AMD’s answer to the Titan lineup, the Radeon Pro Duo was first revealed last month at AMD’s Capsaicin event. Navigating a line fine line between the Radeon and FirePro lineups, the new graphics cards combines two of AMD’s top end Fiji GPUs. According to VideoCardz, we may see the first Radeon Pro Duos out in the wild sooner than expected. The chip will launch in just a couple of weeks on April 29th.
The Radeon Pro Duo features a pair of 28 nm Fiji GPUs, with two sets of 4,096 stream processors, 256 TMUs, 64 ROPs, and 4 GB of 4096-bit HBM memory. This means a total of 8192 stream processors, 512 TMUs , 128 ROPs and 8GB of HBM1. While the price is a hefty $1499, you do get a very nice custom Cooler Master water cooler with it. Peak performance is a high 16TFLOPS which is still 4.4TFLOPS more than Nvidia’s Tesla P100.
From AMD’s internal benchmarks of 3DMark, the Radeon Pro Duo should smash any other card on the market by a significant margin. Games, however, tend to be more fickle and the Radeon Pro Duo does rely on CrossFire for much of its performance. Given the many issues plaguing SLI and CrossFire this year, it will be interesting to see real world performance once the card becomes available.
One of the first applications that came to mind with HBM was pairing it up with an AMD APU. Proven to work as VRAM with the Fiji GPUs last year, HBM also has possible applications to act as a high-speed cache for other applications where density is important. While we’ve known that AMD has been planning APUs with HBM, the latest report points to Raven Ridge, the 2017 series of APUs that follow Bristol Ridge, to have HBM.
According to the source, Raven Ridge will utilize AMD’s upcoming Zen CPU cores likely paired with Polaris GCN iGPU. With 14nmLPP and Polaris, AMD can stuff a much larger iGPU with their APUs without worrying too much about extra costs or ballooning die size. However, even with the current generation of APUs, the iGPU is bottlenecked at the high-end, something even DDR4 won’t fully solve.
In order to keep growing APU GPU performance, AMD also needs to increase the memory bandwidth. One way, of course, is to use eDRAM as Intel has done with notable success. That, however, is expensive, leading to the top SKUs costing near $400. In comes HBM to the rescue at a relatively lower cost, allowing a large yet budget friendly cache pool to help reduce bandwidth constraints. To produce this, AMD has tapped Amkor, the same firm that worked on Fiji interposers to package Raven Ridge.
With at least, 1GB HBM buffer, the APU will be very well fed, allowing for the iGPU to grow to at least R7 370 performance levels before running out of steam. AMD is also probably working on HMC to supplant HBM in the future as well. If AMD manages to pull this off, Raven Ridge will be the most potent APUs yet, securing the crown against Intel.
In the few days after AMD first demoed Polaris 10 to us at Capsaicin, more details about the upcoming graphics cards have been revealed. Set to be the big brother to the smaller Polaris 11, the better performing chip will drop sometime after June this year.
First off, we’re now able to bring you more information about the settings Hitman was running at during the demo. At Ultra Settings and 1440p, Polaris 10 was able to keep to a constant 60FPS, with VSync being possible. This means the minimum FPS did not drop below 60 at any point. This puts the card at least above the R9 390X and on par if no better than the Fury and Fury X. Of course, the demo was done with DX12 but the boost is only about 10% in Hitman.
Another detail we have uncovered is the maximum length of the engineering sample. Based on the Cooler Master Elite 110 case used, the maximum card length is 210mm or 8.3 inches. In comparison, the Nano is 6 inches and the Fury X 7.64 inches. Given the small size, one can expect Polaris 10 to be as power efficient as Polaris 11 and potentially be using HBM. Given that Vega will be the cards to debut HBM2, Polaris 10 may be limited to 4GB of VRAM. Finally, display connectivity is provided by 3x DP 1.3, 1x HDMI 2.0 and 1 DVI-D Dual Link though OEMs may change this come launch unless AMD locks it down.
Even as Polaris approaches us quickly within 3 months, the planning for its successor has long been in the works. At their Capsaicin event, AMD took off the wraps for their upcoming GPU plans with a roadmap detailing the planned releases up till 2019. In keeping with the star nomenclature that started with Polaris and ditching the islands, we will have Vega and then Navi following Polaris.
Starting off with Polaris later this year, AMD’s main selling point it seems is the 2.5x performance per watt the new GCN architecture will bring. This is no doubt due to the combination of improved hardware itself, the new 14nm LPP process and DX12 finally making use of the previously wasted hardware resources like asynchronous controllers and shaders.
Moving along, we have Vega to release in what looks to be early 2017. The biggest change it seems is the use of HBM2, replacing GDDR5(X) and HBM1 no doubt. This means we can no doubt expected all Vega releases to utilize HBM2. While this may suggest Polaris won’t be using HBM2, it could also mean that only certain Polaris chips, likely only the high-end ones, will use HBM2.
