A new architecture with GP104
Table of Contents
- Asynchronous compute discussion
- Is only 2-Way SLI supported?
- Overclocking over 2.0 GHz
- Dissecting the Founders Edition
- Benchmarks begin
- VR Testing
- Impressive power efficiency
- Performance per dollar discussion
- Ansel screenshot tool
The summer of change for GPUs has begun with today’s review of the GeForce GTX 1080. NVIDIA has endured leaks, speculation and criticism for months now, with enthusiasts calling out NVIDIA for not including HBM technology or for not having asynchronous compute capability. Last week NVIDIA’s CEO Jen-Hsun Huang went on stage and officially announced the GTX 1080 and GTX 1070 graphics cards with a healthy amount of information about their supposed performance and price points. Issues around cost and what exactly a Founders Edition is aside, the event was well received and clearly showed a performance and efficiency improvement that we were not expecting.
The question is, does the actual product live up to the hype? Can NVIDIA overcome some users’ negative view of the Founders Edition to create a product message that will get the wide range of PC gamers looking for an upgrade path an option they’ll take?
I’ll let you know through the course of this review, but what I can tell you definitively is that the GeForce GTX 1080 clearly sits alone at the top of the GPU world.
NVIDIA's Ansel Technology
“In-game photography” is an interesting concept. Not too long ago, it was difficult to just capture the user's direct experience with a title. Print screen could only hold a single screenshot at a time, which allowed Steam and FRAPS to provide a better user experience. FRAPS also made video more accessible to the end-user, but it output huge files and, while it wasn't too expensive, it needed to be purchased online, which was a big issue ten-or-so years ago.
Seeing that their audience would enjoy video captures, NVIDIA introduced ShadowPlay a couple of years ago. The feature allowed users to, not only record video, but also capture the last few minutes. It did this with hardware acceleration, and it did this for free (for compatible GPUs). While I don't use ShadowPlay, preferring the control of OBS, it's a good example of how NVIDIA wants to support their users. They see these features as a value-add, which draw people to their hardware.
Subject: General Tech, Graphics Cards | May 16, 2016 - 03:19 PM | Ryan Shrout
Tagged: video, tom petersen, pascal, nvidia, live, GTX 1080, gtx, GP104, geforce
Our review of the GeForce GTX 1080 is LIVE NOW, so be sure you check that out before today's live stream!!
Get yourself ready, it’s time for another GeForce GTX live stream hosted by PC Perspective’s Ryan Shrout and NVIDIA’s Tom Petersen. The general details about consumer Pascal and the GeForce GTX 1080 graphics card are already official and based on the traffic to our stories and the response on Twitter and YouTube, there is more than a little pent-up excitement. .
On hand to talk about the new graphics card, answer questions about technologies in the GeForce family including Pascal, SLI, VR, Simultaneous Multi-Projection and more will be Tom Petersen, well known in our community. We have done quite a few awesome live steams with Tom in the past, check them out if you haven't already.
NVIDIA GeForce GTX 1080 Live Stream
10am PT / 1pm ET - May 17th
Need a reminder? Join our live mailing list!
The event will take place Tuesday, May 17th at 1pm ET / 10am PT at http://www.pcper.com/live. There you’ll be able to catch the live video stream as well as use our chat room to interact with the audience, asking questions for me and Tom to answer live.
Tom has a history of being both informative and entertaining and these live streaming events are always full of fun and technical information that you can get literally nowhere else. Previous streams have produced news as well – including statements on support for Adaptive Sync, release dates for displays and first-ever demos of triple display G-Sync functionality. You never know what’s going to happen or what will be said!
UPDATE! UPDATE! UPDATE! This just in fellow gamers: Tom is going to be providing two GeForce GTX 1080 graphics cards to give away during the live stream! We won't be able to ship them until availability hits at the end of May, but two lucky viewers of the live stream will be able to get their paws on the fastest graphics card we have ever tested!! Make sure you are scheduled to be here on May 17th at 10am PT / 1pm ET!!
Don't you want to win me??!?
If you have questions, please leave them in the comments below and we'll look through them just before the start of the live stream. Of course you'll be able to tweet us questions @pcper and we'll be keeping an eye on the IRC chat as well for more inquiries. What do you want to know and hear from Tom or I?
So join us! Set your calendar for this coming Tuesday at 1pm ET / 10am PT and be here at PC Perspective to catch it. If you are a forgetful type of person, sign up for the PC Perspective Live mailing list that we use exclusively to notify users of upcoming live streaming events including these types of specials and our regular live podcast. I promise, no spam will be had!
Subject: Graphics Cards | May 10, 2016 - 07:29 PM | Ryan Shrout
Tagged: video, pascal, nvidia, GTX 1080, gtx 1070, geforce
After the live streamed event announcing the GeForce GTX 1080 and GTX 1070, Allyn and I spent a few minutes this afternoon going over the information as it was provided, discussing our excitement about the product and coming to grips with what in the world a "Founder's Edition" even is.
If you haven't yet done so, check out Scott's summary post on the GTX 1080 and GTX 1070 specs right here.
Subject: Graphics Cards | May 10, 2016 - 04:06 PM | Sebastian Peak
Tagged: video card, reference cooler, pascal, nvidia, GTX 1080, graphics card, GeForce GTX 1080, Founder's Edition
The first non-reference GTX 1080 has been revealed courtesy of Galax, and the images (via VideoCardz.com) look a lot different than the Founder's Edition.
Galax GTX 1080 (Image Credit: VideoCardz)
The Galax is the first custom implementation of the GTX 1080 we've seen, and as such the first example of a $599 variant of the GTX 1080. The Founder's Edition cards carry a $100 premium (and offer that really nice industrial design) but ultimately it's about performance and the Galax card will presumably offer completely stock specifications.
(Image Credit: VideoCardz)
Expect to see a deluge of aftermarket cooling from EVGA, ASUS, MSI, and others soon enough - most of which will presumably be using a dual or triple-fan cooler, and not a simple blower like this.
Subject: Graphics Cards | May 9, 2016 - 05:00 PM | Scott Michaud
Tagged: pascal, nvidia, GTX 1080, geforce
During the GeForce GTX 1080 launch event, NVIDIA announced two prices for the card. The new GPU has an MSRP of $599 USD, while a Founders Edition will be available for $699 USD. They did not really elaborate on the difference at the keynote, but they apparently clarified the product structure for the attending press.
According to GamersNexus, the “Founders Edition” is NVIDIA's new branding for their reference design, which has been updated with the GeForce GTX 1080. That is it. Normally, a reference design is pretty much bottom-tier in a product stack. Apart from AMD's water-cooling experiments, reference designs are relatively simple, single-fan blower coolers. NVIDIA's reference cooler though, at least on their top-three-or-so models of any given generation, are pretty good. They are fairly quiet, effective, and aesthetically pleasing. When searching for a specific GPU online, you will often see a half-dozen entries based on this, from various AIB partners, and another half-dozen other offerings from those same companies, which is very different. MSI does their Twin Frozr thing, while ASUS has their Direct CU and Poseidon coolers.
If you want the $599 model, then, counter to what we've been conditioned to expect, you will not be buying NVIDIA's reference cooler. These will come from AIB partners, which means that NVIDIA is (at least somewhat) allowing them to set a minimum product this time around. They expect reference cards to be intrinsically valuable, not just purchased because they rank highest on a “sort by lowest price” metric.
This is interesting for a number of reasons. It wasn't too long ago that NVIDIA finally allowed AIB vendors to customize Titan-level graphics cards. Before that, NVIDIA's reference cooler was the only option. When they released control to their partners, we started to see water cooled Titan Xs. There is two ways to look at it: either NVIDIA is relaxing their policy of controlling user experience, or they want their personal brand to be more than the cheapest offering of their part. Granted, the GTX 1080 is supposed to be their high-end, but still mainstream offering.
It's just interesting to see this decision and rationalize it both as a release of control over user experience, and, simultaneously, as an increase of it.
Subject: Graphics Cards | May 6, 2016 - 10:38 PM | Scott Michaud
Tagged: pascal, nvidia, GTX 1080, gtx 1070, GP104, geforce
So NVIDIA has announced their next generation of graphics processors, based on the Pascal architecture. They introduced it as “a new king,” because they claim that it is faster than the Titan X, even at a lower power. It will be available “around the world” on May 27th for $599 USD (MSRP). The GTX 1070 was also announced, with slightly reduced specifications, and it will be available on June 10th for $379 USD (MSRP).
Pascal is created on the 16nm process at TSMC, which gives them a lot of headroom. They have fewer shaders than the Titan X, but with a significantly higher clock rate. It also uses GDDR5X, which is an incremental improvement over GDDR5. We knew it wasn't going to use HBM2.0, like Big Pascal does, but it's interesting that they did not stick with old, reliable GDDR5.
The full specifications of the GTX 1080 are as follows:
- 2560 CUDA Cores
- 1607 MHz Base Clock (8.2 TFLOPs)
- 1733 MHz Boost Clock (8.9 TFLOPs)
- 8GB GDDR5X Memory at 320 GB/s (256-bit)
- 180W Listed Power (Update: uses 1x 8-pin power)
We do not currently have the specifications of the GTX 1070, apart from it being 6.5 TFLOPs.
It also looks like it has five display outputs: 3x DisplayPort 1.2, which are “ready” for 1.3 and 1.4, 1x HDMI 2.0b, and 1x DL-DVI. They do not explicitly state that all three DisplayPorts will run on the same standard, even though that seems likely. They also do not state whether all five outputs can be used simultaneously, but I hope that they can be.
They also have a new SLI bridge, called SLI HB Bridge, that is supposed to have double the bandwidth of Maxwell. I'm not sure what that will mean for multi-gpu systems, but it will probably be something we'll find out about soon.
