The NVIDIA GeForce RTX 2060 6GB Founders Edition Review: Not Quite Mainstream
by Nate Oh on January 7, 2019 9:00 AM ESTCompute & Synthetics
Shifting gears, we'll look at the compute and synthetic aspects of the RTX 2060 (6GB). As a cut-down configuration of the TU106 GPU found in the RTX 2070, we should expect a similar progression in results.
Starting off with GEMM tests, the RTX 2070's tensor cores are pulled into action with half-precision matrix multiplication, though using binaries originally compiled for Volta. Because Turing is backwards compatible and in the same compute capability family as Volta (sm_75 compared to Volta's sm_70), the benchmark continues to work out-of-the-box, though without any particular Turing optimizations.
For Turing-based GeForce, FP32 accumulation on tensors is capped at half-speed, thus resulting in the observed halved performance. Aside from product segmentation, that higher-precision mode is primarily for deep learning training purposes, something that GeForce cards wouldn't be doing in games or consumer tasks.
Moving on, we have CompuBench 2.0, the latest iteration of Kishonti's GPU compute benchmark suite offers a wide array of different practical compute workloads, and we’ve decided to focus on level set segmentation, optical flow modeling, and N-Body physics simulations.
Moving on, we'll also look at single precision floating point performance with FAHBench, the official Folding @ Home benchmark. Folding @ Home is the popular Stanford-backed research and distributed computing initiative that has work distributed to millions of volunteer computers over the internet, each of which is responsible for a tiny slice of a protein folding simulation. FAHBench can test both single precision and double precision floating point performance, with single precision being the most useful metric for most consumer cards due to their low double precision performance.
Next is Geekbench 4's GPU compute suite. A multi-faceted test suite, Geekbench 4 runs seven different GPU sub-tests, ranging from face detection to FFTs, and then averages out their scores via their geometric mean. As a result Geekbench 4 isn't testing any one workload, but rather is an average of many different basic workloads.
We'll also take a quick look at tessellation performance.
Finally, for looking at texel and pixel fillrate, we have the Beyond3D Test Suite. This test offers a slew of additional tests – many of which use behind the scenes or in our earlier architectural analysis – but for now we’ll stick to simple pixel and texel fillrates.
Facebook
Twitter
RSS
134 Comments
View All Comments
JRW - Saturday, February 23, 2019 - link
2060 is considerably faster than a 580 tho, I recently upgraded from an R9 290X to EVGA RTX 2060 XC Black and love it, the 290X served me very well tho great card even with todays games @ 1080P but struggled a bit trying to hit my monitors 144hz refresh.PeachNCream - Monday, January 7, 2019 - link
Turing's MSRP makes the benchmark performance meaningless.jrs77 - Monday, January 7, 2019 - link
Midrange card for 350 bucks... :facepalm:I don't care if it's as fast as a 1070ti. A xx60 series card should never cost more than 250 and the 1060 was allready overpriced for most of the time, due to all that bitcoin-fuckery.
Manch - Monday, January 7, 2019 - link
The Vegas are a good bit cheaper than what the scale shows. Not just on sale but regular price reductions. Even mentioned in the article so why tye discrepancy? Also I thoight Vega was a bit slower than the vanilla1080. Its showing to be faster than the FE?sing_electric - Monday, January 7, 2019 - link
I'm not sure what you're referring to, since the best deal I've heard of on the Vega 56 was ~$320 on Black Friday, and today, I can't find a card for less than $370 (at NewEgg on one model, all others are $400+). I like AMD but given today's prices, the only price category where I think AMD wins right now is with the ~$200 580. The ~$280 RX 590 is most of the way to the 2060's MSRP but offers significantly less performance.Manch - Monday, January 7, 2019 - link
Per the article, ". In the mix are concurrent events like AMD-partner Sapphire’s just-announced RX Vega price cuts, which will see the RX Vega 64 Nitro Plus moved to $379 and the RX Vega 56 Pulse to $329, and both with an attached 3-game bundle" Thats even better than what Ive seen.I just bought a MSI vega 64 from amazon for $399 with the 3 game bundle in Dec. Ive seen on avg 400-450 for Vega 64 and a good bit lower for Vega 56.
The chart has Vega 56 at 499 which isnt the case.
Manch - Monday, January 7, 2019 - link
Vega 64 $399, Vega 56 $368 new egg. Plus 3 games.Manch - Monday, January 7, 2019 - link
vega 64 $399 on amazon as well. There are higher pri ed cards but who cares is theyre readily available at these prices?Vayra - Wednesday, January 9, 2019 - link
They also take twice as much power at the wall. *poof* there go the savings. And you get free extra noise and heat in the case to boot.Manch - Friday, January 11, 2019 - link
Double?! LOLCompared to a 2060? The avg diff according to Anand's Bench is 130watts.
Avg price of electricity in the US is 12 cents a kilowatt hour. That means it would cost you 1.2 cents per 100watts an hour. It would cost you on average 1.668 cents more an hour to run a VEGA 64 at full bore balls out compared to the 2060. If we then calculate the difference for an entire year @ 100% power draw for 365 days or 8760hrs the total comes out to $146.12 Here in Germany it would be about double that.
Lets be real no one does that. (Miners?)
Avg is 12hrs a week! Highly doubtful the card is running 100% for 12hrs a week but if it were.
52 weeks in a year, 12 hrs a week for 624hrs for a soul crushing total of $10.41
So yes it cost more to run a higher power card....duh, but it's not double. Stop the FUD.