Solid DNA blog

Blog about stuff on Solid Edge CAD software

New year, new software, new hardware

Has informative this article will be it will also be controversial, so let’s plunge and commit my self

There two basic personalities that we can refer when it comes time to buy hardware for CAD station:

  •         No budget limit

The «no budget limits» type usually does not care about performance. Because of «the more you paid the more you got» philosophy. Those are the first to blame software’s poor performances. Hey! They have bought the best hardware possible.

  •         I need to have good ROI

The «I need good ROI» usually make a trade off toward the budget, so they will pay less. Those usually overestimate their system performance. Therefore, they will blame software’s performances also because they have the feeling they have made a good deal on the hardware.

In a world where everything’s should go fast, even the fastest, the newest or the most recent pc cannot compete.

In a world where capitalism, dividend and profit reign, we look forward to pay less.

Have your mind free about your hardware purchase; position yourself between those two types.

Miracle recipe do not exist, what I offer is my point of view and a small guide to help choose.

Budget 

It seems obvious to fix a realistic budget.

However how many of you fix that budget after beginning shopping or seeing the rebates?

Proud of the budget you have decide, you start seeing other vendors to get the best price. From there your budget change at every place you go. And usualy it goes up……

My point of view, this is like an open vendor’s buffet. No matter how good you are at negotiate, if vendors fix your budget there is a good chance that you lose somewhere.

It is important that you fix your budget base on what you can or are willing to afford.

Peace of mind

The next one is more important, it will ensure peace of mind toward your purchase. This peace of mind determines your satisfaction. 

Usually those who paid more get more satisfaction, simply because they are emotionally ready to pay that price. Thus if they pay less they have the feeling they have made a better deal.

Remember the world of capitalism, dividend and profit.

For those who price is more important, they lost their emotivity and rationality in front of a price that melt like ice in sun. Vendors are experts detecting those behaviors; some even make a specialty in that field of lower prices.

Many of you do not care about theory, you do a Cartesian work… we want numbers🙂

Your rational and comfort zone should be reflect in your budget. All currency will be in Canadian and are base on phone calls, discussions and mails while doing technical support.

Expect your budget to go around 1000$ and 2000$. The 2000$ seem to be a consensus has a psychological barrier. Those with special needs will be willing to pay more because they do not have the same average needs then the rest of the market. Pass the mark of 2000 the emotive side grow exponential.

In our financial context, this upper limit is closer to 1750$.

When context is optimal this upper limit goes around 2250/2500.

I heard some reader’s murmur coming to my ears. Each of you has a slight different situation then the neighbor. This is not the recipe that you follow eyes blind. This guide reflects questions ask thru out the years.

First stop in the budget – Video card

It is not my intention to go in-depth here, however let say that we are in a transition. Open GL is no longer a requirement. Introduction of Vista have start the transition by using DirectX 3D. Open GL organization have launch OpenGL 3.0 will see in 2009 what impact this has on the CAD. Major players seem to go along with DirectX.

Siemens support Direct X since V19/20

Autodesk thru and extensive PDF have explain their position for their next release.

For Solidworks, I have not read any position, so my best guess they will stick with Open GL

Get back to a more technical stuff….

What is important make sure you have dedicated memory (onboard RAM). Usually this starts at 256 and go up to 1 gig.

Also important;

  •         Number of tessellation (number of triangles per second)
  •         Texel/sec (how fast a card render those triangles)

 With 200$ to 500$ you can get the lower professional card ( FX 570 256 MB). On the other hand, with 100 and 200 you could get a good general card (9500 GT 1 gig)

For those who need facts…The rule of thumb still apply go with a certified video card for CAD.

Personally, I found Nvidia with its line of Quadro a good choice.

I found this article about Nvidia card date 2008-07-02

http://www.deskeng.com/articles/aaajxm.htm

Here two short texts extract from article found on the net.

«…..Quadro versus GeForce

The NVIDIA Quadro family of professional graphics cards is very, very expensive. They are generally 2-5X more expensive than their gaming counterparts, the NVIDIA GeForce graphics cards. However, everyone knows that Quadro and GeForce graphics cards are virtually identical in hardware.

Yes, you read that right. Even with the unbelievable price tag, the NVIDIA Quadro is really no different from their desktop GeForce counterparts. So why is NVIDIA charging you so much for a Quadro?

It all revolves around the driver support for professional 3D applications like 3ds Max or Maya. Quadro drivers allow the Quadro to be used to accelerate the rendering operations of such professional 3D applications while GeForce drivers do not. This is the basis for the premium prices NVIDIA (and ATI) charge for their professional-grade graphics cards…..»

This one is more technical….

«…..What is AGP? PCI/x16?

These terms will likely appear on any box of a video card you wish to buy, so make sure to look at them before you buy it.

PCI – This is a standard port inside your computer. Any computer made within the last 15 years has these ports inside of it. They were once used for Video Cards, but they just couldn’t keep up with increased demands for speed and power. They maxed out at 256 MB (MegaBytes). The most powerful PCI Video cards by today’s standards are mediocre at best.

AGP – Successor to PCI. Since the PCI port couldn’t keep up, AGP (Accelerated Graphics Port) was created to work with increasing demand. Nevertheless, in recent years even AGP hasn’t been able to keep up with the more powerful Video Cards of the day. They max out at 512 MB

PCI x16 – This is today’s standard. Almost all new video cards use this kind of connection. Their limits have not been reached yet. Currently, top video cards have as much as 1 GB (GigaByte, or 1000 MegaBytes)…..»

