesting input on technology. Stop by our live page at http://www.pcper.com/live at 10pm EST / 7pm EST for all the fun!
Make sure you don't miss it by signing up for our PC Perspective Live Mailing List!
Subject: Editorial, General Tech | May 29, 2014 - 02:17 AM | Scott Michaud
It should not pass anyone's smell test but it apparently does, according to tweets and other articles. Officially, the TrueCrypt website (which redirects to their SourceForge page) claims that, with the end of Windows XP support (??), the TrueCrypt development team wants users to stop using their software. Instead, they suggest a switch to BitLocker, Mac OSX built-in encryption, or whatever random encryption suite comes up when you search your Linux distro's package manager (!?). Not only that, but several versions of Windows (such as 7 Home Premium) do not have access to BitLocker. Lastly, none of these are a good solution for users who want a single encrypted container across multiple OSes.
A new version (don't use it!!!) called TrueCrypt 7.2 was released and signed with their private encryption key.
The developers have not denied the end of support, and its full-of-crap reason. (Seriously, because Microsoft deprecated Windows XP almost two months ago, they pull support for a two year old version now?)
They have also not confirmed it. They have been missing since at least "the announcement" (or earlier if they were not the ones who made it). Going missing and unreachable, the day of your supposedly gigantic resignation announcement, does not support the validity of that announcement.
To me, that is about as unconfirmed as you can get.
Still, people are believing the claims that TrueCrypt 7.1a is not secure. The version has been around since February 2012 and, beyond people looking at its source code, has passed a significant portion of a third-party audit. Even if you believe the website, it only says that TrueCrypt will not be updated for security. It does not say that TrueCrypt 7.1a is vulnerable to any known attack.
In other words, the version that has been good enough for over two years, and several known cases of government agencies being unable to penetrate it, is probably as secure today as it was last week.
"The final version", TrueCrypt 7.2, is a decrypt-only solution. It allows users to unencrypt existing vaults, although who knows what else it does, to move it to another solution. The source code changes have been published, and they do not seem shady so far, but since we cannot even verify that their private key has not leaked, I wouldn't trust it. A very deep compromise could make finding vulnerabilities very difficult.
So what is going on? Who knows. One possibility is that they were targeted for a very coordinated hack, one which completely owned them and their private key, performed by someone(s) who spent a significant amount of time modifying a fake 7.2 version. Another possibility is that they were legally gagged and forced to shut down operations, but they managed to negotiate a method for users to decrypt existing data with a neutered build.
One thing is for sure, if this is a GoG-style publicity stunt, I will flip a couple of tables.
We'll see. ┻━┻ \_(ツ)_/ ┻━┻
The AMD Argument
Earlier this week, a story was posted in a Forbes.com blog that dove into the idea of NVIDIA GameWorks and how it was doing a disservice not just on the latest Ubisoft title Watch_Dogs but on PC gamers in general. Using quotes from AMD directly, the author claims that NVIDIA is actively engaging in methods to prevent game developers from optimizing games for AMD graphics hardware. This is an incredibly bold statement and one that I hope AMD is not making lightly. Here is a quote from the story:
Gameworks represents a clear and present threat to gamers by deliberately crippling performance on AMD products (40% of the market) to widen the margin in favor of NVIDIA products. . . . Participation in the Gameworks program often precludes the developer from accepting AMD suggestions that would improve performance directly in the game code—the most desirable form of optimization.
The example cited on the Forbes story is the recently released Watch_Dogs title, which appears to show favoritism towards NVIDIA GPUs with performance of the GTX 770 ($369) coming close the performance of a Radeon R9 290X ($549).
It's evident that Watch Dogs is optimized for Nvidia hardware but it's staggering just how un-optimized it is on AMD hardware.
Watch_Dogs is the latest GameWorks title released this week.
I decided to get in touch with AMD directly to see exactly what stance the company was attempting to take with these kinds of claims. No surprise, AMD was just as forward with me as they appeared to be in the Forbes story originally.
The AMD Stance
Central to AMD’s latest annoyance with the competition is the NVIDIA GameWorks program. First unveiled last October during a press event in Montreal, GameWorks combines several NVIDIA built engine functions into libraries that can be utilized and accessed by game developers to build advanced features into games. NVIDIA’s website claims that GameWorks is “easy to integrate into games” while also including tutorials and tools to help quickly generate content with the software set. Included in the GameWorks suite are tools like VisualFX which offers rendering solutions like HBAO+, TXAA, Depth of Field, FaceWorks, HairWorks and more. Physics tools include the obvious like PhysX while also adding clothing, destruction, particles and more.