Finally, we come to Navi, which should debut in early 2018. This release will have scalability and use of next-gen memory like Hyper-Memory Cube for instance. The scalability mention suggests either the use of smaller GCN units used to build the chip to better suit the market or a new process node. For now, we are probably better off trying to figure out what Polaris will be
After many months of waiting, AMD has finally unveiled their dual Fiji graphics card. Though not called FuryX2 as we originally expected, the name Radeon Pro Duo is just as fitting. For now, AMD still has not revealed the full specifications but the most important one, price, is a lofty $1,499 USD. For 2 Fiji GPUs and 16TFLOPs of performance, it may well be enough to entice the VR developers AMD is targeting.
As expected of the most powerful single card GPU yet, the power requirements are massive. The 2 GPUs draw power over 3 PCIe 8pin power connectors for up to 525W of power. Memory bandwidth is doubled but the memory remains split as with CFX or dual-GPU cards, with 4GB of HBM1 each over a 4096bit bus. In total, the card has 8192 Shader Cores, 512 TMUs and 128 ROPs with 4DP display connectors. Cooling is provided by a Cooler Master CLC unit with a 120mm extra-thick radiator.
Targetted towards VR and game developers, it makes sense as it offers the performance necessary to the run the most demanding of titles, especially in their unoptimized form. Furthermore, the use of AMD’s affinity multi-GPU, LiquidVR and DX12 will all serve to limit the impact of having 2 separate GPUs. This should allow for better scaling than we usually see with CFX and other solutions.
By targeting VR developers, AMD is able to get away with the hefty price tag, just as Nvidia was able to do the same with their Titan series. This may limit the market though some AMD fans and prosumers won’t mind too much. The price is only $200 more than 2 Fury X’s but it will use up fewer slots and be less of a hassle to arrange the cooling for it. The late launch however, is more of a problematic issue as Polaris and Pascal are fast approaching. It remains to be seen if AMD’s gamble will pay off.
For the upcoming graphics generation, both Nvidia and AMD are set to use HBM2 for their upcoming Pascal and Polaris graphics cards. While Samsung has already revealed that they have kicked off mass production for HBM2, we’re getting word on competitor SK Hynix’s plans for the new memory technology. According to Golem.de, the 4GB modules will start mass production in Q2 while the 8GB variants in Q3.
With HBM2, the maximum capacity per stack jumps from 1GB in HBM1 to 8GB with HBM2. This will allow GPUs to have up to 32GB of VRAM using the 8GB modules in the Fiji/Fury configuration, 16GB using the 4GB modules and 8GB using the lowest 2GB stacks. Each stack also doubles the bandwidth from 128GBps to 256GBps. Due tot he bandwidth increase all coming from a 1Ghz clock rate boost, we can expect latency to be severely reduced.
Given that SK Hynix along with AMD were the pioneers for HBM1, it is surprising that Samsung was the first one to reach mass production for the 4GB modules. On the other hand, SK Hynix may still have locked in the 2GB stacks first, the ones more likely to be used for consumer GPUs. With 4 stacks, this will allow for 8GB of VRAM, plenty for 4K and VR GPUs, while 2 stacks will do fine with 4GB for mainstream cards. Hopefully, this means that HBM2 cards will arrive sooner rather than later.
The initial unveiling of AMD’s Fury X was eagerly anticipated due to the advent of high bandwidth memory, and potential to revolutionize the size to performance ratio of modern graphics cards. This new form of stackable video RAM provided a glimpse into the future and departure from the current GDDR5 standard. Although, this isn’t going to happen overnight as production costs and sourcing HBM on a mass scale has to be taken into consideration. On another note, JEDEC recently announced GDD5X with memory speeds up to 14 Gbps which helps to enhance non-HBM GPUs while catering to the lower-mid range market. The Fury X and Fury utilizes the first iteration of high bandwidth memory which features a maximum capacity of 4GB.
There’s some discussion regarding the effect of this limitation at high resolutions but I personally haven’t seen it cause a noticeable bottleneck. If anything, the Fury range is capable of outperforming the 980 Ti during 4K benchmarks while it tends to linger behind at lower resolutions. AMD’s flagship opts for a closed-loop liquid cooler to reduce temperatures and minimize operating noise. In theory, you can argue this level of cooling prowess was required to tame the GPU’s core. However, there are some air-cooled variants which allow us to directly compare between each form of heat dissipation.
Clearly, the Fury X’s water cooling apparatus adds a premium and isn’t suitable for certain chassis configurations. To be fair, most modern case layouts can accommodate a CLC graphics card without any problems, but there’s also concerns regarding reliability and the possibility of leaks. That’s why air-cooled alternatives which drop the X branding offer great performance at a more enticing price point. For example, the Sapphire Nitro OC R9 Fury is around £60 cheaper than the XFX R9 Fury X. This particular card has a factory overclocked core of 1050MHz, and astounding cooling solution. The question is, how does it compare to the Fury X and GTX 980 Ti? Let’s find out!
Packing and Accessories
The Sapphire Nitro OC R9 Fury comes in a visually appealing box which outlines the Tri-X cooling system, factory overclocked core, and extremely fast memory. I’m really fond of the striking robot front cover and small cut out which provides a sneak peek at the GPU’s colour scheme.
