Gunning for Broadwell-E
As I walked away from the St. Regis in downtown San Francisco tonight, I found myself wandering through the streets towards my hotel with something unique in tow. It was a smile. I was smiling, thinking about what AMD had just demonstrated and showed at its latest Zen processor reveal. The importance of this product launch can literally not be overstated for a company struggling to find a foothold to hang on to in a market that it once had a definitive lead. It’s been many years since I left a conference call, or a meeting, or a press conference feeling genuinely hopefully and enthusiastic about what AMD has shown me. Tonight I had that.
AMD’s CEO Lisa Su, and CTO Mark Papermaster, took stage down the street from the Intel Developer Forum to roll out a handful of new architectural details about the Zen architecture while also showing the first performance results comparing it to competing parts from Intel. The crowd in attendance, a mix of media and analysts, were impressed. The feeling was palpable in the room.
It’s late as I write this, and while there are some interesting architecture details to discuss, I think it is in everyone’s best interest that we touch on them lightly for now, and instead refocus on the deep-dive once the Hot Chips information comes out early next week. What you really want to know is clear: can Zen make Intel work again? Can Zen make that $1700 price tag on the Broadwell-E 6950X seem even more ludicrous? Yes.
The Zen Architecture
Much of what was discussed from the Zen architecture is a re-release of what has been out in recent months. This is a completely new, from the ground up, microarchitecture and not a revamp of the aging Bulldozer design. It integrated SMT (simultaneous multi-threading), a first for an AMD CPU, to better take efficient advantage of a longer pipeline. Intel has had HyperThreading for a long time now and AMD is finally joining the fold. A high bandwidth and low latency caching system is used to “feed the beast” as Papermaster put it and utilizing 14nm process technology (starting at Global Foundries) gives efficiency, and scaling a significant bump while enabling AMD to scale from notebooks to desktops to servers with the same architecture.
By far the most impressive claim from AMD thus far was that of a 40% increase in IPC over previous AMD designs. That’s a HUGE claim and is key to the success or failure of Zen. AMD proved to me today that the claims are real and that we will see the immediate impact of that architecture bump from day one.
Press was told of a handful of high level changes to the new architecture as well. Branch prediction gets a complete overhaul. This marks the first AMD processor to have a micro-op cache. Wider execution width with broader instruction schedulers are integrated, all of which adds up to much higher instruction level parallelism to improve single threaded performance.
Performance improvements aside, throughput and efficiency go up with Zen as well. AMD has integrated an 8MB L3 cache and improved prefetching for up 5x the cache bandwidth available per core on the CPU. SMT makes sure the pipeline stays full to prevent “bubbles” that introduce latency and lower efficiency while region-specific power gating means that we’ll see Zen in notebooks as well as enterprise servers in 2017. It truly is an impressive design from AMD.
Summit Ridge, the enthusiast platform that will be the first product available with Zen, is based on the AM4 platform and processors will go up to 8-cores and 16-threads. DDR4 memory support is included, PCI Express 3.0 and what AMD calls “next-gen” IO – I would expect a quick leap forward for AMD to catch up on things like NVMe and Thunderbolt.
The Real Deal – Zen Performance
As part of today’s reveal, AMD is showing the first true comparison between Zen and Intel processors. Sure, AMD showed a Zen-powered system running the upcoming Deus Ex running at 4K with a system powered by the Fury X, but the really impressive results where shown when comparing Zen to a Broadwell-E platform.
Using Blender to measure the performance of a rendering workload (a Zen CPU mockup of course), AMD ran an 8-core / 16-thread Zen processor at 3.0 GHz against an 8-core / 16-thread Broadwell-E processor at 3.0 GHz (likely a fixed clocked Core i7-6900K). The point of the demonstration was to showcase the IPC improvements of Zen and it worked: the render completed on the Zen platform a second or two faster than it did on the Intel Broadwell-E system.
Not much to look at, but Zen on the left, Broadwell-E on the right...
Of course there are lots of caveats: we didn’t setup the systems, I don’t know for sure that GPUs weren’t involved, we don’t know the final clocks of the Zen processors releasing in early 2017, etc. But I took two things away from the demonstration that are very important.