Subject: Editorial, General Tech | May 14, 2014 - 09:56 PM | Scott Michaud
Tagged: ultraviolet, mozilla, DRM, Adobe Access, Adobe
Needless to say, DRM is a controversial topic and I am clearly against it. I do not blame Mozilla. The non-profit organization responsible for Firefox knew that they could not oppose Chrome, IE, and Safari while being a consumer software provider. I do not even blame Apple, Google, and Microsoft for their decisions, either. This problem is much bigger and it comes down to a total misunderstanding of basic mathematics (albeit at a ridiculously abstract and applied level).
Simply put, piracy figures are meaningless. They are a measure of how many people use content without paying (assuming they are even accurate). You know what is more useful? Sales figures. Piracy figures are measurements, dependent variables, and so is revenue. Measurements cannot influence other measurements. Specifically, measurements cannot influence anything because they are, themselves, the result of influences. That is what "a measure" is.
Implementing DRM is not a measurement, however. It is a controllable action whose influence can be recorded. If you implement DRM and your sales go down, it hurt you. You may notice piracy figures decline. However, you should be too busy to care because you should be spending your time trying to undo the damage you did to your sales! Why are you looking at piracy figures when you're bleeding money?
I have yet to see a DRM implementation that correlated with an increase in sales. I have, however, seen some which correlate to a massive decrease.
The thing is, Netflix might know that and I am pretty sure that some of the web browser companies know that. They do not necessarily want to implement DRM. What they want is content and, surprise, the people who are in charge of the content are definitely not enlightened to that logic. I am not even sure if they realize that the reason why content is pirated before their release dates is because they are not leaked by end users.
But whatever. Technical companies, who want that content available on their products, are stuck finding a way to appease those content companies in a way that damages their users and shrinks their potential market the least. For Mozilla, this means keeping as much open as possible.
Since they do not have existing relationships with Hollywood, Adobe Access will be the actual method of displaying the video. They are clear to note that this only applies to video. They believe their existing relationships in text, images, and games will prevent the disease from spreading. This is basically a plug-in architecture with a sandbox that is open source and as strict as possible.
This sandbox is intended to prevent a security vulnerability from having access to the host system, give a method of controlling the DRM's performance if it hitches, and not allow the DRM to query the machine for authentication. The last part is something they wanted to highlight, because it shows their effort to protect the privacy of their users. They also imply a method for users to opt-out but did not go into specifics.
As an aside, Adobe will support their Access DRM software on Windows, Mac, and Linux. Mozilla is pushing hard for Android and Firefox OS, too. According to Adobe, Access DRM is certified for use with Ultraviolet content.
I accept Mozilla's decision to join everyone else but I am sad that it came to this. I can think of only two reasons for including DRM: for legal (felony) "protection" under the DMCA or to make content companies feel better while they slowly sink their own ships chasing after numbers which have nothing to do with profits or revenue.
Ultimately, though, they made a compromise. That is always how we stumble and fall down slippery slopes. I am disappointed but I cannot suggest a better option.
Subject: Editorial, General Tech | May 5, 2014 - 08:08 PM | Scott Michaud
Tagged: mozilla, net neutrality
Recently, the FCC has been moving to give up Net Neutrality. Mozilla, being dedicated to the free (as in speech) and open internet, has offered a simple compromise. Their proposal is that the FCC classifies internet service providers (ISPs) as common carriers on the server side, forcing restrictions on them to prevent discrimination of traffic to customers, while allowing them to be "information services" to consumers.
In other words, force ISPs to allow services to have unrestricted access to consumers, without flipping unnecessary tables with content distribution (TV, etc.) services. Like all possibilities so far, it could have some consequences, however.
"Net Neutrality" is a hot issue lately. Simply put, the internet gives society an affordable method of sharing information. How much is "just information" is catching numerous industries off guard, including ones which Internet Service Providers (ISPs) participate in (such as TV and Movie distribution), and that leads to serious tensions.
On the one hand, these companies want to protect their existing business models. They want consumers to continue to select their cable and satellite TV packages, on-demand videos, and other services at controlled profit margins and without the stress and uncertainty of competing.
On the other hand, if the world changes, they want to be the winner in that new reality. Yikes.
A... bad... photograph of Mozilla's "UP" anti-datamining proposal.
Mozilla's proposal is very typical of them. They tend to propose compromises which divides an issue such that both sides get the majority of their needs. Another good example is "UP", or User Personalization, which tries to cut down on data mining by giving a method for the browser to tell websites what they actually want to know (and let the user tell the browser how much to tell them). The user would compromise, giving the amount of information they find acceptable, so the website would compromise and take only what they need (rather than developing methods to grab anything and everything they can). It feels like a similar thing is happening here. This proposal gives users what they want, freedom to choose services without restriction, without tossing ISPs into "Title II" common carrier altogether.