On the opposite side, there’s a detailed description of the R9 Fury range and award-winning Tri-X cooling. Furthermore, the packaging outlines information regarding LiquidVR, FreeSync, and other essential AMD features. This is displayed in an easy-to-read manner and helps inform the buyer about the graphics card’s functionality.
In terms of accessories, Sapphire includes a user’s guide, driver disk, Select Club registration code, and relatively thick HDMI cable.
When AMD launched their Fiji based lineup last year, many were pleased with the performance. The use of HBM helped the Fiji cards helped them achieve better power efficiency while still maintaining the advantages of GCN. The biggest concern at the time was that Nvidia had just cut prices on their GTX 980 and 980Ti, making the Fury and Fury X somewhat disadvantaged. With the launch and holiday season behind us, it looks like AMD is finally deciding to cut prices on the vanilla Fury.
According to the rumours, this price drop is set to happen imminently and meant to better position the Fury against the GTX 980. That card currently retails about 10-20% cheaper than the Fury though the Fury does manage about 10-15% better performance overall. If the price drop comes, the Fury may offer more value relative to the GTX980.
A price drop now does make sense as Polaris is going to arrive in a few months. Cutting the prices to get rid of some inventory will help AMD and their partners be better prepared once Polaris arrives. AMD also recently cut prices on the R9 Nano as well so a cut for the Fury isn’t out of the question. Who knows, maybe the Fury X may its prices slashed as well.
Even as this generation’s GPUs are continuing to fly off the shelves, Nvidia is already gearing up for their Pascal launch. Despite being quieter than AMD, it looks like Nvidia will launch their Pascal cards around the same time, in 2H 2016 as AMD’s Polaris will. What’s more, 2H 2016 will see Nvidia’s flagship Pascal GPU based on TSMC’s 16nmFF+ process and utilizing HBM2. This is still a rumour right now but it does fit the time frame since 1H 2016 would be too soon and 2017 too late.
The biggest question is what does”flagship” mean exactly. Ever since GTX 680 was launched, Nvidia has been playing around with the word flagship. Traditionally, the big dies like GF110 would launch first with the smaller mainstream GF104 launching after. Kepler and Maxwell saw that switch with GK104 and GM104 launching ahead of GK110 and GM200 respectively. This suggests that the so-called “flagship” may only be GP104 and not GP100. Even if it is GP100, it may well be a cut-down version, similar to how the GTX 780 was the cut-down variant of the later GTX 780Ti. This strategy does maximize sales for Nvidia but isn’t that great for consumers.
Whatever the card is, be it GP104 or GP100, it is going to use HBM2, giving it at least 512GB/s with 8GB of VRAM but potentially much higher at 16-32GB with 1TB/s+ of bandwidth. With AMD set to launch Polaris around the same time, Q3 2016 should make for exciting times as a slew of new GPUs arrive.
With both AMD’s Polaris and Nvidia’s Pascal both launching in 2016, the GPU market is very lively this year. Both competing platforms are also set to use HBM2 which JEDEC finally released the final standard for. Within days of JEDEC’s release, Samsung has just started mass production of their HBM2 memory chips.
Starting off with 4GB, Samsung’s HBM2 uses their advanced 20nm process. Each HBM2 package stacks four 8Gb core dies on top of a buffer die at the bottom. Each of these packages will offer 256 GB/s of bandwidth, quadruple that of HBM1. For a more dramatic comparison, each 4Gb GDDR5 die only offers 1/7 the memory bandwidth and 1/2 the energy efficiency. HBM2 also has 5000 TSV (Through Silicon Via), 36x more than what GDDR5 has.
Just like HBM1 did for AMD’s Fury lineup, Samsung is expecting HBM2 to bring 95% space savings compared to GDDR5. In addition to regular HBM2, Samsung is also planning HBM2 with ECC, likely for GPGPU and enterprise work. This means not only will consumer GPUs get numerous benefits, but GPUs found in supercomputers and data centers will soon have HBM2 as well, where it is arguably more useful. Samsung is expecting to continue to ramp up production as demand increases over the year.
Ever since HBM1 was revealed and launched with Fury X, many have been looking forwards to what HBM2 would bring along in 2016. While HBM1 brought large power savings and a major boost in memory bandwidth, it was largely limited to a relatively low 4GB capacity. HBM2, however, is set to provide a boost in capacity and bandwidth by increasing the number of stackable dies. We’re now getting reports that AMD’s upcoming Polaris chips will utilize HBM2.
As a major revamp of the GCN architecture, a Polaris flagship GPU would be the natural product to debut HBM2. A flagship GPU much more powerful than current generation chips due to the new architecture and process node would likely require more memory bandwidth to feed it and a high memory capacity as it would be meant for VR and 4K gaming. Being the largest chip in the lineup, the flagship would also benefit from the major power savings, helping offset its core power consumption. The confirmation of HBM2 also suggests that we will be getting high-end Polaris chips this year.
