Jump to content

Alonzo

Members
  • Content Count

    1720
  • Joined

  • Last visited

Community Reputation

1889 Excellent

4 Followers

About Alonzo

  • Rank
    Member

Profile Information

  • Gender
    Male
  • Location
    Canada

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Based on the AMD numbers, their 6900XT basically makes the RTX 3090 a joke. I was never going to buy one anyway, but I did back-order a nice EVGA 3080 FTW3. But now we have the 6800XT and 6900XT. I play in VR, so that's my whole use-case for the card. Against the 6800XT, the 6900XT for +53% cost and +10% performance seems like poor value. But with VR I need every ounce of performance available. It does seem like the 6800XT is the sweet spot. Could the 6900XT be "worth it" as a step up?
  2. This works better when there are lots of players on the server, both those that want to fly (and protect) transports, and those who want to get an easy kill on an A-20 that is landing to deliver supplies. I'm not sure how often players really get into fights around the supply fields, it was intended as something a little different than 'normal' missions, hopefully interesting.
  3. The problem with the 8700 (non-K) chip is that it's a bit too slow for good VR frame rates. You need to hit 90 FPS for the Reverb G2, 80 FPS for the Rift S (and Valve Index). You're also a little too slow with the RTX 2070, even for the Rift S, and definitely too slow for the G2. RAM speed is also a factor, you want at least 3200 Mhz or better. With your current setup I think you can reliably hit half-framerate mode, (40 FPS doubled to 80, or 45 doubled to 90). With the frame doubling mode you're going to see "ghosting" on planes that are moving rapidly past your field of view, such as in a dogfight. For some players they don't mind the ghosting, for some players (like me) they can't stand it. So I reduce my settings a lot in order to try to hit the full frame rate. You might be able to buy or borrow a Rift S to test it out on your rig. If you don't like the frame doubling mode, you can return it (depends on the retailer -- check their policy). I would 100% recommend VR but for IL2 (and the other flight sim) you need a very powerful rig to hit full frame rate. I was playing the other sim at 60 FPS doubled to 120, though, and it looked fine to me. The ghosting is more of a factor for up-close WW2 knife fights.
  4. Correct, it's players attempting to fly a supply plane and land at those airfields.
  5. It's definitely worth being patient (hard, I know, I want faster pixels today please!). I think a lot of enthusiasts watching this video are going to consider an AMD card, and I think NVidia is going to have to respond with some price reductions, at least once AMD actually has inventory in the sales channels. I think the 3070 or 6800 (non-XT) are about where you should be looking for solid 4K60 performance. Someone yesterday said "the best deal on a 3070 is to save up $200 more and buy a 3080" and I think I agree with that assessment, but this AMD release could cause NVidia to drop the price on the 3070 a little. It's ~2080ti performance which has always been good for 4K -- the new marketing hype around "finally 4K ready cards" is BS, it's moving goalposts, this year's games are more demanding, but you've been able to play "today's 4K games" for several years on the top cards.
  6. Yes, exactly. Rumor is that NVidia is prepping a cut-down 3090 with 12GB VRAM and slightly less performance, and will call it the 3080ti. That's already a good result for gamers. NVidia is really hurting from their decision to go with 8nm, the chips are hella power hungry. Apparently they played some kind of game of chicken with TSMC, didn't want to pay what TSMC wanted to charge for the 7nm process, and got stuck with Samsung's 8nm and hence hotter, more power-hungry cards. Rumor mill is NVidia is going to cut their losses on this generation and try to get a 4000-series out mid/late next year on the 7nm node.
  7. I'm sure anyone following this thread will already have seen it, but AMD is releasing the following cards: 6900 XT -- trades blows with the 3090, 16GB VRAM (vs 24GB), 300W total board power (vs $LOTS), $999 (vs $1400+ for the 3090) 6800 XT -- trades blows with the 3080, 16GB VRAM (vs 10GB), 300W total board power (vs 350+), $649 (vs $699 for the 3080) some other one I didn't pay attention to but it's the 3070 competitor, similar vibes We still don't know the availability of these cards or NVidia's response, but this is some serious competition from AMD.
  8. I'm interested in the naming. It'll help me decide which leakers to listen to next time around. My favorite guy (Tom from Moore's Law is Dead) maintains that it's not necessarily called the 6900XT, it might be the 6800XTX, and he thinks it'll be an AMD exclusive (no board partner models). He's saying they'll only call it the 6900XT if it beats the 3090, otherwise they'll stick with a 6800 moniker. Either way, we as consumers win. Well, all of us who didn't buy a scalped 3080 for $1800. Competition is good for all of us. And frankly I'm happy if AMD charge a bit more for their cards (since NVidia has set the price benchmarks) and make some money and continue to be a strong competitor in future.