- The IPC of Zen is on-par or better than Broadwell.
- Zen will scale higher than 3.0 GHz in 8-core configurations.
AMD obviously didn’t state what specific SKUs were going to launch with the Zen architecture, what clock speeds they would run at, or even what TDPs they were targeting. Instead we were left with a vague but understandable remark of “comparable TDPs to Broadwell-E”.
Pricing? Overclocking? We’ll just have to wait a bit longer for that kind of information.
There is clearly a lot more for AMD to share about Zen but the announcement and showcase made this week with the early prototype products have solidified for me the capability and promise of this new microarchitecture. We have asked for, and needed, as an industry, a competitor to Intel in the enthusiast CPU space – something we haven’t legitimately had since the Athlon X2 days. Zen is what we have been pining over, what gamers and consumers have needed.
AMD’s processor stars might finally be aligning for a product that combines performance, efficiency and scalability at the right time. I’m ready for it –are you?
It always feels a little odd when covering NVIDIA’s quarterly earnings due to how they present their financial calendar. No, we are not reporting from the future. Yes, it can be confusing when comparing results and getting your dates mixed up. Regardless of the date before the earnings, NVIDIA did exceptionally well in a quarter that is typically the second weakest after Q1.
NVIDIA reported revenue of $1.43 billion. This is a jump from an already strong Q1 where they took in $1.30 billion. Compare this to the $1.027 billion of its competitor AMD who also provides CPUs as well as GPUs. NVIDIA sold a lot of GPUs as well as other products. Their primary money makers were the consumer space GPUs and the professional and compute markets where they have a virtual stranglehold on at the moment. The company’s GAAP net income is a very respectable $253 million.
The release of the latest Pascal based GPUs were the primary mover for the gains for this latest quarter. AMD has had a hard time competing with NVIDIA for marketshare. The older Maxwell based chips performed well against the entire line of AMD offerings and typically did so with better power and heat characteristics. Even though the GTX 970 was somewhat limited in its memory configuration as compared to the AMD products (3.5 GB + .5 GB vs. a full 4 GB implementation) it was a top seller in its class. The same could be said for the products up and down the stack.
Pascal was released at the end of May, but the company had been shipping chips to its partners as well as creating the “Founder’s Edition” models to its exacting specifications. These were strong sellers throughout the end of May until the end of the quarter. NVIDIA recently unveiled their latest Pascal based Quadro cards, but we do not know how much of an impact those have had on this quarter. NVIDIA has also been shipping, in very limited quantities, the Tesla P100 based units to select customers and outfits.
Subject: Graphics Cards | August 12, 2016 - 09:44 PM | Jeremy Hellstrom
Tagged: rx 470, LatencyMon, dpc, amd
When The Tech Report first conducted their review of the RX 470 they saw benchmark behaviour very different from any other GPU in that family but could not figure out what it was and resolve it before the mob arrived with pitchforks and torches demanding they publish or die.
As it turns out there was indeed something rotten in benchmark; incredibly high DPC on the test machine. Investigation determined the culprit to be the beta BIOS on their ASRock Z170 Extreme7+, specifically the BIOS which allowed you to overclock locked Intel CPUs. They have just released their new findings along with a look at LatencyMon and DPC in general. Take a look at the new benchmarks and information about DPC, but also absorb the consequences of demanding articles arrive picoseconds after the NDA expires; if there is a delay in publishing there might just be a damn good reason why.
"We retested our RX 470 to account for this issue, and we also updated our review with DirectX 12 benchmarks for Rise of the Tomb Raider and Hitman, plus full OpenGL and Vulkan benchmarks for Doom."