At the same time, AMD is also confirming that they will continue to use GDDR5 and likely GDDR5X as well. At CES, AMD showed off a low powered Polaris chip using GDDR5 that was able to provide the same performance as Nvidia’s GTX 950 but with a significantly lower power consumption. With such a leap in efficiency, the HBM2 chips will likely be light years ahead of current cards in terms of efficiency if GDDR5 already shows such massive gains.
When AMD launched the Radeon R9 Nano mini-ITX graphics card back in September 2015, the card gained a lot of popularity for several reasons. It was one of the smallest graphics cards available and at the same time delivers a punch that is similar to other high-end cards. The performance is thanks to the HBM memory and Fiji GPU while the small size was made possible by the use of optimal clocking of both the GPU and memory for both power consumption and heat. However, the card did have one fault and that was the price. At a price of $649, the AMD R9 Nano was a hard pill to swallow.
It is now January 2016 and it looks like AMD wants a larger cut of the GPU market. Maybe the manufacturing process has caught up and has been refined for cheaper productions costs, or something else. We don’t know why, but that doesn’t matter anyway as long as the result comes in. And it does and AMD announced a price cut from $649 and all the way down to $499. The $150 price cut equals about 23 percent which is quite nice.
There is no doubt that this price cut on the AMD Radeon R9 Nano will convince quite a few people to get this card that previously wanted it but thought it was too expensive. I am sure there will be quite a few more hardcore mITX gaming builds out there this coming time, and most of them will be powered by this little marvel of a graphics card.
After all, the R9 Nano has a peak power draw of 175W and only measures 6-inch while delivering a level of performance that are on par with larger and more power-hungry GPUs. It also blows the GTX 970 Mini-ITX away with up to 30 percent better performance.
Is this enough to convince you to switch to an AMD Radeon R9 Nano graphics card? $499 is still a respectable price and it might not be in everyone’s budget. Oh, and we should naturally consider that these are the manufacturers suggested retail prices. We have previously seen the cards drop in price at various outlets.
Ever since Llano launched back in 2011, AMD has been pushing their APUs as being the next big thing. Combining a powerful CPU and GPU on one die, the APU allows even budget users to enjoy strong graphics to play games or use it for compute. In 2016, AMD is looking to combine their new Zen CPU architecture with their Polaris based graphics and it looks like something revolutionary will happen. According to a leaked paper, AMD may be planning a multi-core Zen APU coupled with a massive iGPU and an HBM cache with 128GB/s of bandwidth
For AMD, APUs have stayed strong even as their CPU line faltered, with their iGPU beating out Intel solutions. This changed though with Intel’s Iris Pro graphics that came with an eDRAM cache, with the high-speed cache helping put Intel neck and neck with AMD’s top APUs. The only bright side for AMD is that Intel’s chips have cost significantly more than AMD’s offerings. With the launch of a competitive CPU architecture in Zen, a new GPU architecture in Polaris and a large HBM cache, AMD has a chance to surge past Intel in this critical segment.
At a peak of 128GB/s, the HBM cache performs similarly to that of the memory of graphics cards like the GTX 760 and 960. If AMD adds in a high-performance iGPU, the APU has the chance to match or even exceed $150 graphics cards. With this APU, AMD is truly bringing gaming to the masses. Maybe in 2016, we may finally see AMD APUs powering Apple products.
The graphics card market is full of interesting power struggles and if recent reports are true, it seems 2016 will be one of the biggest battles yet. AMD may have already put out some cards with HBM memory, and we’ve heard that Nvidia will be doing the same soon too, but don’t count GDDR memory out just yet! It seems that the upcoming GDDR5 standard is being developed by Micron, which will power mid-end graphics cards, while HBM2 will likely remain for higher end cards.
Of course, there’s some confusion here as JEDEC are already working on the GDDR5X standard, so where Micron fits in really remains to be seen, but that’s something we’ll have to wait and see. GDDR5X is said to double the bandwidth, so is GDDR6 is new standard, or just a further refinement of 5X? Either way, we can expect it to adopt a lower node, most likely starting from 20nm and working down from there, allowing for higher clocks, and lower voltages, although these kinds of improvements are the obvious targets for any increase in performance these days.
Our guess is that the revised GDDR standards will be acting as a bridge until HBM matures enough to cover a wider range of cards and budgets. Either way, 2016 is shaping up to be an exciting time in the GPU market, with new memory, new architectures, new cards and so much more on the horizon.
It looks like AMD are cooking up a big product launch for December, one that could knock Nvidia into the stone age! The as yet un-named graphics card from the red team is currently not officially named, but let’s call this powerhouse the Fury X2 for conversations sake. We’ve already known about this card for a little while now, as Lisa Su was keen to show off a PCB that had not one, but two Fiji GPUs on board. The only problem is when, where and how much it will be, are still unknown.
AMD are to hold a special event in December, where it is believed the Fury X2 will be released. With no rumours pointing to a similar dual-GPU monster from the Nvidia camp, a dual Fiji and HBM equipped card could quickly put AMD back at the top of the GPU power food chain.
The first new teaser comes courtesy of Johan Andersson of DICE, always a reliable source for leaks as he previously teased the 295X2 and the R9 290X. Unfortunately for Johan and no doubt AMD, it seems his “pre-release GPU” had some leaks of its own.
