It only depends on what condition your cooling system is in.
To obtain this information download some software which shows CPU temperatures and see what they are after say 20-30 minutes of work. Anything below 80C is ok for a desktop, if it's 90-100 you need to worry (CPU won't die as it has good self-defence mechanism but it will be slower).
A good fan costs ~10-15$, you shouldn't see temperatures much above 60C even after running it for 12 hours straight if you use it.

Hello,

I made some tests with the free version of Jesolver.
I run several scripts in PioSolver and than the same scripts in Jesolver and compared the results. There are big difference in the result on some combos. It is interesting witch solver algorithm is more close to the best GTO strategy.
Did someone else who have Jesolver test this?

Differences on given combos are completely normal and don't mean much.
The solution is measured in distance to Nash and there are many ways for a strategy to be say 0.1 from Nash.
That being said you are able to get better accuracy in shorter time with Jesolver as of today as it both runs faster and is more precise.
That both solvers measure it in exactly the same way is easy to test:
-build some sizeable tree on the flop
-click "calculate results"
-notice that max exploits for both are exactly the same

It's possible that in the future the saves will have common format as well although this is a bit problematic (possible for full trees painful for small ones).

As expected the preflop solver had some issues in the first release, some of the most urgent issues will be fixed early next week, please report weird things/bugs so I may have a chance to squeeze fixes in the release. The fixes so far include:

1)performance: the solver didn't utilize 6+ cores very well; it will be now between 20% (on quads) to 35% (on 16+ cores) faster
2)convergence: will be improved so it converges faster and preflop strategies are a bit smoother;
3)tree building: stuff like "change raise size" option, error checking for missed folds/calls, error checking for stack sizes etc. is now added
4)errors with calc_ev_line an calc_ev (now impossible during solving) will be now handled nicely
5)Sometimes the solver crashes during saving of huge trees (usually 100+GB) - it remains to be seen if it's hardware or software issue; all the signs point to hardware as of today as the solver is run on hundreds of VMs every day and it never crashes for people who do high volume (and have ecc RAM). That being said it's entirely possible there is some use path which leads to crashes and some error checks are missing. For how to report it, see below.

There seems to be an issues with pure preflop games with many bet sizes. It is possible but please when you report any weird problem always:

1)describe steps which led to it
2)include the config (save current preflop tree parameters and then go to PreflopTreeBuilding and attach the .txt file)

Otherwise the report is not useful as I can't do anything about it.
Sending me screenshot with "aha, this doesn't look right!" is useless as well (99% it's user's error, remaining 1% I don't know where to start investigating).

Overall please refrain from reporting issue directly to me on Skype (the group is of course fine). I get 20-30+ questions/suggestions/reports every time I log in and I can't handle the volume. Write a a report according to the instructions above and send an email to support@piosolver.com instead.
If you ask about general computer support not related to the solver you will be instantly removed from my Skype contacts as you are wasting time which is better spent doing stuff which helps more people.

I understand those are busy times because everyone wants optimal preflop for all spots ASAP but please understand I am doing everything I can to deliver better experience so patience is appreciated.

Thank for your attention :-)

"Saves don't work"

The reason: there is a bug in the viewer which causes a crash when loading saves caused by presence of .txt files in the same folder (it tries to load them as tree config).

Explanation: the mechanism was necessary when solving batches of trees (there is only one config there for many trees) and was implemented a bit hastily.

Workaround for now: keep saves in different folder than your .txt files or move them to a fresh folder if they are already there. Everything should work then.

Hi, I was wondering when the Xmas discount would occur ? After Xmas ?

Same

Starting tomorrow and lasting to the end of 2015.
Btw, Christmas here span from 25th to26th is that that atypical?

Yes, Christmas in the US lasts from the day after thanksgiving until kids go back to school.

Will all versions feature the preflop solver after a while?

how can i 'unregister' one of my licenses and move it to another computer? i m pretty sure i ve seen it somewhere but cant find it atm :/

Does the preflop solver also work for limit holdem? Any chance you are going to cooperate with Tammelin in nearby future and have his engine/solver built in, say to the Edge version of PioViewer?

http://piosolver.myshopify.com/pages/faq (point 5)

If that doesn't work drop us an email with your key and info which activation you would like to reset (in case of pro version earlier/latter/both).