Here are some more Graphics Card articles from around the web:
- AMD & NVIDIA GPU VR Performance in Trials on Tatooine @ [H]ard|OCP
- AMD's Radeon RX 460 @ The Tech Report
- 18-Way GPU Linux Benchmarks, Including The Radeon RX 460 & RX 470 On Open-Source @ Phoronix
- ASUS Radeon RX 460 STRIX OC 4 GB @ techPowerUp
- MSI RX 470 Gaming X 8G @ Kiguru
- MSI GTX 1060 6GB Gaming X @ Kitguru
- MSI GeForce GTX 1070 Gaming Z @ Modders-Inc
- Nvidia Titan X (Pascal) Extended Overclock Guide @ Guru of 3D
- Nvidia Titan X @ Kitguru
- MSI GeForce GTX 1080 Gaming Z 8G Review @HiTech Legion
- Zotac GTX 1080 AMP! Edition 8 GB @ techPowerUp
Subject: Graphics Cards | August 10, 2016 - 08:59 PM | Scott Michaud
Tagged: amd, graphics drivers
Alongside the release of the Radeon RX 460 and RX 470 graphics cards, AMD has released the Radeon Software Crimson Edition 16.8.1 drivers. Beyond adding support for these new products, it also adds a Crossfire profile for F1 2016 and fixes a few issues, like Firefox and Overwatch crashing under certain circumstances. It also allows users of the RX 480 to overclock their memory higher than they previously could.
AMD is continuing their trend of steadily releasing graphics drivers, and rapidly fixing important issues as they arise. Also, they have been verbose in their release notes, outlining fixes and known problems as they occur. Users can often track the bugs that affect them as they are added to the Known Issues, then graduated to Fixed Issues. While this often goes unrecognized, it's frustrating as a user to experience a bug and not know whether the company even knows about it, or they are just refusing to acknowledge it.
Useful release notes, like AMD has been publishing, are very helpful in that regard.
Subject: General Tech | August 10, 2016 - 06:13 PM | Jeremy Hellstrom
Tagged: gaming, starseed, VR, amd, nvidia, htc vive
When [H]ard|OCP looks at the performance of a VR game, be it a Vive or Rift title, they focus on the gameplay experience as opposed to benchmarks. There are numerous reasons for this, from the fact that these games do not tend to stress GPUs like many triple A titles but also because the targets are different, steady render times below 11.1ms are the target as opposed to higher frame counts. AMD initially had issues with this game, the newest driver release has resolved those issues completely. The takeaway quote in [H]'s conclusions provide the most telling part of the review, "If we were to perform a blind gaming test, you would not be able to identify which GPU you were gaming with at the time."
"We are back this week to take another objective look at AMD and NVIDIA GPU performance in one of the the top selling games in the VR-only realm, The Gallery Episode 1: Call of Starseed. This is another GPU-intensive title that has the ability to put some GPUs on their heels. How do the new RX 480 and GeForce 1000 series perform?"
Here is some more Tech News from around the web:
- Battlefield 1 weapons of war detailed in video trailer @ HEXUS
- No Man’s Sky Launch Update: Exploits Removed, Sea Beds Souped-Up, Sunsets Intensified… @ Rock, Paper, SHOTGUN
- No Man’s Sky isn’t the game I expected: thoughts on the first 10 hours @ Polygon
- Deus Ex: Mankind Divided PC to support Tobii Eye Tracking @ HEXUS
- Sudden Strike 4 Is A Slower More Thoughtful RTS @ Rock, Paper, SHOTGUN
- Survive This Bundle @ Humble Bundle
- Dead Rising Being Remastered And Coming To PC @ Rock, Paper, SHOTGUN
Subject: Graphics Cards | August 8, 2016 - 09:25 PM | Jeremy Hellstrom
Tagged: htc vive, amd, nvidia, raw data
Raw Data is an early access game for the HTC Vive, one which requires space to move and which allows the Vive to show off its tracking ability. [H]ard|OCP wanted to see how the GPUs found in most high end systems would perform in this VR game and so grabbed several AMD and NVIDIA cards to test out. Benchmarking VR games is not an easy task, instead of raw performance you need to focus on the dropped frames and unstable fps which result in nausea and a less engrossing VR experience. To that end [H] has played the game numerous times on a variety of GPUs with settings changing throughout to determine the sweet spot for the GPU you are running. VR offers a new gaming experience and new tests need to be developed to demonstrate performance to those interested in jumping into the new market. Check out the full review to see what you think of their methodology as well as the raw performance of the cards.
"Both AMD and NVIDIA have had a lot to say about "VR" for a while now. VR is far from mainstream, but we are now seeing some games that are tremendously compelling to play, putting you in middle of the action. Raw Data is one of those, and it is extremely GPU intensive. How do the newest GPUs stack up in Raw Data?"