Dat feeling when your just arrived closed liquid cooling pre-release GPU turns out to not be so closed after all pic.twitter.com/umpVUtdVwc
AMD are in a really strong position right now, with a new series of graphics cards that may not have toppled all of the Nvidia flagships, at least not in every test, but it’s certainly put them back on the map in terms of high-end gaming. Their 3xx series of cards are knocking Nvidia down the charts, with better performance and lower prices than their competitor across a wide range of performance brackets, the new Crimson drivers and now it seems the battle is about to heat up even more, as AMD cards are seeing a range of price cuts.
Everything from the 300 series of cards, as well as the new Fury series which feature HBM memory are currently sporting some exciting deals at major retailers. Of course, it won’t be long before Nvidia play a similar price-cut game, but for now, AMD are tempting peoples wallets a little more.
R9 Fury X was originally $649, but that’s down to $589 and even down to $569 after a rebate! The R9 Nano, dropping from $649, down to $569, $549 after rebate. The R9 Fury, $549 down to a very tasty $499.
What about the 300 series? The R9 390X and 390 are down from $429 and $329 to $359 and $259 after a rebate. That’s exceptional value for money giving the performance of these cards! The R9 380 2GB and the 4GB model are down to a wallet friendly $169 and $179, or even lower with rebate at $139 and $159.
Most major retailers are already honouring these new prices, although Newegg and Amazon are currently the best deals I’ve seen, but it’s certainly worth shopping around and who knows, with Black Friday and Cyber Monday, maybe we’ll see even lower prices.
Following on from our highly popular ’12K’ (Triple 4K Monitor) Upgrade, we have new graphics cards which we can update the results with. Since the original article, things got a bit hectic and cards were coming and going extremely quickly. This meant that we didn’t have enough time in one sitting to correctly configure and run the tests as the second (or even third) card needed to be sent on to another media. We are now happy to bring you a long-awaited update featuring graphics cards such as the R9 Fury, R9 Nano and SLI GTX 980Ti’s. The list still isn’t complete with gaps such as SLI Titan X and CrossFire R9 Fury, but once we get these cards in for long enough, we will carry out another update.
With 4K monitors becoming the norm in today’s enthusiast gaming set-up, thanks to the ever decreasing price of these monitors and the increasing performance supplied by single cards; it’s not surprising that some users are combining multiple units. Some will have these monitors for the simply epic screen size and productivity potential, others will simply use them for an upgrade to the current surround gaming experience. Personally I don’t like 4K resolution unless it’s on a large screen, anything under 32″ makes the pixels so small they are hard to see and then you would just have to increase the sizes of font, which defeats (some of) the object.
We all have a lot of chips inside our systems, but companies such as AMD and Nvidia rarely show us what that looks like and we have to rely on third-parties such as ChipWorks to show us what the insides look like. They have now done that with the Fiji chip from the Radeon R9 Fury series and we finally get to see a die shot of it.
A lot of the time this isn’t that interesting as they’re using the same old technology, but this time it is particularly interesting because the Fiji chip is the first to use the HBM JEDEC standard for stacked memory. While the GDDR5 standard has been great for many years, it is getting old and has more or less reached its limit. Granted, a GDDR5x standard has been rumoured for a release next year, but HBM is the future.
HBM offers extremely high memory bandwidth at a considerably lower thermal and power cost than GDDR5, and it was also said to require fewer transistors than GDDR5. That was a thing that we hadn’t been able to verify up until now, but it looks like it was correct. Below we see the Fiji chip next to the Tonga chip and in size relation. While the Fiji HBM bus takes up a little more overall space, it also provides a lot more with its 1024bit over the 64bit GDDR5 bus.
AMD and SK Hynix have spent seven years on the development of the HBM standard and it looks like it’s going to pay off, especially with what we’ve already heard of the HBM2 generation that will take the whole thing to a completely new level.
Here at eTeknix, we strive to give the consumer the best possible advice in every aspect of technology. Today is no different and we are excited to bring you the CrossFireX review of the highly anticipated R9 Nano 4GB graphics cards.
The R9 Nano is the third release in the Fiji GPU core range and the third official graphics card to utilise High Bandwidth Memory (HBMv1). We’ve been impressed with the performance of the Fiji range so far with the fully unlocked R9 Fury X providing a good alternative to the NVIDIA GTX 980Ti, the R9 Fury providing a good step up from the R9 390X and the GTX 980 and the R9 Nano being the perfect option for small form factor builds. A single R9 Nano provides the perfect balance of performance, power consumption and mobility, but will combining two still be a worthwhile option?
When we test in CrossFireX, we aim to use two identical graphics card to ensure that everything is as similar as possible. When using the same cards, you can almost guarantee the same cooling capabilities, power draw, core clock and other variables. This then gives us the best possible outcome for maximum performance as the computer does not need to compensate for any differences.
eTeknix has fought hard over the last few months to be able to bring you the Fiji articles that we have, some may have been a little late, but we have managed to get them out to you one way or another. Stock levels of the Fiji core and HBM have been extremely limited, so AMD had to make the tough decision to only allow an severely limited number of media samples and plumb the rest to the consumer market.