Subject: Graphics Cards | May 5, 2016 - 02:38 PM | Scott Michaud
Tagged: nvidia, pascal, geforce
We're expecting a major announcement tomorrow... at some point. NVIDIA create a teaser website, called “Order of 10,” that is counting down to a 1PM EDT. On the same day, at 9PM EDT, they will have a live stream on their Twitch channel. This wasn't planned as long-term as their Game24 event, which turned out to be a GTX 970 and GTX 980 launch party, but it wouldn't surprise me if it ended up being a similar format. I don't know for sure whether one or both events will be about the new mainstream Pascal, but it would be surprising if Friday ends (for North America) without a GPU launch of some sort.
VideoCardz got a hold of 3DMark Fire Strike Extreme benchmarks, though. It is registered as an 8GB card with a GPU clock of 1860 MHz. While synthetic benchmarks, let alone a single benchmark of anything, isn't necessarily representative of overall performance, it scores slightly higher than a reasonably overclocked GTX 980 Ti (and way above a stock one). Specifically, this card yields a graphics score of 10102 on Fire Strike Extreme 1.1, while the 980 Ti achieved 7781 for us without an overclock.
We expected a substantial bump in clock rate, especially after GP100 was announced at GTC. This “full” Pascal chip was listed at a 1328 MHz clock, with a 1480 MHz boost. Enterprise GPUs are often underclocked compared to consumer parts, stock to stock. As stated a few times, overclocking could be a huge gap, too. The GTX 980 Ti was able to go from 1190 MHz to 1465 MHz. On the other hand, consumer Pascal's recorded 1860 MHz could itself be an overclock. We won't know until NVIDIA makes an official release. If not, maybe we could see these new parts break 2 GHz in general use?
Subject: General Tech | April 28, 2016 - 01:47 PM | Jeremy Hellstrom
Tagged: nvidia, GP100, pascal
The Tech Report takes you on a walk through NVIDIA's HPC products to show you just what is interesting about the Tesla P100 HPC which Jen-Hsun Huang introduced us to. The background gives you an idea of how much has changed from their first forays into HPC to this new 16nm process, 610mm² chip with 56 SMs. If you missed out on the presentation or wanted some more information about how they pulled off FP16 on natively FP32 hardware or how the cache of this chip was set up then click on over and read it for yourself.
"Nvidia's GP100 "Pascal" GPU launched on the Tesla P100 HPC accelerator a couple weeks ago. Join us as we take an in-depth look at what we know about this next-generation graphics processor so far, and what it might mean for the consumer GeForces of the future."
Here is some more Tech News from around the web:
- Microsoft delivers new previews of Windows Server 2016 and System Centre 2016 @ The Inquirer
- Time for a patch: six vulns fixed in NTP daemon @ The Register
- Searching for USB Power Supplies that Won’t Explode @ Hack a Day
- Hackers so far ahead of defenders it's not even a game @ The Register
- Trouble at t'spinning rust mill: Disk drive production is about to head south @ The Register
- Tech ARP 2016 Power Bank Giveaway
Subject: Graphics Cards | April 22, 2016 - 10:16 AM | Sebastian Peak
Tagged: rumor, report, pascal, nvidia, leak, graphics card, gpu, gddr5x, GDDR5
According to a report from VideoCardz (via Overclock.net/Chip Hell) high quality images have leaked of the upcoming GP104 die, which is expected to power the GeForce GTX 1070 graphics card.
Image credit: VideoCardz.com
"This GP104-200 variant is supposedly planned for GeForce GTX 1070. Although it is a cut-down version of GP104-400, both GPUs will look exactly the same. The only difference being modified GPU configuration. The high quality picture is perfect material for comparison."
A couple of interesting things have emerged with this die shot, with the relatively small size of the GPU (die size estimated at 333 mm2), and the assumption that this will be using conventional GDDR5 memory - based on a previously leaked photo of the die on PCB.
Alleged photo of GP104 using GDDR5 memory (Image credit: VideoCardz via ChipHell)
"Leaker also says that GTX 1080 will feature GDDR5X memory, while GTX 1070 will stick to GDDR5 standard, both using 256-bit memory bus. Cards based on GP104 GPU are to be equipped with three DisplayPorts, HDMI and DVI."
While this is no doubt disappointing to those anticipating HBM with the upcoming Pascal consumer GPUs, the move isn't all that surprising considering the consistent rumors that GTX 1080 would use GDDR5X.
Is the lack of HBM (or HBM2) enough to make you skip this generation of GeForce GPU? This author points out that AMD's Fury X - the first GPU to use HBM - was still unable to beat a GTX 980 Ti in many tests, even though the 980 Ti uses conventional GDDR5. Memory is obviously important, but the core defines the performance of the GPU.
If NVIDIA has made improvements to performance and efficiency we should see impressive numbers, but this might be a more iterative update than originally expected - which only gives AMD more of a chance to win marketshare with their upcoming Radeon 400-series GPUs. It should be an interesting summer.