Here are some more Graphics Card articles from around the web:
- PowerColor Red Devil Radeon RX 470 @ [H]ard|OCP
- Radeon RX 470 @ The Tech Report
- PowerColor Radeon RX 470 Red Devil Review @HiTech Legion
- Sapphire Nitro+ RX 470 OC @ eTeknix
- The AMD Radeon RX 470 4GB Review @ Hardware Canucks
- AMD RX 470 @ Hardware Heaven
- Sapphire RX 470 Nitro + OC 4GB @ Kitguru
- Asus RX 470 Strix Gaming OC Aura RGB 4GB @ Kitguru
- ASUS Radeon RX 470 STRIX OC 4 GB @ techPowerUp
Subject: Graphics Cards | August 8, 2016 - 05:08 PM | Sebastian Peak
Tagged: amd, radeon, RX460, rx 460, graphics, gpu, gaming, benchmark, 1080p, 1920x1080, gtx 950, gtx 750 ti
HEXUS has posted their review of Sapphire's AMD Radeon RX 460 Nitro 4GB graphics card, pitting it against the NVIDIA GTX 950 and GTX 750 Ti in a 1920x1080 benchmarking battle.
Image credit: HEXUS
"Unlike the two previous AMD GPUs released under the Polaris branding recently, RX 460 is very much a mainstream part that's aimed at buyers who are taking their first real steps into PC gaming. RX 460 uses a distinct, smaller die and is to be priced from £99. As usual, let's fire up the comparison specification table and dissect the latest offering from AMD."
Image credit: HEXUS
The results might surprise you, and vary somewhat based on the game selected. Check out the source link for the full review over at HEXUS.
Subject: Graphics Cards | August 6, 2016 - 06:24 PM | Tim Verry
Tagged: sapphire, rx 470, polaris 10, dual x, amd
Following the official launch of AMD's Radeon RX 470 GPU, Sapphire has unleashed its own custom graphics card with the Nitro+ RX 470 in 4GB and 8GB factory overclocked versions. Surprisingly, the new cards are up for purchase now at various retailers at $210 for the 4GB model and $240 for the 8GB model (more on that in a bit).
The new Nitro+ RX 470 uses the same board and cooler design as the previously announced Nitro+ RX 480 which is a good thing both for Sapphire (less R&D cost) and for consumers as they get a rather beefy cooler that should allow them to push the RX 470 clocks quite a bit. The card uses the same Dual X cooler with two 95mm quick connect fans, three nickel plated copper heatpipes, and an aluminum fin stack. The card features the same black fan shroud and black and silver colored backplate. Out of the box this cooler should keep the RX 470 GPU running cooler and quieter than the RX 480, but it should also enable users to get higher clocks out of the smaller GPU (less cores means less heat and more overclocking headroom assuming you get a good chip from the silicon lottery).
Sapphire is using Black Diamond 4 chokes and a 4+1 power phase design that is driven by a single 8-pin PCI-E power connector (and up to 75W from the motherboard slot). This mirrors the design of its RX 480 sibling.
Display outputs include a single DVI, two HDMI 2.0b, and two DisplayPort 1.4 ports.
The chart below outlines the comparison between the Nitro+ RX 470 cards, RX 470 reference specifications, and the RX 480.
Nitro+ RX 470 4GB
|Nitro+ RX 470 8GB||RX 470 Reference||RX 480|
|GPU Clock (Base)||1143 MHz||1121 MHz||926 MHz||1120 MHz|
|GPU Clock (Boost)||1260 MHz||1260 MHz||1206 MHz||1266 MHz|
|Memory||4GB GDDR5 @ 7 GHz||8GB GDDR5 @ 8 GHz||4 or 8 GB GDDR5 @ 6.6 GHz||4 or 8 GB GDDR5 @ up to 8 GHz|
|Memory Bandwidth||224 GB/s||256 GB/s||211 GB/s||256 GB/s|
|GPU||Polaris 10||Polaris 10||Polaris 10||Polaris 10|
|Price||$210||$240||$180+||$200+ ($240+ for 8GB)|
The RX 470 GPU is only slightly cut down from RX 480 in that it features four fewer CUs though the processor maintains the same number of ROP units and the same 256-bit memory bus. Reference clocks are 926 MHz base and 1206 MHz boost. Memory can be up to 8GB of GDDR5 with reference memory clocks of 6.6 GHz (effective). Sapphire has overclocked both the GPU and memory with the NItro+ series. The Nitro+ RX 470 with 4GB of GDDR5 is clocked at 1143 MHz base, 1260 MHz boost, and 7 GHz memory while the 8GB version has a lower base clock of 1121 but a higher memory clock of 8 GHz.