So here it is, our R9 Nano article supplied by Club3D. Right up until launch, we covered a lot about the card and something we knew was almost exactly how the card would look. An R9 Fury X copy with a fan instead of a water cooling solution. From there we took guesses at other specifications, would it feature the Fiji core or a cut-down version like in the R9 Fury with my money on the latter due to the massively decreased size and only single fan; I was extremely surprised when I found out that it uses a full Fiji core as found in the R9 Fury X.
Let’s find out how this miniature monster performs in today’s review.
Packaging and accessories
I’m actually really disheartened by this box. If you are paying £500+ for a graphics card, you’d expect at least a little bit more premium quality to the box. It offers everything you could want in a box, but it just feels cheap.
The back of the box has some key features with some images to be more appealing.
Club3D have really cut down on the accessories with this card, just a simple installation leaflet and a driver disc.
We have previously reported the rumours that Nvidia was planning to use GDDR5X on their upcoming Pascal graphics cards, a rumour that not everyone bought right away. But there are quite a few reasons that we could see this happen and the newest leak seems to support this.
Last time it was a German site that leaked the GDDR5X information and this time we get news from a Russian outlet that got their hands on what looks like leaked slides from Micron of the upcoming GDDR5x memory. The site also suggests that this won’t be limited to only Nvidia, but that AMD also will want to get on board and use this type of memory on some of their graphics cards next year.
HBM, and HBM2 might be very exciting and be the next mainstream graphics memory, but it is still a costly one to produce and the production is also limited. This leaves room for the next GDDR5 standard to make its entry. GDDR5X offers double the data-rate per memory access of 64 byte/access compared to 32 byte/access of the current GDDR5 standard. Where current GDDR5 tops out at around 7Gbps, the new standard will initially offer 10-12 Gbps with a later goal to achieve 16 Gbps.
The implementation of the new memory should be relatively easy and cheap for manufacturers as the new chips will retain the same pin layout. With all this information, we can assume that HBM will be reserved for the top-tier graphics cards for the foreseeable future while GDDR5x will breath more power into mid-level and entry-level cards and allow them to perform better at the ever-increasing resolutions.
While Nvidia’s Pascal is set to launch next year, the graphics company is already setting it’s sights on their next architecture. Codenamed Volta, the Pascal replacement is set to launch sometime in 2018 for the consumer market. For the much more lucrative supercomputer market though, Nvidia is debuting the architecture one year earlier, in 2017. Volta would mark the sixth unified shader architecture to be released by Nvidia.
Volta was originally meant to used HMC or Hybrid Memory Cube as a GDDR5 replacement. That didn’t pan out as HMC suffered delays. This led to Pascal debuting with HBM, with High Bandwidth Memory as a stand in. HMC is finally showing some progress as Micron finally starts rolling with production. Despite this, supply will still short which is probably why Nvidia is choosing to launch first with supercomputers.
Another reason is Nvidia also wants to milk Pascal as much as possible. If Volta were to launch the same year for consumers as supercomputers, Pascal would only have a very short 1-year cycle, making it hard to recoup R&D costs. With Pascal being the scapegoat as well with HBM, Nvidia should be able to learn some lessons for HMC which is relatively more complex. Volta will also utilize the NVLINK GPU interconnect set to debut with Pascal.
Thank you WCCFTech for providing us with this information
Over the next few weeks, both PowerColor and XFX are expected to launch their own R9 Fury GPUs. Based off AMD’s cutdown Fiji die, the Fury initially only launched with cards from ASUS and Sapphire. With AIB partners joining in, it looks like the supply issues behind the Fury may finally have been resolved.
First off, we have PowerColor’s card which has been revealed on their website. At 3584 shaders, 224 TMUs and 64 ROPs, the cut-down Fiji will be clocked at 1000Mhz with the 4GB of HBM untouched. Connectivity features 3 DisplayPort and 1 HDMI output. The card measures 320mm x 125mm x 45mm which is a tad larger than the Sapphire Fury Tri-X overall while only a bit longer than the ASUS Fury Strix. Like the aforementioned cards, the PowerColor features 3 fans which speaks to the level of cooling required for Fury.
Moving on, we have XFX’s implementation which is also based off 3 fans. Interestingly, it looks like the XFX model is essentially the same as the PowerColor one, at least judging from the heatsink and shroud. It may be that the two firms are using the same cooling solution from an OEM. Like the PowerColor, it looks to have 3 DP and 1 HDMI as well.
Rounding off the major partners, we still have no word yet from MSI nor Gigabyte about when their cards might arrive. Given that Sapphire and PowerColor are AMD exclusive partners, it’s not surprising that they are moving in ahead of the last two.
Thank you Videocardz for providing us with this information
It has been rumoured, leaked, and talked about for a while now, the AMD Radeon R9 Nano graphics card, and all those rumours were correct. AMD has just released the Radeon R9 Nano graphics card and it is a beaut.
I could talk a lot about this card right from the start, but let us start with the specifications instead. I’m sure it is those that interest most folks around here, reading this right now. The AMD Radeon R9 Nano comes with a fully configured Fiji GPU based on the 28nm process, with 4096 stream processors, 64 ROPs, 256 Texture Units, and 64 compute units for a compute performance of 8.19TFLOPs. That is also the same chip configuration as the water-cooled Radeon R9 Fury X uses, but I’ll get more into that later. Being a Fiji GPU, it comes with 4GB HBM memory directly on the GPU. The GPU clock can go up to 1000 MHz and the memory is locked at 500MHz/1.0 Gbps on the 4096-bit memory bus for up to 512GB/s bandwidth.
The AMD Radeon R9 Nano isn’t just a very small card, measuring just about 6-inches in length, it’s also very power efficient and that is its key point. As a comparison, a mITX motherboard is 6.7-inch by 6.7-inch and that is already very small. The Nano is even shorter. It only needs a single 8-pin power connector for its typical board power of 175W. With what we know now, it’s easy to say that this is the most powerful mITX card ever created. One of the reasons the card uses so much less power than the R9 290x for example, besides the optimization for performance per watt, is the lower power consumption of the HBM memory. A 4GB GDDR5 equipped card would consume about 50W where the Nano only requires 8w.
The R9 Nano isn’t targeted as a new solo flagship and not intended to perform as good as the Fury X either, as it is. The performance will be around that of the air-cooled R9 Fury, but smaller and more efficient as well as with a lot of headroom for aftermarket cooling solutions. Both the memory and the GPU have been set to the absolute sweet spot where you get maximum performance per watt, and that is this cards strength. While you can overclock the card and get roughly 15% more performance, it will come at a power increase of about 50%. If you want it, you can do it. But whether it’s worth it, is another question.
I’ve previously mentioned that the GPU clock goes up to 1000MHz, but you won’t get that far up under typical situations. It will more likely be around 850-900MHz, again to get the best performance per watt. This can be overwritten in the CCC and set manually, so nothing to worry about if you want to run it to the max.
With such a small card and powerful card, there is worry about the cooling. Both the efficiency, thermal throttling, and noise generation. The GPU is allowed all the way up to 85 degrees before any thermal throttling is happening, and it is designed to run no hotter than 75 degrees in normal setups.
In the first instance, the R9 nano will be released as a reference design only, but it will be opened up to AIB partners later on, allowing them to create their own cooling solutions for the card. Speaking of upgrading it, you can also upgrade the shroud with a custom one like on the Fury X, but it won’t be as easy. AMD promised to release the 3D print files, allowing people to create their own shrouds. But it needs to accommodate for the fan that is mounted on it and isn’t just a square plate like it was on the Fury X.
The Nano comes without any DVI connectors and again, for now it’s reference design only. But this will also be opened up to card partners later on, where a few surely will choose to include the legacy connection. For now you get three DisplayPort connectors and one HDMI. The HDMI port isn’t a 2.0, but you can get that type of connection through a DisplayPort adapter if needed.
Being a mITX design, the AMD R9 Nano is intended to be used mainly in compact and portable gaming rigs, but that doesn’t prevent you from using them in any larger one. The card also features bridgeless CrossfireX support with up to four cards. That ability could create some truly sick mods with all new possibilities in custom case placements.
The three photos below illustrate the cooling solution that is used and that is probably one of the things that many previous AMD customers might be worried about. The Radeon 290x reference card wasn’t exactly what you would call silent.
The first part we see is the dedicated VRM direct touch heat sink, something that probably hasn’t been seen on this type of cards before. A stable and cool VRM goes a long way for a graphics card and having the extra heat sink will allow the fan to spin at a much lower speed.
The second part is the actual cooler that features a dual vapor chamber and heat-pipe thermal solution. This combination again allows the fan to spin at a slower speed as it’s required less. AMD put a lot of work into the cooler, and it shows. Effectively this allows the card to be 16dBA quieter than the R9 290X graphics card and comes in on a noise level of 42 dBA. That is the same noise level you’d find in a library.
The third part of the cooling solution is one that easily could be overlooked and it’s the direction of the fins. Most cards have them turned 90 degrees and that’s simply a bad choice. With this design, most of the hot air will be blown out the rear instead of up onto motherboard and CPU area that is located above the graphics card.
AMD also opted for a matte black PCB on this card next to the full metal shroud and brushed aluminium finish. It sure looks great. It wasn’t entirely sure at the press event what speed the fan will spin on, but it should go up to around 2700 RPM.
So to sum up: The AMD R9 Nano uses the same setup as the R9 Fury X, but it is tuned to a more optimal performance per watt and it’s targeted at compact and mobile gaming rigs. You can overclock it to gain more performance, but it will come at a significant extra power requirement. It is not meant to compete with the Fury X but offer a smaller and more efficient variant of the same. The performance, as it is, will be around that of the air-cooled R9 Fury.
What’s left to say, oh yea the pricing. The card won’t be cheap and it will cost about the same as the Fury X with a $649 USD MSRP. I also got a fun fact, the Amazing AMD micro system displayed at the same time as the Fury X was announced actually used two R9 Nano cards in a dual-board layout. So different PCB, but same GPU configuration.
Are you tempted to get one of the new AMD Radeon R9 Nano cards? I know I am, or perhaps two while I’m at it. Here is also a comparison on what is possible. These two systems have the same CPU, the same amount of memory, same graphics power, and same storage abilities. Which would you pick?
Later today, AMD is expected to finally launch their much-anticipated R9 Nano GPU. Based off of the full Fiji die, the small form factor card will be a sure hit with the HTPC crowd. Having already had some of their slides with benchmarks and images of the card leak, we’re now getting word on the pricing situation. According a report, AMD is setting the price pretty high, with the Nano set to launch with an MSRP of $649 USD. Keep in mind that this is still unconfirmed right now and AMD may yet launch the card at a different price.
To put that in perspective, the R9 Fury, with a slightly higher clocked but gimped Fiji comes in at $549 while the full Fiji flagship R9 Fury X costs about $649. This puts the R9 Nano in a tough spot despite having a full Fiji. Even with similar top clocks as the Fury X, the Nano is still limited by its cooling, 42dB sound rating and its 75C temperature target. This means the Nano will probably perform closer to the vanilla Fury. Against the green side, the Nano does offer 30% more performance than the competing SFF GTX 970, but will set you back double the price.
With those numbers in mind, it really seems that AMD wants to maintain their margins on the Fiji die and HBM. The serious binning they are doing to get such efficient chips also means the Nano is a rarity. The biggest question is whether or not users are willing shell out top cash to get the fastest and most efficient mini-ITX card on the market or will they simply settle for something else. At this price though, there is little chance the Nano will cannibalize the Fury’s so AMD has at least got that covered.
Thank you WCCFTech for providing us with this information
With AMD virtually confirmed to launch their SFF R9 Nano tomorrow, we’re getting word that the Fiji GPU onboard won’t be cut down. Unlike the R9 Fury, the Nano will be like it’s older R9 Fury X sibling and feature the complete Fiji die. This means that the Nano will have the same 4096 shader cores, 256 TMUs, 64 ROPs and 4GB HBM as the full fledged Fury X flagship. The Nano may also feature HDMI 2.0 which will allow 4K 60hz for TVs, something the Furys lacks and great for a card that is perfect for high-end HTPC gaming
What is even better news for SFF fans is that the Nano will feature a top speed of 1000Mhz on the core, giving it the potential to nearly match the Fury X. If the card manages to somehow not throttle (ie under water probably), the performance should be pretty much on par with the Fury X, in a much more compact form factor. The card also features the same display setup as it’s Fury siblings, all in a row which can allow for a single expansion card slot under a watercooling. Stock cooling probably is a combination vapor chamber and several heatpipes though we’ll know more once the card arrives.
The biggest question though is how the stock power limits, the 8pin connector, and the heat sink will impact the card. While the 8pin connector in tandem with the PCIe slot should allow plenty of power, the stock power limits may serve to limit the speeds you’ll get at full throttle. Another issue is whether or not the cooling system can keep up with a full Fiji, keeping in mind that AMD went with watercooling for the Fury X which also features full Fiji at similar speeds. Some sources are saying that at full load the card will usually throttle to about 800mhz with the peak speed only being seen in some lighter loads. Noise levels will also be interesting to see. AMD did run into some serious problems with the stock cooling for the R9 290X, leading to heavy throttling. Hopefully, the lesson has been learned.
With a full Fiji core, AMD can’t afford to sell the card too low but it also somehow has to fit in with the Fury siblings. Given that performance should hover around that of the Fury, AMD will have to price the card carefully to ensure that it does sell, but also preventing it from cannibalizing the rest of the lineup too much. With just a day to go, we hope to bring your more information as it arrives.
Thank you Videocardz for providing us with this information
While SKHynix was the first to produce HBM for AMD, things may change next year. Samsung is leveraging their massive production base and getting into the HBM market with plans to start mass production in 2016. Given that Samsung is planning to provide 4Hi and 8Hi stacks, Samsung is skipping right past HBM1 and straight to the more useful HBM2. Samsung is marketing the use of HBM as a fast cache like eDRAM for both normal consumers but also targeted towards high-performance computing as well.
Having more suppliers for HBM is obviously a good thing for AMD as it ensures that SkHynix won’t become a bottleneck both for current HBM cards but also for their upcoming Greenland flagship in the future as well. This is also bad news as AMD reportedly has a priority agreement to get HBM from SKHynix. It was inferred at the time this meant that AMD could effectively lock out Nvidia from HBM for a while till production ramped up. With Samsung leveraging their massive fabs, it seems that Nvidia will probably be able to get all the HBM they want as well.
With HBM being in plentiful supply, Nvidia probably won’t face any issues on this front for their new Pascal GPUs. It’s important to note that while Samsung does plan to start mass production in 2016, when that ramp up will take place and how long that will take is still unknown. This means that AMD might still have an advantage, albeit now diminished. With widely available HBM2, a new process node and new architectures, the GPU landscape for 2016 looks very promising.
Thank you ComputerBase for providing us with this information