Of course, this probably comes with a few caveats...
The first issue that pops in my mind is, "What is a service?". I see this causing problems for peer-to-peer applications (including BitTorrent Sync and Crashplan, excluding Crashplan Central). Neither endpoint would necessarily be classified as "a server", or at least convince a non-technical lawmaker that is the case, and thus ISPs would not need to apply common carrier restrictions to them. This could be a serious issue for WebRTC. Even worse, companies like Google and Netflix would have no incentive to help fight those battles -- they're legally protected. It would have to be defined, very clearly, what makes "a server".
Every method will get messy for someone. Still, the discussion is being made.
Subject: Editorial, General Tech, Graphics Cards | April 30, 2014 - 10:05 AM | Ryan Shrout
Tagged: hadron air, hadron, gtx 750, giveaway, evga, contest
Congrats to our winner: Pierce H.! Check back soon for more contests and giveaways at PC Perspective!!
In these good old United States of America, April 15th is a trying day. Circled on most of our calendars is the final deadline for paying up your bounty to Uncle Sam so we can continue to have things like freeway systems and universal Internet access.
But EVGA is here for us! Courtesy of our long time sponsor you can win a post-Tax Day prize pack that includes both an EVGA Hadron Air mini-ITX chassis (reviewed by us here) as well as an EVGA GeForce GTX 750 graphics card.
Nothing makes paying taxes better than free stuff that falls under the gift limit...
With these components under your belt you are well down the road to PC gaming bliss, upgrading your existing PC or starting a new one in a form factor you might not have otherwise imagined.
Competing for these prizes is simple and open to anyone in the world, even if you don't suffer the same April 15th fear that we do. (I'm sure you have your own worries...)
- Fill out the form at the bottom of this post to give us your name and email address, in addition to the reasons you love April 15th! (Seriously, we need some good ideas for next year to keep our heads up!) Also, this does not mean you should leave a standard comment on the post to enter, though you are welcome to do that too.
- Stop by our Facebook page and give us a LIKE (I hate saying that), head over to our Twitter page and follow @pcper and heck, why not check our our many videos and subscribe to our YouTube channel?
- Why not do the same for EVGA's Facebook and Twitter accounts?
- Wait patiently for April 30th when we will draw and update this news post with the winners name and tax documentation! (Okay, probably not that last part.)
A huge thanks goes out to friends and supporters at EVGA for providing us with the hardware to hand out to you all. If it weren't for sponsors like this PC Perspective just couldn't happen, so be sure to give them some thanks when you see them around the In-tar-webs!!
Subject: Editorial | April 23, 2014 - 09:51 PM | Josh Walrath
Tagged: TDP, Athlon 5350, Asus AM1I-A, amd, AM1
If I had one regret about my AM1 review that posted a few weeks ago, it was that I used a pretty hefty (relatively speaking) 500 watt power supply for a part that is listed at a 25 watt TDP. Power supplies really do not hit their efficiency numbers until they are at least under 50% load. Even the most efficient 500 watt power supply is going to inflate the consumption numbers of these diminutive parts that we are currently testing.
Keep it simple... keep it efficient.
Ryan had sent along a 60 watt notebook power supply with an ATX cable adapter at around the same time as I started testing on the AMD Athlon 5350 and Asus AM1I-A. I was somewhat roped into running that previously mentioned 500 watt power supply due to comparative reasons. I was using a 100 watt TDP A10-6790 APU with a pretty loaded Gigabyte A88X based ITX motherboard. That combination would have likely fried the 60 watt (12v x 5A) notebook power supply under load.
Now that I had a little extra time on my hands, I was able to finally get around to seeing exactly how efficient this little number could get. I swapped the old WD Green 1 TB drive for a new Samsung 840 EVO 500 GB SSD. I removed the BD-ROM drive completely from the equation as well. Neither of those parts uses a lot of wattage, but I am pushing this combination to go as low as I possibly can.
The results are pretty interesting. At idle we see the 60 watt supply (sans spinning drive and BD-ROM) hitting 12 watts as measured from the wall. The 500 watt power supply and those extra pieces added another 11 watts of draw. At load we see a somewhat similar numbers, but not nearly as dramatic as at idle. The 60 watt system is drawing 29 watts while the 500 watt system is at 37 watts.
So how do you get from a 60 watt notebook power adapter to ATX standard? This is the brains behind the operation.
The numbers for both power supplies are both good, but we do see that we get a nice jump in efficiency from using the smaller unit and a SSD instead of a spinning drive. Either way, the Athlon 5350 and AMD AM1 infrastructure sip power as compared to most desktop processors.