  9. From the Hardware Unboxed review. So the 3070 is basically 2080ti level performance. I'd say the 3070 is a good recommendation for VR users and would do a great job driving a Rift S or Valve Index, and a reasonable job driving a Reverb G2. Although if you're buying a G2 it's probably worth the extra money to move up to a 3080. BUT! Tomorrow is AMD reveal day. They might surprise us, and the rumor is they have a card that will be competitive with the 3070 or even the 3080, and their pricing choices will probably force NVidia to respond. So things could change.
  10. Thank you Rapidus. I have a Valve Index, but other users are reporting the problem also with Oculus Rift S. Here is some testing that Chili did on Combat Box, and some testing also on Berloga server. The frame time spike is exactly every 5 seconds -- maybe in multiplayer there is some kind of periodic message from server to client that causes a CPU spike, or something else the client is doing every 5 seconds, but only in multiplayer.
  11. Not if you know the motherboard BIOS. Just set the limits really high and the chip stays at max boost max power draw. My chip is an 8086K @ 5.1ghz, hyper-threading off, something like 1.38v CPU vcore, RAM at 3800mhz 1.4v. The CPU is cooled by a fairly modest Noctua air cooler, the NH-U14S (needed a smaller one for RAM compatibility). My secret is that I delidded the CPU and replaced the crappy thermal paste with liquid metal, which got be about a 15C temperature reduction. But yeah. Get a 9000 or 10000-series Intel with the soldered TIM and a 240/280/360mm AIO and you'll be fine at 5.0 or 5.1ghz.
  12. Personally I just create a benchmark track each time I need one. Remagen spawn point, some bombers, shoot them up and fly low over the city, 2 minutes. Just like Chili. It only helps me benchmark my system against itself but that's often what I want to do. Find sweet spot between all the graphics options, supersample, etc. I do wish we had an in-game benchmark for comparison between systems.
  13. Maybe buy a headset from a place that has a good return policy, and if it's not worth it, return it. VR will work and you'll get a sense of 'presence' in the airplane, but VR is already much less competitive than flat-screen + TiR. VR without the depth perception might not be worth it.
  14. The statistics reset every month, so it's deaths during the whole month, and it's a personal statistic. It is not really "in proportion to other players deaths" but I guess you could look at it that way. What I mean to say is, the calculation for each pilot is individual, it does not look at someone else's numbers to calculate yours. But yes, numbers will include a factor for how many times each pilot died. Personally I think the 'rating' calculation is a bit flawed, for several reasons: The rating 'death penalty' is quite significant, and is very significant if you compare someone who has died zero times to someone who has died twice. The person who died twice will have half the rating of the pilot who didn't die at all, if the rest of their points are the same. Therefore... ...fighter pilots who fly at 8km and make 30 kills in the first week without dying can end up on the top of the scoreboard and remain there all month, even without flying, ahead of fighter pilots who actually take risks, fly every day, and rack up 200+ kills but die every now and then. Ground attackers who fly on the server at off peak times and face little opposition can 'farm' ground kills and also rack up big scores and get high on the scoreboard, if they don't die. We could change the rating calculation, but then we'd need to agree on what makes someone the 'best' pilot 🙂 So for now, we have left the default IL2 stats calculation. It can be gamed, sure, but it's not awful. I think it was made marginally better by adding a small score for flight time (minutes flown) and increasing ground target points values.
  15. I think you are talking about 'rating'. As discussed earlier in the thread, overall pilot rating is based on missions flown, points scored, total flight time, and deaths/captures. In particular, deaths/captures is the denominator in part of the calculation, so when you die, your calculated rating might be reduced, depending on your other scores. This is the calculation used by IL2 stats: def update_rating(self): # score per death sd = self.score / max(self.relive, 1) # score per hour shr = self.score / max(self.flight_time_hours, 1) self.rating = int((sd * shr * self.score) / 1000) It calculates "score per death" and "score per hour" and uses them to calculate the overall rating.
×
×
  • Create New...