Second stop – the RAM

Basically, the more you have the better it is.

There is a line where you will have to make a decision…. 3GB. If you invest in more than 3GB you will have to make sure, you can use a 64 bits operating system in our day to day.

The other characteristic for the memory is the speed of memory. Two systems with similar price choose the one with RAM speed rate higher.

Without going to much in details…

http://www.cherlocksolutions.com/resources_articles_understanding_RAM.htm

Who can talk about RAM without asking THE question?

What is the size of assemblies manageable with my PC?

In theory, the limit is very high. Especially with 64 bits (264). Here the reservoir is very huge, but at what speed can I crunch the information inside?

32 bits = 4 294 967 296 (4 gig)

64 bits = 18 446 744 073 709 551 616 (18 something)

128 bits = 3.4028236692093846346337460743177e+38 (not before few years)

The real debate should not be around parts. Instead, it should be around:

  •         Total number of faces
  •         Their complexity
  •         How much MB do I need to load to represent them in space

In addition to that did the designer take all means place at his disposal to increase the density of information per memory blocks.

Finally, what is your level of comfort?

Therefore, since I am a blogger and I have free time hihihihhih, I will place myself in trouble especially for those who cannot live without exact number (the Cartesian one)

From a non-scientific study and my unique personal experience.

To set a base point for the debate, base on what I just write …..

For 3gig of RAM you can expect assembly around 10 000@20 000 parts/sub-assemblies. If you prefer 5000@7000 per gig of ram. It is possible to get higher numbers 10, 15, 20, 30, 50 000 the more you increase the density Parts/RAM, the more serious you will have to be managing your project.

I heard customers yelling at me, competition making validations and comparisons. If your type are « My father is better then ours»

Here three charts, instead of evaluating the system, you will evaluate yourself. How good are you at managing assemblies? Versus the others.

  •         With a marker, on the right find the average assembly size (number of parts including sub assembly)
  •         Move the marker to the left until you are align with the system memory at the bottom.
  •         Move down until you intersect the curve.
  •         Count the small mark on the left to extract your index.
  •         Weight our project in MB and divide by the index.

Then multiply by the comfort index (1-incomfort. 2-correct, 3-comfortable)

Choose the league you play in:

Minor league – 2500 (parts/sub-assemblies) with 2 gig of RAM = index of 11

If the project is 300 MB, you got a score of 27

I am comfortable with that project 27×3 = 81

If someone has a similar project with 500 MB:

He is uncomfortable = 45 (when we compare we are happy)

Correct = 90 (I am in the average but I could do better)

Comfortable = 135 (There is better manager then me, I have work to do)

Third step – CPU/motherboard

CPU is simple, unless you have specific needs. For Intel or AMD chose the one with the higher clock speed. The number of core has is importance. However, not has much has the clock speed.

Important, look at the L1 and L2 cache clock speed and size. Pick higher one.

Motherboard many times we do not have the choice

From and internet article…

«….A cache is a small, higher speed memory system which stores the most recently used instructions or data from a larger but slower memory system. Because programs frequently use a subset of instructions or data repeatedly, the cache is a cost effective method of enhancing the memory system in a “statistical” method, without having to resort to the expense of making ALL of the memory system faster.

For example, when the processor (CPU) performs a read (fetch) operation, it will first check to see if the data or instruction it requests is in the cache – if it is, then the cache provides the data very quickly to the processor, without having to access slower main memory (DRAM or ROM). If the data or instruction is NOT in the cache, the CPU fetches the item needed from DRAM or ROM, AND also stores a copy in the cache (assuming the “page” of the address has been marked as “cacheable” by system software). Now this instruction or data item is available in the cache if the processor requests it again. The larger the cache, the more instructions and data the cache can store, and the more likely it is for the requested item to be in the cache.

An n-th level cache is n-1 levels away from the CPU. A 1st Level cache, (also called an internal, or L1 cache) is usually built into the processor chip itself, while 2nd level caches (also called external, or L2 caches) and higher level caches are usually outside the processor chip. Some systems may have levels beyond L2. Each cache level away from the processor (larger value of n) is typically larger, but slower, than levels closer to the processor….»

Hard drive, Ethernet and other

More and faster is better. Any other hardware simply feed the previous ones. If you have complex parts/assemblies that push the CPU to 100% you probably not need 10 000 rpm hardrive a 7200 will do the job. Invest in faster CPU instead. For network card go with a 1000 (10/100 is prehistoric) wireless, go with the N standard dual band

Conclusion

Everything is question of balance and harmony.  If you decide for a brand system, a notebook or station, you build yourself or make it build, follow the steps of the budget has a guideline. If you have specials needs CAM/FEA you probably need more CPU, go with a lower end video card.

Before complain about the system, make sure you adopt good practices. This will be less expensive and will procure recurrent economies.

Personal experience allows to predict and ROI of 200@400 pc/year for a period around 5@ 8 years service. For larger corporation we talk about 3@5 years because pc starts with CAD then move on to other departments and finish with the secretary. Finally, the corporation gives gifts at Christmas….

Do not try to save 25-50 or even 100/200 dollars on a system. Remember that the designer earn between 500@1000 per week

Tools does not make good workers, it help them make good work. Investing in training will give better return.

No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: