Vigile writes: Benchmarks are finally live for the updated NVIDIA Titan X graphics card that uses a brand new GP102 GPU based on the Pascal architecture and the 16nm process at TSMC. Announced just a week or so ago, PC Perspective has done the full suite of performance tests on the card to find that this beast of a GPU that includes 3,584 CUDA cores running at over 1600 MHz in practice, races past the GeForce GTX 1080 by 30-40%. This makes the new Titan X the highest performance GPU on the market. Users of the GeForce GTX 980 Ti will see 50-80% performance gains with the Titan X and AMD's highest performing graphics card, the Fury X, gets bested by 80-120% in all games but one! Clearly the Titan X is the GPU king but it is going to cost you — NVIDIA is selling it directly for $1200.
Vigile writes: It's been a terribly long news cycle, but today is finally the day reviews and sales start of the new AMD Radeon RX 480 graphics card based on the company's latest Polaris architecture and built on 14nm FinFET process technology. With a starting price tag of $199 for the 4GB model and $239 for the 8GB, the RX 480 has some interesting performance characteristics. Compared to the GeForce GTX 970, currently selling for around $280, the RX 480 performs +/- 5-10% in DX11 games but PC Perspective found that the RX 480 was as much as 40% faster in DX12 titles like Gears of War, Hitman and Rise of the Tomb Raider. Compared to previous AMD products, the RX 480 is as fast as a Radeon R9 390 but uses just 150 watts compared to 275 watts for the previous generation! Chances are that NVIDIA will have a competing product based on Pascal available sometime in July, so AMD's advantage may be short lived; but in the meantime the Radeon RX 480 is clearly the best GPU for $200.
Vigile writes: NVIDIA has now released the second of its two new Pascal architecture GeForce cards to the market, this one with lower prices and performance slightly from the GTX 1080 to create the GTX 1070. Based on the same GP104 GPU, but with 1920 CUDA cores rather than 2560, and slightly lower base and Boost clock speeds, the GTX 1070 is offering performance above the GeForce GTX 980 Ti, a card that launched at $649. PC Perspective tested it at both 1920x1080 and 2560x1440 to address the segments that the $379 MSRP and $449 Founders Edition prices target, and found it was 20% faster than the GTX 980 Ti, 30-50% faster than the GTX 970 and much faster than the R9 Nano or R9 390X from AMD. Even more impressive is that it matches or beats the GTX 980 Ti with 100 watt lower TDP! Cards are expected to go on sale June 10th and if you can find one at the $379 price point, it looks to be the best enthusiast gaming card for those of us without a GTX 1080 budget.
Vigile writes: Though details have been coming out for over a week from NVIDIA, today is the day that reviews of the new GeForce GTX 1080 finally hit. The results are pretty stunning — the GTX 1080 is the new fastest GPU in the world, beating out the likes of NVIDIA's own GTX 980 Ti and AMD's Fiji XT GPU built into the R9 Fury X. The GP104 GPU has 2560 CUDA cores (25% higher than GTX 980) and runs at a base clock of 1607 MHz (42% higher than the GTX 980) while introducing the first GDDR5X memory interface that runs at 10 Gbps. The result is performance 25-40% higher than the GTX 980 Ti and 65-100% faster than the GTX 980! Better yet, it does this while only drawing about 10 watts more than the GM204 part used in the GTX 980 thanks to the new 16nm FinFET process tech. NVIDIA also adds in new features that improve VR performance, a new screenshot tool that is actually pretty impressive and more. Starting price is $599 EVENTUALLY but the "Founders Edition" parts launching on May 27th will be $699.
