Ampere Discussion Thread

[SteelCrysis]

https://www.extremetech.com/gaming/31589...-into-2021

Jensen also reiterated that the problem with Ampere was overwhelming demand, not an issue of supply. So many people apparently want the cards, it’s impossible to keep them on store shelves. I wouldn’t be surprised if this was part of the reason because Turing uptake has been quite low. A lot of gamers are back on 10xx GPUs, and the RTX 30xx family look like great upgrades relative to those cards. Could Ampere demand be the entire reason stocks are so hard to come by? It could be — but I suspect it isn’t the entire explanation.

First, we know bots were a problem at the RTX 3080 and RTX 3090 launches, and it’s not as if those users are going to have quit using them in the intervening period. Every retailer who isn’t running effective bot detection is going to be a problem.

Second, according to semiconductor analyst Daniel Nenni, Samsung’s 8nm yields are anything but good. In an August 30 podcast, Nenni said: “Samsung has had yield problems – serious, serious yield problems – throughout their history because they were first to a node, but TSMC is always first to high volume manufacturing, so you really have to separate the two.”

This does not automatically mean that Samsung’s 8nm is a low-yielding node, but there have certainly been questions about what yields look like, and no clear answers yet. Samsung has struggled to land major customers other than its own business for more recent advanced nodes. IBM and Nvidia are both fairly recent announcements, especially Nvidia’s high-end manufacturing.
...
The big question now is: “Will AMD have a similar problem?” If RDNA2 and Ampere are relatively well-matched and one of them is $500 theoretically and $800 practically, that’s going to impact people’s buying decisions. AMD probably wins some sales based on Nvidia’s difficulty supplying the market, if prices on Nvidia GPUs stay inflated. Alternately, it’s possible that RDNA2 will get hit by exactly the same wave of bots or upgrade demand. It won’t really matter how RDNA2 compares with Ampere if both of them are impossible to find, and the reason won’t really matter to customers who don’t get to buy one.

Either way, I wouldn’t necessarily pin your hopes on a new GPU this Christmas.

https://www.tomshardware.com/news/evga-i...ing-system

Even now that a few weeks have passed since the RTX 3080 and RTX 3090 launches, availability is still a huge issue. Per Nvidia's own admission, supply problems are likely to persist into 2021, and that isn't good news -- especially with the bots grabbing every purchase the moment cards are available. But, EVGA has come up with a creative workaround to give everyone a fair chance: classic british queueing.
...
EVGA's product management director, Jacob Freeman has also been actively keeping the community updated about RTX 30-series orders through their Twitter account, and credit where credit is due, they are doing a stellar job managing expectations.

Of course, when you receive the email that it's your turn to order an RTX 30-series card, you still only have 5 hours to actually place the order -- so you better pray that it comes at a time when your boss isn't watching.

https://www.techpowerup.com/273010/evga-...hics-cards

Check after the break for the full explanation on how this system works, which was given by EVGA's own Jacob Freeman. The system is first being rolled-out for the US, with other regions following suit according to its success.

https://www.tomshardware.com/news/crysis...benchmarks

A couple of days ago, word got out that a gamer was running Crysis from a 3090 GPU RAM disk. It worked, but of course, we had questions. Specifically, how well does Crysis 3 run from a RAM disk compared to boring old SSD storage? That's easy enough to test, so we set about downloading Crysis 3, VRAM Drive, and ImDisk Toolkit—those last two were required for the GPU and system RAM disk testing.
...
Yeah, that earlier bit about Crysis 3 not being very storage limited? This is the result. Launch times varied by about 0.2 seconds in our testing, but some of that might be human error. No one is likely to notice a 0.2 second difference in load times, and the SATA SSD actually outperformed the NVMe SSD. The GPU RAM Disk ended up coming in last, perhaps just due to software overhead. The VRAM Drive ought to perform as well as other storage options, but again, a few tenths of a second aren't particularly meaningful. The time to load a save game was effectively tied.

As for actual in-game performance, there's a bit more variability between runs, with the VRAM Drive coming out just a hair ahead of the two SSDs. Running the game off of system RAM ended up being the slowest, which again doesn't make much sense, but it was consistently nearly 1 fps slower than the other storage options. There were also still occasional stutters on all of the test options (particularly on the first run, where minimum fps dropped into the single digits), so extreme RAM drive storage didn't fix that.

The issue with RAM drives is that the applications have no idea they're residing on blazingly fast storage. This is why letting the application or game or even OS manage memory is usually a better overall solution. Think about what we've done here in our testing of Crysis 3.

For the RAM drive, we've allocated a chunk of system memory as storage. When we launch the game, the CPU reads data out of that portion of memory, copies it into another section of RAM, then processes the data in various ways and loads some portions of the data into the GPU memory. There's a lot of wasted resources and effort.

For the VRAM drive, it's even worse. Data gets copied over the PCIe bus to the system RAM when the game launches, then it gets processed there, and eventually, textures and other portions of the game get copied back to the VRAM over the PCIe bus. We're using DDR4-3600 memory that provides 57.6 GBps of bandwidth, but the PCIe Gen3 bus only manages 16 GBps. PCIe Gen4 might help in this case, but even so, we're still generally going to end up being limited by other elements rather than storage throughput.

The takeaway is that, besides being an extremely expensive storage solution, VRAM and RAM disks typically aren't necessary for games. It would be better for games to optimize their use of memory better than to try and pre-allocate a fixed portion of memory—system or GPU—as storage, then copy over files to that drive, and maybe gain some benefit. There are situations where a RAM disk can be more beneficial, particularly in the server realm. But for games like Crysis 3? Not so much.