The 8GB model having a lower base overclock is a bit strange to me, but at least they are rated at the same boost clock. These specifications are very close to the RX 480 actually and with a bit of user overclocking beyond the factory overclock you could get even closer to the performance of it.
The problem with this RX 470 that gets so close to the RX 480 though is that the price is also very close to reference RX 480s! The Sapphire Nitro+ RX 470 4GB is priced at $209.99 while the Nitro+ RX 470 8GB is $239.99.
These prices put the card well into RX 480 territory though not quite up to the MSRPs of factory overclocked RX 480s (e.g. Sapphire's own Nitro+ RX 480 is $219 and $269 for 4GB and 8GB respectively). The company has a nice looking (and hopefully performing) RX 470, but it is going to be tough to choose this card over a RX 480 that has more shaders and TMUs. One advantage though is that this is a card that will just work without having to manually overclock (though where is the fun in that? heh) and it is actually available right now unlike the slew of RX 480 cards that have been launched but are consistently out of stock everywhere! If you simply can't wait for a RX 480, this might not be a bad option.
EDIT: Of course the 8GB model goes out of stock at Newegg as I write this and Amazon's prices are higher than MSRP! hah.
Subject: Graphics Cards | August 2, 2016 - 07:50 AM | Tim Verry
Tagged: sapphire, rx 460, polaris 11, nitro, amd
AMD and its board partners will officially launch the first Polaris 11 GPU and the Radeon RX 460 graphics cards based around that processor on August 8th. Fortunately Videocardz.com got a hold of an image that shows off Sapphire's take on the RX 460 in the form of a factory overclocked and custom cooled RX460 Nitro OC. This gives us a hint at the kinds of cards we can expect and it appears to be good news for budget gamers as it suggests that there will be several options around this firm $100 price point that are a bit more than the bare necessities.
In the case of Sapphire's RX 460 Nitro OC, it uses a custom dual fan cooler with two copper heatpipes, an aluminum fin stack (that is much larger than reference), and two 90mm fans. Display IO includes one DVI, one HDMI, and one DisplayPort. The card itself uses a physical PCI-E x16 connector that is electrically PCI-E 3.0 x8. The x8 connection will be more than enough for this GPU though it also enables partners to cut costs.
Clockspeeds are not yet known, but the Polaris 11 GPU (896 cores, 56 TMUs, 16 ROPs) will be paired with 4GB GDDR5 memory.
It is encouraging to me to see custom cards at this price point out of the gate with the full 4GB of memory (AMD allows 2GB or 4GB versions). Gamers that simply can't justify spending much more than a hundred dollars on a GPU should have ample options to choose from and I am looking forward to seeing what all the partners have to offer.
Are you looking at Polaris 11 and the RX 460 for a super budget gaming build? What do you think about Sapphire's card with the company's custom cooler?
Subject: Graphics Cards | August 1, 2016 - 02:16 PM | Sebastian Peak
Tagged: amd, radeon, radeon software, Crimson Edition 16.7.3, driver, graphics, update, rx480, rise of the tomb raider
AMD has released the Radeon Software Crimson Edition 16.7.3 driver, with improved performance in Rise of the Tomb Raider for Radeon RX 480 owners, as well as various bug fixes.
Radeon Software Crimson Edition is AMD's revolutionary new graphics software that delivers redesigned functionality, supercharged graphics performance, remarkable new features, and innovation that redefines the overall user experience. Every Radeon Software release strives to deliver new features, better performance and stability improvements.
Radeon Software Crimson Edition 16.7.3 Highlights
Rise of the Tomb Raider performance increase up to 10% versus Radeon Software Crimson Edition 16.7.2 on Radeon RX 480 graphics