You know, in 2018 the solver's 10x faster with user interface with a lot more features than the current ones will cost 150$. The question is when that happens.
Preflop solver is a tool for a niche market (because you need a very decent computer and/or rent one) and that's why the price tag is what it is today - it wouldn't make sense to develop it for less.

Oh, I see. Here we have a Christmas break from like 23th and it lasts to 1st Jan.
Anyway, it starts tomorrow and will last for a week.

thx, seems like it worked, appreciate your fast responses

Xmas sale still isn't on ? I looked around the website but can't find any evidence of it yet

I am sorry I've got overwhelmed with work a bit.
I am writing the announcement right now. Those should be up in few hours (and again last to the end of 2015).

EDIT:
It's here now:
http://piosolver.myshopify.com/blogs...hristmas-sales

I hope the offer is interesting. It's likely we are going to put some cheaper upgrades in as well to give people who bought it before the sales some nice offer as well.
If you bought a basic version in last 2 weeks, contact me, you can get something extra as well.

Hey punter!

Sick promo, im one of many who is very interested in a gto solver, however as im a laptop person i have i5 and 4gb ram available, only. Will i be able to calculate flop solutions in reasonable time?

It depends which i5 (there are mobile i5s and normal i5s and almost normal i5s). 4Gb is a problem though. While a lot of trees fit in 500mb - 2GB range these days there are other things running on a system and you rarely get say 2.5-3gb free. I wouldn't advice going for the solver with such specs at this point unfortunately.

Hey, as always thanks for the quick answers !

I saw this offer and wanted the confirmation that it really includes 2x preflop solver activations like it seems.

Thanks !
http://piosolver.myshopify.com/produ...-2-activations

Hi, in edge in sale are 2 preflop solvers?

Preflop solver is now part of the edge license. So every computer you activate your edge license on you can use both preflop and postflop ones. It means that it in fact includes 2 of them and if you get a standard one you get 3.

I hope that clears things up. Our original plan of only including one preflop solver was withdrew as we thought it wouldn't be convenient enough for our customer to only have one.

Does removing an obviously unused portion of a range speed up preflop calcs? For example, if I remove a significant portion of BB's potential range defending vs EP raiser, will that be helpful? I realise this requires some knowledge about the spots in order to not accidentally cut away +EV hands.

What about going into a node or two once a tree is constructed, and before starting the solver, to quickly make a very rough strategy to jumpstart the solving - perhaps a range saved from a similar calc?

Wondering about any practical ways to speed things along. Keep up the good work!

Hi, I wonder how preflop solver knows that in 6max utgvsbu oop is utg and he act first pre flop, and buvsbb bu is oop but he acts second pre flop.

Not right now but maybe I manage to add it to the next update if not, then to the next after that.

No for reasons different than just me not adding it. Only removing hands from both ranges has potential to speed it up significantly (removing them from one speeds it up but it's a very minor effect).

This idea in general is hard to pull off and didn't work with all my tests so far.
The reason is that say in a flop tre 98%+ of the data is on the river and in preflop solving 99.9%+ of the data is on the flop+. Those places in the tree needs to be solved and it will take comparable time to do that regardless what starting flop strategy is.
There are surely tricks to pull it off but I was never able to make anything significant out if this idea.

For now the best one is to wait 2-3 days, the updater will be significantly faster (20%-35% depending on computer + a bit faster convergence).

In the full preflop tab you can choose blinds for IP and OOP (those describe postflop play), so:

1)SB vs BB on a HU table:
-IP money in the pot: 5
-OOP money in the pot: 10
-dead money: 0

2)SB vs BB on a 6max table:
-IP money in the pot: 10
-OOP money in the pot: 5
-dead money: 0

3)BTN vs BB on a 6max table:
-IP money in the pot: 0
-OOP money in the pot: 10
-dead money: 5

4)UTG vs BTN on a 6max table:

In general impossible to solve because there are two more players to act but if you want to solve after BTN already 3bets:

4)BTN vs UTG on a 6max table, after BTN 3bet:
-IP money in the pot: 0
-OOP money in the pot: 0
-dead money: 15

and force (add only those lines to the tree) say bet 30 and 3bet to 90, after that add 4bet/call/fold.

Also always build a tree on 5 flops first and go browse it for a while before solving to confirm it is what you want. After that go back, rebuild on as many flops as you want and start solving.

The last case will be way slower than it should be in this release so maybe it's better to wait with those.