Vigile writes: While NVIDIA is talking to the world about the benefits of Pascal for high performance computing, PC Perspective dove into the information with more of an ear towards what it might mean for GeForce products and PC gaming later in the year. The specifications alone are impressively insane: 16GB of HBM2 with a 4096-bit memory bus, 3840 CUDA processing cores and 15.3 billion transistors. It will be built on the 16nm FinFET process technology with a massive die size of 610 mm^2; and that is without the memory. The GP100 configuration that was announced sets the clock speed up to 1480 MHz with boost, a significant jump over a Maxwell-based card like the Titan hit just over 1000 MHz. It's unknown what benefits, if any, NVLink technology will provide for gamers, though direct GPU to GPU communication might help with the hurdles of SLI scaling. There are still lots of technological questions as well as even the inclusion of HBM2 may not make it to flagship GeForce products, instead sticking with GDDR5 or GDDR5X based on cost structures. NVIDIA will likely balance power consumption and performance down some, as the Tesla P100 part has a TDP of 300 watts. If current rumors from both AMD and NVIDIA turn out to be correct, it appears we may have a "one big chip (NVIDIA) versus many little chips (AMD)" battle coming this summer.
Vigile writes: With the release of the HTC Vive and SteamVR to the masses today, the hunt is on to find the best and most accurate way of measuring the experiences and performance of VR games and applications. For traditional PC games we have an abundance of tools including frame time measurements, direct screen capture analysis and decades of explaining and detailing the experience a user sees on the screen. But with VR, a lot changes. With the inclusion of the Oculus and OpenVR SDKs, frames aren't just copied or flipped to the screen; instead we have to deal with dropped frames, timewarps, varying animation smoothness levels, etc. PC Perspective has been diving into this new world to try and find a way to measure performance of VR with both experience-based methods (user testing and camera testing) as well on-system metrics like Present() times and GPU render times. There are still a lot of questions and a lot of work to do to get VR testing to the same level as traditional gaming, but the path has been paved. Link to Original Source
Vigile writes: One of the concerns surrounding the recent debate of the Unified Windows Platform and games being released on it, such as the recent Gears of War Ultimate Edition, was the inability for media and consumers, and even entry level developers, to properly profile the performance of those applications. All of the standard testing applications like Fraps, FCAT and other overlays are locked out of UWP games. A Intel graphics engineer released a tool called PresentMon on GitHub yesterday that accesses event timers in Windows to monitor Present commands in any API, including DX11, DX12, Vulkan as well as games built on the Windows Store platform. Using this data, PC Perspective was able to profile the performance of the new Gears of War on PC, comparing frame time variability between the two flagship parts from NVIDIA and AMD. While it's not a perfect utility yet, there is hope now that this open source code will allow for performance metrics on any and all gaming titles.
Vigile writes: Microsoft has big plans to try and merge the experiences of the Xbox One and Windows for gaming but the push back from the community and from major developers and personalities is mounting. Earlier this week PC Perspective posted a story that detailed the controversy around DX12 performance analysis without an exclusive full screen mode, changes to multi-GPU configurations and even compatibility issues with variable refresh that crop up from games from the Windows Store. Microsoft's only official response so far as been that it is listening to feedback and plans to address it with upcoming changes. Now today, Epic's Tim Sweeney has posted an editorial at The Guardian with an even more dramatic tone, saying that UWP (Unified Windows Platform) "can, should, must and will, die..." Clearly the stakes are being placed in the ground and even damage control from Phil Spencer on Twitter isn't likely to hold back angry PC users.
Vigile writes: Valve took another step to prepare the world for VR gaming by releasing the SteamVR Performance Test today. This application that is free to download through Steam, runs a portion of the Aperture Science Robot Repair demo originally built for the HTC Vive VR headset, and reports back performance metrics and a grade for your PC's hardware. Scores include a Not Ready, Capable and Ready result as well as an "average fidelity" numeric score that is even more interesting. Valve integrated a dynamic fidelity feature "that adjusts image quality of the game in a way to avoid dropped frames and frame rates under 90 FPS" — a target for an acceptable VR experience. Early results put the GeForce GTX 980 Ti at the top of the GPU stack though AMD's Radeon products do very well at every price point below $600. Is your wallet ready?
Vigile writes: Even with Samsung pushing forward into 3D NAND with 32-layer technologies used in SSDs like the 850 Pro and the recently released M.2 PCIe NVMe 950 Pro, there is still plenty of traditional 2D planar memory being fabbed on production lines. To utilize that inventory Samsung is shifting its low capacity SSDs back to it, announcing the 750 EVO drives today available in 120GB and 250GB capacities. Though based largely on the very popular, but sometimes troubled, 840 EVO specs, the new drives are faster and start with some impressively low prices. The starting MSRP for the 250GB 750 EVO will be just $75.
Vigile writes: The gang over at PC Perspective just posted a story that looks at a set of three M.2 form factor Samsung 950 Pro NVMe PCIe SSDs in a RAID-0 array, courtesy of a new motherboard from Gigabyte that included three M.2 slots. The pure bandwidth available in this configuration is amazing, breaching 3.3 GB/s on reads and 3.0 GB/s on writes. But what is more interesting is a new testing methodology that allows for individual storage IO latency capturing, giving us a look at performance of SSDs in all configurations. What PC Perspective proved here is that users often claiming that RAIDs "feel faster" despite a lack of bandwidth result to prove it, are likely correct. Measurements now show that the latency of IO operations improves dramatically as you add drives to an array, giving a feeling of "snappiness" to a system beyond even what a single SSD can offer. PC Perspective's new testing demonstrates the triple RAID-0 array having just 1/6th of the latency of a single drive.
Vigile writes: Samsung just released its first non-OEM, consumer level NVMe enabled SSD, the 950 Pro series. This drive will ship in an M.2 form factor rather than a 2.5-in drive size that is the standard for users today, allowing installation into notebooks, small form factor PCs and desktop PCs that have at least one M.2 slot on-board. It peaks at 512GB capacity today but Samsung promises a 1TB version using 48-layer VNAND in 2016. The NVMe protocol allows much better performance directly over the PCIe bus without the overhead of the AHCI protocol used in hard drives and previous SSDs. PC Perspective's review has performance breaking the 2.5GB/s read speed level while also introducing an entirely new type of performance evaluation for SSDs centered around latency distribution of IOs. By measuring how long each IO takes, rather than reporting only an average, the performance of an SSD can be determined on a per-workflow basis and drives can be compared in an entirely new light. There is a lot of detail on to be read over and digested but again the new NVMe Samsung 950 Pro impresses.
Vigile writes: In preparation for the release of the free-to-play Fable Legends game on both Xbox One and PC this winter, Microsoft and Lionhead Studios released a benchmark today that allows users to test performance of their PC hardware configuration with a DirectX 12 based game engine that pushes the boundaries of render quality. Based on a modified UE4 engine, Fable Legends includes support for asynchronous compute shaders, manual resource barrier tracking and explicit memory management, all new to the DX12 API. Unlike the previous DX12 benchmark, Ashes of the Singularity, that focused mainly on high draw call counts and mass quantities of on-screen units, Fable Legends takes a more standard approach, attempting to improve image quality and shadow reproduction with the new API. PC Perspective has done some performance analysis with the new benchmark and a range of graphics cards, finding that while NVIDIA still holds the lead at the top spot (GTX 980 Ti vs Fury X), the AMD Radeon mid-range products offer better performance (and better value) than the comparable GeForce parts.
Vigile writes: Back when AMD announced it would be producing an even smaller graphics card than the Fury X, but based on the same full-sized Fiji GPU, many people wondered just how they would be able to pull it off. Using 4096 stream processors, a 4096-bit memory bus with 4GB of HBM (high bandwidth memory) and a clock speed rated "up to" 1000 MHz, the new AMD Radeon R9 Nano looked to be an impressive card. Today PC Perspective has a review of the R9 Nano and though there are some quirks, including pronounced coil whine and a hefty $650 price tag, it offers nearly the same performance as the non-X Radeon R9 Fury card at 100 watts lower TDP! It is able to do that by dynamically adjusting the clock speed from ~830 MHz to 1000 MHz depending on the workload, always maintaining a peak power draw of just 175 watts. All of this is packed into a 6 inch PCB — smaller than any other enthusiast class GPU to date, making it a perfect pairing for SFF cases that demand smaller components. The R9 Nano is expensive though with the same asking price as AMD's own R9 Fury X and the GeForce GTX 980 Ti.